PPC & Advertising

Create a Sponsored Products Campaign

Sponsored Products is the ad type that matters most. It is the majority of most accounts' spend, it looks like an organic result, and it is the one that feeds organic rank. Creating one takes about five minutes. Structuring it so you can still read it in three months takes a little thought, and that is what this covers. Watch the walkthrough Prefer to watch? One hundred and thirty-eight seconds on automatic versus manual targeting, the three bidding strategies, and the structure that survives contact with a real account. Before you open the console Two decisions made now save a rebuild later. One match type per campaign. Exact, phrase and broad each get their own. Mixing them means you can never tell which earned the sale, and you cannot bid them apart. One theme per ad group. Products that need a similar bid belong together. A £12 item and a £60 item cannot share a sensible bid. Creating the campaign Amazon Ads console → Campaign manager → Create campaign → Sponsored Products. 1. Name it so it reads itself. Something like Brand — Product — SP — Exact. In six months the name is the only context you have. 2. Set the daily budget. Start lower than feels right. You can raise a budget in seconds; you cannot un-spend. Sites: Amazon and beyond, or Amazon Business only 3. Choose where the ads run. Amazon and beyond is the default and covers Amazon retail, Amazon Business and selected sites off Amazon. Amazon Business restricts the campaign to business shoppers only. Leave it on the default unless you are deliberately running a B2B strategy. 4. Choose the products. One product, or a tight group that shares a price band. Automatic or manual targeting 5. Pick your targeting. Automatic targeting lets Amazon choose, across four buckets you can bid separately: close match, loose match, substitutes, complements. Cheap discovery, imprecise. Manual keyword targeting is your keywords at your bids, in one match type. Manual product targeting aims at ASINs or categories — exact for a specific ASIN, expanded to reach similar products. The three campaign bidding strategies 6. Choose a bidding strategy. | Strategy | What Amazon does | When | |---|---|---| | Down only | lowers the bid when a sale looks unlikely | start here | | Up and down | may raise your bid up to 100% | proven exact campaigns only | | Fixed | nothing | when you want clean data | Up-and-down hands Amazon the decision about how much of your money to spend. It is not wrong, but it is not a starting position. Both negative keyword and negative product targeting are available here, at creation. Most people discover them months later, after paying for the drift. Bid adjustments by placement and by audience 7. Bid adjustments. Amazon lets you raise bids by placement — top of search (first page), rest of search, product pages — and, more recently, by audience. Leave them at zero to begin. You cannot set a sensible multiplier before you have data, and you will have it in a fortnight. See Audit Your Placement Multipliers. A structure that survives The pattern most accounts converge on: An auto campaign on a small budget, running permanently as a discovery engine. A broad or phrase campaign for controlled expansion. An exact campaign holding the proven terms, taking the bulk of the budget. Terms are promoted upward as they prove themselves, and negated in the campaign they came from so you stop bidding against yourself. That routine is Harvest Winning Search Terms into Exact-Match Campaigns. Then judge it in TrackIQ Creating the campaign is the easy half. This is the half the console makes hard, because it scatters the answer across screens and resets your date range. Campaigns, with spend, ACoS and conversion in one table Amazon Ads → Campaigns puts every campaign in one table with consistent columns and a date range that holds. See Campaigns: Budgets, Placement and Audience Adjustments. Wait before judging. Sponsored Products attributes over 7 days, so the last week always understates. A campaign is not readable for a fortnight. Then, in order: 1. Is it spending? No spend usually means the bid is under the going rate, or the budget caps at 9am. 2. Is it converting? Clicks with no orders is a listing problem, not a bid problem. Bidding up buys more of the same disappointment. Search terms — what shoppers actually typed 3. What did it actually run on? Amazon Ads → Search Terms shows the phrases shoppers typed, which under auto and broad are rarely your keywords. This is where harvest candidates and negative candidates both come from. See Search Terms and Aggregated Search Terms. 4. Where did it appear? Placements splits Top of Search from Rest of Search and Product Pages, which behave like three different products. Common mistakes Mixing match types, as above. The most expensive structural error available. Judging in three days. Attribution has not landed. Raising bids on a conversion problem. More expensive traffic to a page that is not converting. Never negating. Auto and broad drift. Without negatives you pay for the drift indefinitely. Budget capping early. A campaign that spends its budget by mid-morning is not being optimised, it is being switched off for most of the day. Related - Amazon PPC: Core Terms and Metrics — the vocabulary this assumes. - Keywords and Match Types — reading them once live. - Cut Wasted Ad Spend in an Afternoon — the first optimisation pass. - Diagnose an ACoS Spike in Ten Minutes — when it moves against you. - Profit Campaigns and Ranking Campaigns — deciding what the campaign is for before you build it. - Restructure Your Campaigns Before You Optimise — the campaigns every product should have, and what to name them. - Easy-Win Campaigns — six cheap structures with exact starting bids.

Create a Sponsored Brands Campaign

Sponsored Brands is the banner at the top of search — your logo, a headline, and several products. It sells the brand rather than one item, and it is the only ad type that can send a shopper to your storefront instead of a product page. It also needs Brand Registry, so if the option is missing that is why. Watch the walkthrough Prefer to watch? One hundred and thirty-two seconds on the three goals and what each actually buys, the ad formats, and the measure that is fairer to a brand banner than ACoS. Creating the campaign Amazon Ads console → Campaign manager → Create campaign → Sponsored Brands. The three Sponsored Brands goals, priced differently 1. Choose the goal. Three are offered, and they are priced completely differently: | Goal | Success metric | You pay for | |---|---|---| | Drive page visits | Clicks | cost per click (CPC) | | Grow brand impression share | Top-of-search impression share | 1,000 viewable impressions (vCPM) | | Reserve share of voice | Top-of-search impression share | a fixed price | Choose drive page visits almost every time. It bills per click, which is measurable, and every optimisation covered here assumes it. The other two buy placement, not response. Impression share is a visibility target, so you are committing budget to appear at the top of search whether or not anyone acts. That is a brand-awareness purchase with far weaker attribution, and it is not where to learn what works. 2. Name, dates, budget. No end date unless the promotion genuinely ends. Daily budget conservative. 3. Amazon or Amazon Business. Leave it on Amazon unless you are genuinely B2B-heavy, in which case a separate Business-targeted campaign is worth running. Collections, Store spotlight and Video 4. Pick the format. Collections — several products, landing on a page you compose. The usual choice. You then pick a collection type: automatic, where Amazon assembles the products from your catalogue against your keywords, or manual, where you choose them. Start manual. Automatic is convenient, but it decides what shoppers see and you lose the ability to say why a banner worked. Store spotlight — sends shoppers into your storefront. Good for a broad brand, weaker for a single hero product. Video — the same campaign type with a video creative. Covered separately in Create a Sponsored Brands Video Campaign. 5. Bidding. Turn off Amazon's bid optimisation and control your own bids. You can add placement and audience adjustments later, once there is data to justify them. 6. Targeting. Keyword (exact, phrase, broad) or product (ASIN, category). Start with exact. Broad on a brand banner spends quickly and teaches you little. 7. The creative. This is where SB differs from everything else: Logo — to Amazon's stated pixel specification, showing your brand name. Headline — up to 50 characters. It is the only copy you control at the top of search, so it deserves more than a product name. Products — you will select three for display, drawn from the products on the campaign. Images — use them. A banner of bare product shots performs worse than one with considered imagery. Raising bids on an audience built in AMC 8. Bid adjustments. You can raise bids by placement, and by audience. The audience option is worth knowing about: alongside Amazon's own prebuilt audiences, you can raise bids on a custom audience you built in Amazon Marketing Cloud. That is the point where your own data starts steering your bids rather than Amazon's segments. See Understanding AMC Cloud: New-to-Brand and Path-to-Purchase Insights. Then submit. Sponsored Brands goes through creative moderation, so it is not live immediately — the button reads Submit for review, not Launch. Judging it in TrackIQ Campaign performance across every ad type Sponsored Brands appears alongside your other campaigns in Amazon Ads → Campaigns, with the same columns and the same date range. Three things to hold in mind that do not apply to Sponsored Products: Attribution is 14 days, not 7. Recent days understate more than you are used to. Do not compare its ACoS with Sponsored Products. They are doing different jobs. SB reaches people earlier, and its last-click ACoS will usually look worse. Judged only on that number, you will switch off the thing building the brand. It overlaps with Sponsored Products. One shopper can see both, and both can claim the sale. This is why the ad types cannot be added together — see Why Sponsored Products, Brands and Display Numbers Do Not Simply Add Up, and, if you suspect real cannibalisation, Spot Cannibalisation Between Sponsored Brands and Sponsored Products. The better measure: new customers Since SB works higher in the funnel, the honest question is not "what was the ACoS" but "did it bring people who did not know us". That is a new-to-brand question, and it is answered in AMC Cloud: Are these new customers? See NTB Customers: New vs Repeat, Counted as People. Did it cause branded search to rise? If people saw the banner and later searched your name, that is the banner working. See Brand Keywords: Proving Upper-Funnel Spend Works. Does it appear early in the path? See Path to Purchase: Which Ad Sequences Actually Convert. Common mistakes Choosing an impression-share goal because the numbers look impressive. Impressions are not sales, and neither is share of voice. A headline that repeats the product title. Fifty characters of prime placement, spent saying nothing. Judging on ACoS alone. It reliably under-credits upper-funnel work. Broad match from day one. Expensive discovery on your most visible placement. Related - Amazon PPC: Core Terms and Metrics — the vocabulary. - Create a Sponsored Brands Video Campaign — the same campaign type, video creative. - Prove Your Display Campaigns Created Demand — the method, applied to upper funnel.

Create a Sponsored Display Campaign

Sponsored Display is the odd one out. Sponsored Products and Sponsored Brands target searches. Sponsored Display targets people — shoppers who viewed something, bought something, or fit a behaviour Amazon has inferred — and it can follow them off Amazon. That difference changes how you judge it, which is where most Display budgets get cut by mistake. Watch the walkthrough Prefer to watch? One hundred and eighteen seconds on the optimisation strategy that quietly softens your numbers, the five targeting families, and the budget that gets cut by mistake. Creating the campaign Amazon Ads console → Campaign manager → Create campaign → Sponsored Display. 1. Name, dates, budget. Start the budget low — Display reaches broadly and can spend quickly. Something in the region of a quarter of what you would give a Sponsored Products campaign is a sensible opening. Sponsored Display ad format: image or video 2. Creative type: image or video. Either can land on a specific product or on your storefront. Product page is the usual choice; storefront suits a broad range. Reach, Page visits and Conversions 3. Optimisation strategy. Three, and the pricing differs: | Strategy | Cost type | Conversions attributed to | |---|---|---| | Reach | vCPM | viewable impressions and clicks | | Page visits | CPC | clicks | | Conversions | CPC | clicks | Optimise for conversions. This is the one to use. Avoid reach. It bills per thousand viewable impressions, and because it credits conversions to impressions as well as clicks, its reported results are far softer than they look. You will be buying exposure with no reliable way to tell whether it did anything. 4. Set your bids. Sponsored Display bids per target, so each audience you add gets its own bid in the targeting table. There is also an optional cost control setting, currently in beta, that lets Amazon manage bids to hold a metric at or below a value you name. Leave it off until the campaign has a baseline — it takes around 72 hours to settle, and you cannot read a campaign that is being steered while you are still learning what it does. The five Sponsored Display targeting families 5. Choose your targeting. Five families, and they behave nothing alike: | Family | Who it reaches | Realistic job | |---|---|---| | Contextual | shoppers on chosen products or categories | conquesting, defending | | Remarketing | people who viewed or purchased, over a lookback window | the warmest — start here | | In-market | people shopping the category now | reach, cold | | Interest and lifestyle | long-run behavioural groups | coldest, broadest | | Custom audiences | a segment you built yourself in AMC | precise, and yours alone | Start with views-remarketing. Smallest, warmest, cheapest to prove. Note the lookback window on each remarketing target — 30 days by default — which decides how far back a view still counts. Expand outward once it works. Custom audiences are the interesting one. They are built in Amazon Marketing Cloud rather than picked from a list, which means you can target a segment defined by your own data instead of one of Amazon's presets. See Understanding AMC Cloud: New-to-Brand and Path-to-Purchase Insights. 6. Build the creative. Headline to 50 characters, logo showing your brand name, image or video to Amazon's stated specification. Then launch. Judging it in TrackIQ Audiences, segmented by type Amazon Ads → Audiences reports one row per campaign per segment, with a Segment type column that groups them into those four families. See Audiences: Segment Performance. Compare within a family, never across it. Remarketing beating lifestyle on ACoS is not a finding, it is arithmetic. The warm audience was always going to win a last-click comparison. Judge each family against its own job: Remarketing should convert. If it does not, the problem is the listing, not the audience. In-market and lifestyle should bring new people. Held to a conversion target they will always fail, and cutting them removes the thing that fed remarketing its audience in the first place. Attribution is 14 days. Recent days understate more than Sponsored Products does. The measure that actually settles it For the cold families, last-click ACoS is the wrong instrument. The right question is whether they brought customers who did not know you. AMC Cloud → NTB Customers gives new versus repeat as people, not orders. See NTB Customers: New vs Repeat, Counted as People. AMC Cloud → Path to Purchase shows whether Display appeared first in the journey or merely intercepted a sale already in progress. That distinction is the whole argument. See Path to Purchase: Which Ad Sequences Actually Convert. The full method is Prove Your Display Campaigns Created Demand. Sponsored Display is not Amazon DSP They are frequently confused. Sponsored Display is self-service, sits in the Ads console beside your other campaigns, and reports under Amazon Ads → Audiences. Amazon DSP is a separate programmatic platform with its own console, its own audiences and its own reporting, under Amazon DSP in TrackIQ. See DSP Audience Segments. Common mistakes Judging cold audiences on ACoS. The most common way a working Display budget gets cut. Starting with lifestyle audiences because the reach numbers are large. Start warm. Adding segment sales together. Segments overlap; one shopper can sit in several. Reading a segment with forty clicks. Not enough to mean anything. Choosing the reach objective. You will not be able to prove it worked. Related - Audiences: Segment Performance — the screen, in full. - Shift Budget from Harvesting to Acquisition — when efficiency has crowded out growth. - What 'New to Brand' Actually Means — defined differently in different places.

Create a Sponsored Brands Video Campaign

Sponsored Brands Video is a Sponsored Brands campaign with a video creative instead of a banner. It is not a separate ad type, which matters: it lives in the same campaign manager, uses the same targeting, and reports as Sponsored Brands. It is also, for many sellers, the best-performing format they run — an autoplaying video in the middle of search results is hard to scroll past. Watch the walkthrough Prefer to watch? One hundred and thirteen seconds on why it is a format rather than an ad type, what a silent autoplaying ad has to do in two seconds, and the metric that moves first. Creating the campaign Amazon Ads console → Campaign manager → Create campaign → Sponsored Brands. 1. Goal: drive page visits. Cost per click, measurable. As with any Sponsored Brands campaign, avoid the two impression-share goals — they buy placement rather than response. 2. Name, dates, budget. No end date unless it genuinely ends. 3. Amazon or Amazon Business. Amazon unless you are deliberately targeting B2B. Ad format: Video, alongside Collections and Store spotlight 4. Choose the video format and upload your creative. 5. Destination. The video can land on the product detail page or your storefront. Send it to the product page. Someone who watched a video about one product should arrive at that product, not a directory. 6. Bidding. Turn off Amazon's optimisation, set your own bids, add placement adjustments only once you have data. 7. Targeting. Keyword (exact, phrase, broad) or product (ASIN, category). Start with exact, same as any Sponsored Brands campaign. 8. Upload the video to Amazon's stated specification for length, dimensions and file size, then submit for review. Video creative goes through moderation, so it will not be live immediately. What makes the creative work The mechanics are easy; the video is the whole campaign. It autoplays silently. If it needs sound to make sense, it does not make sense. Caption anything essential. The first two seconds decide it. Lead with the product doing the thing it does. A logo animation spends your most valuable moment on nothing. Show use, not features. Video's advantage over a static banner is demonstration. If your video is a slideshow of stills, you have bought a video slot to run a banner. Keep it short. Attention is gone long before a long video ends. Judging it in TrackIQ Campaigns, with video alongside every other ad type SBV reports as Sponsored Brands in Amazon Ads → Campaigns, so name it clearly at creation — something with SBV in it — or you will not be able to separate video from banner performance later. Compare video against your banner Sponsored Brands campaigns, not against Sponsored Products. Same ad type, same attribution window, same job. That is the comparison that answers whether the video was worth producing. Attribution is 14 days. Recent days understate. Watch CTR first. It is the fastest signal on whether the creative is working, and it moves long before ACoS becomes readable. A video with weak CTR is a creative problem, and no bid change will fix it. The upper-funnel measure Video reaches people earlier than a search ad, so last-click ACoS under-credits it in the same way it under-credits Sponsored Brands generally. Did it bring new customers? NTB Customers: New vs Repeat, Counted as People. Did branded search rise after it ran? Brand Keywords: Proving Upper-Funnel Spend Works. Common mistakes A video that needs sound. It autoplays muted. Opening on a logo. The two seconds that mattered, spent on branding nobody is looking for yet. Sending traffic to the storefront after a single-product video. Not naming it distinctly, then being unable to tell video from banner in reporting. Judging it against Sponsored Products. Different job, different window, different funnel position. Related - Create a Sponsored Brands Campaign — the banner version, and the shared settings. - Amazon PPC: Core Terms and Metrics — the vocabulary. - Why Sponsored Products, Brands and Display Numbers Do Not Simply Add Up — before you total anything. - Amazon's AI Video Generator — producing the creative these campaigns need.

Amazon PPC: Core Terms and Metrics

Amazon advertising has a vocabulary problem. The same idea has three names depending on which screen you are on, and a few terms mean something different from what they sound like. This is the vocabulary the rest of the training assumes. Learn it once and every later article gets easier. Watch the walkthrough Prefer to watch? One hundred and seventeen seconds on the four campaign types, the three match types, and the two attribution windows that decide how everything reads. Create campaign: the ad type you pick first The campaign types Amazon's campaign chooser offers four self-service types, plus one in beta. Three of them account for nearly all of most sellers' spend. Sponsored Products (SP) — a single product, shown in search results and on product pages, looking much like an organic result. This is the one that most directly moves organic rank for a keyword, and for most sellers it is the majority of spend. It now also offers Amazon Business and Across retailers variants, which place the same ads in front of business buyers and beyond Amazon respectively. Sponsored Brands (SB) — a banner with your logo, a headline and several products, at the top of search. It sells the brand rather than one item. Sponsored Brands Video is a format inside SB, not a separate type. Display (SD) — targets people rather than searches, on and off Amazon. Retargeting and audience work lives here. Amazon labels it simply Display in the chooser, though it is still Sponsored Display everywhere else. Streaming TV — video ads on premium streaming such as Prime Video, sold self-service with no minimum spend. It is a brand-awareness buy, and it is the newest of the four. Full-Funnel Campaigns, currently in beta, is not a fifth ad type so much as an orchestrator: it runs display, video and sponsored formats together and decides the mix for you. A note on what this training covers. SP, SB and SD are what the rest of these articles mean by "the three ad types" — they share the sponsored-ads reporting, and they are what TrackIQ ingests under Amazon Ads. Streaming TV is bought and measured more like Amazon DSP than like a sponsored campaign, so it is judged on reach and new-to-brand rather than ACoS. Targeting, and why match types matter Sponsored Products has two targeting modes, and the distinction runs through everything else. Automatic targeting — Amazon decides what to show you against, in four buckets: close match, loose match, substitutes and complements. Cheap to launch, useful for discovery, imprecise. Manual targeting — you decide, either by keyword or by product. Manual keyword targeting has three match types: | Match type | Triggers on | Use it for | |---|---|---| | Exact | that phrase, near enough | your proven winners | | Phrase | the phrase inside a longer search | controlled expansion | | Broad | the words in any order, plus related | discovery | Manual product targeting comes in two shapes: exact (a specific ASIN, or a category) and expanded, which reaches out to products Amazon judges similar. Do not mix match types in one campaign. It is the single most common structural mistake. Once broad and exact share a campaign you can no longer tell which one earned the sale, and you cannot bid them differently, which was the entire point of separating them. Where the ad appears: placements Top of Search — the first row of results. Converts best, costs most. Rest of Search — further down. Cheaper, usually most of your impressions. Product Pages — on someone else's detail page. A different intent entirely: the shopper was looking at something else. Amazon lets you bid these differently with placement multipliers. See Placements: Top of Search vs Rest of Search. Bidding strategies Three, and the names undersell the risk. Dynamic bids — up and down. Amazon may raise your bid by as much as 100% when it predicts a conversion. That is Amazon deciding how much of your money to spend. Dynamic bids — down only. Amazon lowers the bid when a conversion looks unlikely, never raises it. The safe default. Fixed bids. Your bid, always. Useful when you are deliberately gathering clean data. Start on down-only unless you have a specific reason not to. Up-and-down can work on proven exact-match campaigns; it is a poor place to begin. The metrics, and the two that get confused The metric columns as TrackIQ reports them | Term | Is | Watch for | |---|---|---| | Impressions | times shown | not a goal | | Clicks / CTR | clicks, and clicks ÷ impressions | low CTR is usually the image or price | | CPC | spend ÷ clicks | rises with competition | | Orders / CVR | orders, and orders ÷ clicks | low CVR is the listing, not the ad | | Spend / Sales | money out, ad-attributed money in | | | ACoS | spend ÷ ad sales | judges the advertising | | TACoS | spend ÷ total sales | judges the business | | ROAS | ad sales ÷ spend | the inverse of ACoS | ACoS and TACoS are the pair people mix up. ACoS asks whether the ads are efficient. TACoS asks how dependent the business has become on them. A campaign can hold a fine ACoS while TACoS climbs, which means advertising is buying a larger share of a shrinking organic base. Full definitions in Glossary: ACoS, TACoS, ROAS, NTB and Every Other Column and Understanding ROAS and Key Amazon Advertising Metrics. Attribution: why nothing adds up A click today can produce a sale days later, and Amazon credits it backwards to the click. Sponsored Products: 7 days. Sponsored Brands, Display and DSP: 14 days. Two consequences you will meet constantly: Recent days always look worse than they were. Attribution has not finished landing. Judge nothing on the last few days. The sponsored ad types cannot simply be added together. They use different windows and one shopper can touch several ads. See Why Sponsored Products, Brands and Display Numbers Do Not Simply Add Up. Keyword or search term? Worth separating now, because it causes trouble later. A keyword is what you chose to bid on. A search term is what a shopper actually typed. Under broad and auto targeting they are frequently different phrases. Everything you learn about harvesting depends on that gap. See Search Terms and Aggregated Search Terms. Where to see all of this Campaigns are built in Amazon's Ads console. They are read in TrackIQ, which pulls every ad type into one place with the same columns and date controls: Amazon Ads → Campaigns for the top level, then Ad Groups, Keywords, Targets, Product Ads, Search Terms, Placements and Audiences beneath it. See Monitoring Your Sponsored Ads Performance. Related - Keywords and Match Types — match types in depth. - Understanding ROAS and Key Amazon Advertising Metrics — the metrics in depth. - Time to Conversion: Checking Your Attribution Window — measuring your real window. - What Amazon PPC Is Actually For — why any of this matters to profit. - Profit Campaigns and Ranking Campaigns — the two jobs a campaign can have.

What Amazon PPC Is Actually For

Most Amazon brands have three things holding them up. The products themselves. The inventory to sell them. And advertising, which is what makes anyone see either. Get the first two right and a bad advertising account will still cap your growth. Get advertising right and it becomes the single largest lever you have on profit. Watch the walkthrough Prefer to watch? One hundred and four seconds on why advertising and organic rank are the same conversation, and why efficiency has to come before ranking. What advertising is actually doing An Amazon listing does not get discovered on merit. It gets discovered because it ranks, and it ranks largely because it sells. Advertising is how you buy the first sales that start that loop. So sponsored ads do two jobs at once: They sell directly. Someone searches, sees your ad, buys. They raise organic rank. Those sales feed the sales velocity that decides where you sit organically for that search — which then brings sales you did not pay for. That second job is the one people underrate. A campaign judged only on its own ACoS is being judged on half of what it did. Growing the brand means growing profit This is worth being blunt about, because it changes every decision that follows. Revenue is easy to buy. You can double ad spend tomorrow and revenue will rise. Whether you kept any of it is a different question. The number that matters is profit, and advertising affects it from both directions: it adds sales, and it is a cost. A brand can grow revenue for a year while its advertising quietly eats the margin that growth was supposed to produce. Which is why TACoS — ad spend as a share of total sales — is the honest measure of the account. ACoS tells you whether a campaign is efficient. TACoS tells you whether the business is. Advertising performance, at account level Efficiency first, ranking second There are two distinct pieces of work, and doing them in the wrong order wastes a lot of money. Stage one is making existing spend efficient. Most accounts are carrying targets that have not converted in months, placements bid above what they return, and search terms that were never turned into keywords. None of that requires new budget to fix. It requires finding it. Stage two is ranking, which means deliberately buying traffic on chosen keywords to move organic position, accepting a worse short-term ACoS in exchange for a position that keeps paying afterwards. Do stage one first. Ranking campaigns are expensive by design, and running them on top of an account that is already leaking means you cannot tell which cost was the investment and which was the waste. Fix the leak, then open the tap. What that looks like in practice The work is unglamorous and mostly consists of reading the same handful of reports on a schedule: What spent and did not convert. See Cut Wasted Ad Spend in an Afternoon. What converted and is not yet a keyword. See Harvest Winning Search Terms into Exact-Match Campaigns. Where the ads appeared, and whether the bid matched. See Audit Your Placement Multipliers. How spend is distributed across ad types and placements. Done every week or two, that is most of the job. The reason accounts drift is not that the work is hard, it is that nobody is scheduled to do it. Why the reports matter more than the tactics Advertising advice tends to arrive as tactics — a bid rule, a match type, a structure. Tactics are fine, but they only help if you can see what is happening. An account you cannot read is an account you cannot improve, and the console makes reading it harder than it needs to be: the answer is scattered across screens, the date range resets, and the ad types report on different attribution windows. That is the gap TrackIQ exists to close. Every screen in Amazon Ads holds the date range you set, keeps consistent columns, and puts campaigns, keywords, search terms and placements in one place so a fortnightly review takes minutes rather than an afternoon of exports. Where to start Score what you have before changing anything. Score Your PPC Performance walks through eight reports that tell you, numerically, which parts of the account are being managed and which are not. Then fix the lowest scores first. That is almost always where the money is. Related - Profit Campaigns and Ranking Campaigns — the two jobs a campaign can have, and why they are judged differently. - Amazon PPC: Core Terms and Metrics — the vocabulary this assumes. - Understanding ROAS and Key Amazon Advertising Metrics — the metrics, defined. - The Nine-Report PPC Routine — the routine that keeps existing spend efficient. - Setting Target TACoS per Product — turning the TACoS argument into a number per product. - Before You Add a Product — maximising what you have before adding more.

Profit Campaigns and Ranking Campaigns

Every campaign you run is doing one of two jobs. Either it is trying to make money now, or it is trying to buy organic position for later. They look identical in the console. They are judged by completely different numbers, and confusing them is why perfectly good campaigns get switched off. Watch the walkthrough Prefer to watch? One hundred and fourteen seconds on the two jobs a campaign can have, the KPIs that belong to each, and the two ways people confuse them. Profit campaigns Purpose: sell efficiently. These are the majority of your account — usually the large majority of spend — and they are supposed to be profitable on their own terms. What lives here: auto campaigns, research campaigns, your exact-match workhorses, ASIN-targeting campaigns, and most Sponsored Brands and Sponsored Display. The KPIs are ACoS and TACoS. ACoS tells you whether the campaign is efficient. TACoS tells you whether the account is affordable. The goal is not the lowest possible ACoS. It is the most sales you can generate while staying inside your TACoS target. Those are different objectives, and chasing the first one shrinks the account. An ACoS of 8% sounds excellent until you notice it was achieved by bidding so low that the campaign barely spends. A campaign that spends £4,000 at 28% may be worth far more than one that spends £300 at 12%, provided 28% still clears your margin. What "profitable" means depends on your gross margin and your niche. A 35% ACoS is comfortable on a 60%-margin product and ruinous on a 25%-margin one. The more competitive the category, the harder profitable gets — which is a fact about the category, not a failure of the campaign. Campaigns, with spend, ACoS and conversion in one table Ranking campaigns Purpose: move organic position on chosen keywords. These are deliberately not trying to be profitable in the short term. You are buying sales velocity on a specific search, because velocity is what moves organic rank, and organic rank keeps paying after you stop. They have a recognisable shape: Sponsored Products only. It is the ad type most directly tied to keyword-level rank. Exact match only. You are ranking for a specific phrase, so you cannot afford the traffic to wander. Top of Search placement, aggressively. That is where the velocity is. Short-lived. A ranking campaign is a push, not a permanent fixture. The KPIs are different, and this is the important part. You judge these on conversion rate, click-through rate, and sales velocity on the target keyword — not on ACoS. ACoS on a ranking campaign will look bad, because it is supposed to. You are paying above the efficient rate on purpose. When it works, you stop. Once the keyword holds the organic position you wanted, you wind the campaign down and move the budget to the next keyword. Ranking campaigns that are still running six months later have usually stopped being ranking campaigns and become expensive profit campaigns nobody re-labelled. Why mixing them up is expensive Two failure modes, and both are common. Judging a ranking campaign on ACoS and switching it off two weeks in. You paid the expensive part — the initial velocity — and quit before the organic position that was supposed to repay it had settled. Letting a profit campaign quietly become a ranking campaign. Bids creep up, Top of Search multipliers get raised, ACoS drifts, and nobody has decided that is the plan. The account is now buying rank by accident, at a price nobody agreed. The fix is naming. Put the intent in the campaign name at creation — the same discipline covered in Create a Sponsored Products Campaign. In six months the name is the only context you have, and "is this supposed to be profitable?" is the first question you will ask. Which to do first Get the profit campaigns efficient before you run ranking campaigns. Ranking campaigns are expensive by design. Run them on top of an account that is already leaking money into non-converting targets and mis-set placement bids, and you will not be able to tell which cost was the investment and which was the waste. Score the account first — Score Your PPC Performance — fix what the low scores expose, and only then start buying rank. Reading them apart in TrackIQ Because the two types are judged on different numbers, you want to look at different columns. For profit campaigns, Amazon Ads → Campaigns gives spend, sales, ACoS and conversion in one table with a date range that holds. See Campaigns: Budgets, Placement and Audience Adjustments. For ranking campaigns, the useful views are keyword-level and rank-level: Keywords and Match Types for CTR and CVR on the target phrase, and the organic rank tracking that tells you whether the position actually moved. If rank is not moving after a fortnight of real velocity, the problem is usually the listing rather than the bid. Related - What Amazon PPC Is Actually For — why advertising and organic rank are the same conversation. - Amazon PPC: Core Terms and Metrics — ACoS, TACoS and attribution. - Placements: Top of Search, Rest of Search and Product Pages — where ranking campaigns need to win. - Bid Optimisation: Six Rules — steering a profit campaign toward its target ACoS. - Setting Target TACoS per Product — the budget envelope profit campaigns work inside. - What a Relaunch Is, and When to Run One — the full process a ranking campaign belongs to. - Running a Launch: Stabilise, Scale, Scale Back — how a ranking campaign is actually driven.

Score Your PPC Performance

Before changing anything in an advertising account, it is worth knowing how well it is currently being run — not by impression, but with a number. This is a scoring exercise. Eight reports, each marked out of ten, giving a total out of eighty. It takes an afternoon, it is repeatable, and it tells you exactly which parts of the account have a process behind them and which do not. Watch the walkthrough Prefer to watch? Ninety-nine seconds on the eight reports, the benchmarks worth scoring against, and what the total actually tells you. What the score is actually measuring Each report answers one question: is somebody doing this job regularly? That is the whole trick. You are not scoring the account's ACoS, which depends heavily on your category and margin. You are scoring whether the routine maintenance is happening — because a large backlog of obvious, uncorrected problems is the reliable signal of an account nobody is working. A high score means few corrections are outstanding. A low score means the report found a lot to fix, which means nobody has been looking. The eight reports | # | Report | What a bad result looks like | |---|---|---| | 1 | Bleeders | many targets with 20+ clicks and no sales, last 65 days | | 2 | Bleeders 2.0 | many targets, terms or campaigns at extreme ACoS, last 30 days | | 3 | Lifetime bleeders | targets still enabled that have never converted, over the account's life | | 4 | Placement optimisation | placement bids that do not match placement performance | | 5 | Search term harvesting | many converting search terms never promoted to keywords | | 6 | Ad type mix | Sponsored Products below 80% of spend | | 7 | Placement mix | Top of Search below 60% of Sponsored Products spend | | 8 | Main keyword coverage | barely spending on the keywords that define your niche | Reports 1–3 are covered in The Bleeder Reports. Reports 4, 6 and 7 are covered in Ad Type and Placement Mix. Report 5 is Harvest Winning Search Terms. How to score each one Mark each report from 1 to 10, where 10 means "this report found essentially nothing to fix" and 1 means "this report found a great deal". Two of the eight have hard numeric thresholds, so score those against the benchmark rather than by judgement: Ad type mix. Sponsored Products should be 85–90% of sponsored spend. 80% is the floor. Below 80 scores poorly, and the reason is rank: SP is the ad type most directly tied to keyword-level organic position, so under-weighting it means buying sales without buying rank. Placement mix. Top of Search should take 60% or more of Sponsored Products spend. Below 60 scores poorly, for the same reason — Top of Search is where the velocity that moves rank actually happens. For the other six, the judgement is proportional: a handful of outstanding corrections on a large account is a good score; hundreds is not. Campaigns, spend and efficiency in one table Reading the total 50 or above, out of 80. The account has a working process. The right move is targeted: take the two or three reports that scored lowest and build a routine around those specifically. Do not rebuild what is already working. Below 50. The account is not being maintained in any systematic way. Individual fixes will not hold, because the same problems will regenerate within weeks. What is needed is a repeating schedule covering all eight, not a one-off cleanup. Any single report at 6 or below deserves attention regardless of the total. A strong total can hide one badly neglected area — most often lifetime bleeders, which nobody thinks to look for because the campaigns concerned are old and quiet. Doing it in TrackIQ The original version of this exercise involves exporting bulk files and rebuilding each report by hand in a spreadsheet. That is roughly a day's work per pass, which is why most accounts do it once and never again. Most of the eight are already screens: Bleeders and lifetime bleeders — Amazon Ads → Search Terms and Targets, sorted by spend with orders at zero, over whatever date range you choose. The date range holds, so "last 65 days" and "lifetime" are two clicks apart rather than two exports apart. Search terms, with spend and orders side by side Placement performance and bids — Amazon Ads → Placements, which puts Top of Search, Rest of Search and Product Pages beside their multipliers. Ad type mix — Amazon Ads → Campaigns, grouped by ad type, which gives the SP share of spend directly. Search term harvesting — Amazon Ads → Search Terms, filtered to converting terms. The point is not that the tool does the thinking. It is that the eight questions become an afternoon rather than a project, which is the difference between scoring the account once and scoring it every quarter. What to do next Fix the lowest scores first, in this order of likely payback: Bleeders, if they scored low. This is money already being spent on traffic that has never converted, and stopping it is immediate. See Cut Wasted Ad Spend in an Afternoon. Harvesting, if it scored low. Converting search terms sitting unpromoted are sales you are already paying for but not compounding. Mix, if it scored low. Rebalancing toward SP and Top of Search changes what your existing budget buys without needing more of it. Then re-score in a quarter. The number moving is the evidence that the routine is real. Related - The Bleeder Reports — the first three reports, in detail. - Ad Type and Placement Mix — the benchmarks, and why they are what they are. - Profit Campaigns and Ranking Campaigns — score profit campaigns; ranking campaigns are judged differently. - The Nine-Report PPC Routine — the same reports, on a repeating schedule. - Restructure Your Campaigns Before You Optimise — fixing the structure the score exposes. - Setting Target TACoS per Product — the profit target the score is ultimately serving. - Acting on Your PPC Score — what to do with the score once you have it.

The Bleeder Reports

A bleeder is a target or search term that takes clicks and returns nothing. Not one that converts poorly — one that has never converted at all. Every account accumulates them. They are the cheapest thing in advertising to fix, because fixing them requires no new budget, no creative work and no strategy: you are simply stopping payment for traffic that has already proven it does not buy. Watch the walkthrough Prefer to watch? One hundred and twelve seconds on the twenty-click threshold, the three windows that catch different bleeders, and why you negate and pause rather than one or the other. Three passes, not one The reason this is three reports rather than one is that the same problem hides at three different time scales, and a single date range misses two of them. | Report | Window | Finds | |---|---|---| | Bleeders | last 65 days | targets with meaningful clicks and zero orders | | Bleeders 2.0 | last 30 days | targets converting, but at ruinous ACoS | | Lifetime bleeders | account lifetime | old, quiet targets that have never converted | Bleeders: clicks with nothing to show The test: 20 or more clicks, zero orders, over 65 days. Twenty clicks is the threshold because it is roughly where "no orders" stops being bad luck. A target with six clicks and no sales tells you nothing. A target with forty tells you something real. Why 65 days rather than 30? Because Sponsored Products attributes over seven days, and low-volume targets need time to accumulate enough clicks to be judged at all. A 30-day window on a slow target produces a verdict on eight clicks, which is not a verdict. What to do with them. Add as a negative in the campaign that produced it, so you stop paying for it under auto and broad, then pause the keyword if it is one you added deliberately. Negating without pausing leaves you bidding against yourself later; pausing without negating means auto match finds it again next week. Search terms: what shoppers typed, and what it cost Twenty clicks, or fewer? Twenty is the bar for a judgement — enough evidence to pause something outright, and the threshold worth using when you are auditing whether an account is being maintained. The weekly routine is gentler, because it runs again in seven days. Rather than one cut-off it tapers, reducing bids before it pauses anything: | Clicks, no orders, last 30 days | Action | |---|---| | 6–7 | reduce bid by 5% | | 8–14 | reduce bid by 10% | | more than 14 | pause the target | Both are the same idea at different confidence levels. Use the taper weekly; use twenty clicks when you are deciding whether a process exists at all. And on a product converting below 10%, raise every one of these numbers — fourteen clicks without a sale is weak evidence when one in twenty clicks is a normal sale rate. Bleeders 2.0: converting, but not at that price The first report catches zeroes. This one catches the more insidious case: targets that do convert, occasionally, at an ACoS that is far past anything your margin supports. The test: extreme ACoS over the last 30 days, at target, search term and campaign level. "Extreme" is relative to your break-even ACoS, which is set by your gross margin. If you clear 40% margin, your break-even is roughly 40% ACoS — anything sustained well beyond that is losing money on every sale it makes. Thirty days, not 65, because these targets have volume. They are spending enough to be judged quickly, and the longer you leave them the more they cost. What to do with them. These are usually a bid problem rather than a relevance problem: the target is right but you are paying too much to win it. Cut the bid substantially and watch for a fortnight before deciding to remove it. If it still bleeds at a sensible bid, then it is a relevance problem after all. The exception worth remembering: a high ACoS on a target you deliberately chose for ranking is not a bleeder. It is the plan. See Profit Campaigns and Ranking Campaigns. Lifetime bleeders: the ones nobody looks for This is the report that most often surprises people, and it is the one most often missing entirely. The test: still enabled, and has never produced a sale, across the account's whole history. These are not new mistakes. They are keywords added during a launch two years ago, in campaigns that still run on small budgets, quietly spending a few pounds a week each. Individually invisible. Collectively, often thousands a year. They survive because every routine review uses a recent date range, and a target spending £3 a week never appears near the top of a 30-day sort. You only find them by asking the lifetime question explicitly. What to do with them. Pause, and negate. There is no diagnostic subtlety here — a target with hundreds of lifetime clicks and no lifetime orders is not going to start converting. Running all three in TrackIQ The hand-built version of these reports means exporting bulk sheets and pivoting them, once per window. The reason lifetime bleeders get skipped is that it is the most tedious export of the three. In Amazon Ads → Search Terms and Amazon Ads → Targets, all three are the same operation with a different date range: Set the date range. 65 days, 30 days, or the widest range available for the lifetime pass. The range holds while you work, so you are not re-setting it on every screen. Sort by spend, descending. The expensive problems surface first. Read the orders column. Zero orders with meaningful spend is report one and report three. High spend with a handful of orders is report two. Targets and their spend, orders and ACoS Because the same screen answers all three questions, the lifetime pass costs you a date-range change rather than an afternoon — which is the difference between a report that gets run and one that does not. For the full walkthrough of the first pass, see Cut Wasted Ad Spend in an Afternoon. How often Bleeders and Bleeders 2.0: fortnightly. Frequent enough that nothing runs up a large bill, spaced enough that attribution has landed. Lifetime bleeders: quarterly. They accumulate slowly, and a quarterly pass catches them before they compound. Automating the first two is reasonable once you trust the thresholds — see Rule-Based Optimization, and read Guardrails Before You Automate first. Common mistakes Judging on too few clicks. Under about twenty, "no orders" is noise. Negating without pausing, or pausing without negating. Do both. Using a 30-day window for everything. It hides exactly the targets that have been bleeding longest. Treating a ranking campaign's high ACoS as a bleeder. Check the campaign's purpose before you cut it. Cutting a cold Sponsored Display audience for a poor ACoS. Different job, different measure — see Create a Sponsored Display Campaign. Related - Score Your PPC Performance — where these three reports sit in the wider score. - Cut Wasted Ad Spend in an Afternoon — the first pass, step by step. - Search Terms and Aggregated Search Terms — reading the screen these reports come from. - Bid Optimisation: Six Rules — the graduated bid taper that runs weekly. - The Nine-Report PPC Routine — where the three passes sit in the schedule.

Ad Type and Placement Mix

Two questions about an advertising account can be answered without knowing anything about the products, the category or the margin: What share of spend is going to Sponsored Products? and what share of that is going to Top of Search? Both have benchmarks. Both are usually wrong in accounts that have grown without a plan, and both change what your existing budget buys without needing any more of it. Watch the walkthrough Prefer to watch? One hundred and twenty-six seconds on why Sponsored Products should dominate, why Top of Search should take the majority, and the multiplier check almost nobody runs. Ad type mix: Sponsored Products should dominate The benchmark: Sponsored Products should be 85–90% of sponsored spend. 80% is the floor. The reason is rank. Of the sponsored ad types, SP is the one most directly tied to keyword-level organic position — it targets a search, appears in the results for that search, and its sales feed the velocity that decides where you rank organically for it. Sponsored Brands and Sponsored Display do valuable work, but they do not move keyword rank the same way. SB sends traffic to a storefront or a multi-product page, so the keyword-to-product signal is diluted. SD is not keyword-targeted at all. So an account spending 40% of its budget on SB and SD is buying sales without buying rank. It may look efficient. It is not compounding. What "wrong" usually looks like: SB creeping up because the banners are satisfying to look at and their reported sales are flattered by a 14-day attribution window, while SP is capped by budgets nobody has revisited. Reading it: in Amazon Ads → Campaigns, group by ad type and compare spend. One screen, one date range, no export. Every campaign and ad type in one table Before you rebalance, do not simply add the ad types together and treat the total as truth. They use different attribution windows and one shopper can touch several ads. See Why Sponsored Products, Brands and Display Numbers Do Not Simply Add Up. Placement mix: Top of Search should take the majority The benchmark: Top of Search should be 60% or more of Sponsored Products spend. Sponsored Products appears in three places, and they behave like three different products: | Placement | What it is | Typically | |---|---|---| | Top of Search | the first row of results | highest CVR, highest CPC, drives rank | | Rest of Search | further down the results page | moderate everything | | Product Pages | on competitors' detail pages | lowest CVR, cheapest clicks | Top of Search converts best because intent is highest there, and it is where the sales velocity that moves organic rank actually happens. An account with most of its SP spend on Product Pages is buying the cheapest clicks rather than the most valuable ones. Below 60% is a signal to investigate, not an automatic fault — some categories genuinely reward product-page conquesting. But it should be a decision, not an accident. Placements, with performance beside the multipliers The related check: do the bids match the performance? This is the third question, and it catches a specific, common error. Placement multipliers raise your bid for a given placement. The check is simply: does the placement getting the biggest multiplier also have the best performance? Surprisingly often it does not. A campaign carries a +50% Top of Search multiplier set during a launch eighteen months ago, while Top of Search has since become its worst-converting placement. Nobody looked, because nothing broke loudly. Focus on the top-spending campaigns first. A mis-set multiplier on a campaign spending £40 a month is not worth the attention; the same error on your largest campaign is. And the least-performing placement should be getting the least spend. If it is not, the multipliers are working against you. The full routine is Audit Your Placement Multipliers. Where the multipliers are set Placement adjustments are set at campaign creation and editable afterwards. Amazon offers adjustments for top of search (first page), rest of search and product pages, and more recently by audience as well. Bid adjustments by placement and audience Leave them at zero on a new campaign. You cannot set a sensible multiplier before you have data, and you will have data in a fortnight. See Create a Sponsored Products Campaign. Fixing the mix Do not move budget in one step. These are large structural shifts and the account needs time to re-equilibrate. Raise SP budgets before cutting SB and SD. If SP campaigns are budget-capped, the extra spend has somewhere useful to go immediately. Cutting first just reduces total spend. Then trim the ad types that are over-weighted, starting with anything bought on an impression-share or reach basis, which is the least measurable spend in the account. Move placement spend with multipliers, not budgets. Raising the Top of Search multiplier shifts the mix within a campaign; changing budgets shifts it between campaigns, which is a blunter instrument. Re-measure after a fortnight. Attribution needs to land before the new mix means anything. Common mistakes Chasing the cheapest clicks. Product Pages will always look cheap per click and rarely wins on value per click. Rebalancing on last-click ACoS alone. SB and SD reach people earlier, and last-click under-credits both. Cut them on mix, not on ACoS. Setting multipliers from a hunch. They should follow measured placement performance on that campaign. Fixing the mix while bleeders are still running. Rebalancing an account that is leaking just redistributes the leak. Run The Bleeder Reports first. Related - Score Your PPC Performance — where these benchmarks sit in the wider score. - Placements: Top of Search, Rest of Search and Product Pages — the screen, in full. - Audit Your Placement Multipliers — the bid-versus-performance check, step by step. - Placement Bid Optimisation — the weekly bid-level version of the same question. - The Nine-Report PPC Routine — the schedule these monthly reports belong to. - Scaling Profit Campaigns — growing spend once the mix is right.

Restructure Your Campaigns Before You Optimise

Optimising a badly structured account is like tuning an engine with a part missing. You can spend months adjusting bids on campaigns that were never going to give you a clean answer, because the structure itself makes the data unreadable. So before any of the optimisation routines, there is a foundation pass: work out what campaigns you have, what you are missing, and whether the ones you have can be read at all. Watch the walkthrough Prefer to watch? One hundred seconds on the structural rule that makes every downstream report trustworthy, the campaigns you are probably missing, and the launch check almost nobody does. The structural rules Two rules, and the first is not negotiable. One parent ASIN per campaign. This is the rule that matters. A campaign advertising several parent ASINs produces numbers that cannot be attributed to any one product, so you can never tell which product the spend was working for. Every report downstream inherits that ambiguity. One child ASIN per campaign, where practical. Recommended rather than required. Variations often differ enough in price and conversion that averaging them hides the truth. Group every campaign for a parent ASIN under one portfolio. This is what makes product-level budget management possible at all. If your account breaks the first rule, fix that before anything else. Nothing downstream is trustworthy until you do. Get the listing SEO-ready first Advertising and organic search are not separate systems, and a campaign pointed at a listing that does not mention the keyword will underperform for reasons no bid can fix. Before launching campaigns, for each product: Identify the top 3–5 keywords by search volume and relevance, and get them into the title. Identify the top 10–15 keywords, and make sure each appears at least once in the title, bullets or description. Put everything else in the back-end search terms. One caution on titles. A title change affects organic ranking and conversion as well as advertising, and a bad one can undo months of progress. Change titles deliberately, one at a time, with a before-and-after to measure against. The campaigns every product should have Once the structure is clean, the question is what is missing. | Campaign type | What it is for | |---|---| | Exact match | proven keywords, high exposure, ranking pressure | | Broad or phrase | research and profit — finding new search terms | | Auto | discovery of terms closely tied to the listing | | Product targeting (PAT) | competitor ASINs where you convert well | | Single-keyword exact | your biggest keyword, isolated for maximum control | | Easy wins | simple campaigns that reliably turn a profit | The single-keyword campaign is the one people skip. Isolating your largest keyword into its own campaign gives you budget and bid control over the term that matters most, instead of letting it compete for budget with thirty others. On product targeting, aim at two kinds of ASIN: the best-sellers in your niche, and weaker competitors — higher priced, fewer reviews, poorer listings — where you compare well. Naming, so the account reads itself In six months the campaign name is the only context you have. A consistent pattern makes the account sortable and filterable. Product — Ad type — Targeting — Match type — Placement note For example: Product X — SP — KWs — Exact — TOS Only For single-keyword campaigns, include the keyword itself: Product X — SP — 1KW — "keyword" — Exact — TOS Note anything unusual in the name — exact-only, Top of Search only, dynamic bidding. Those are exactly the details you will want to filter on later, and exactly the ones you will not remember. Campaigns, where a naming system either helps or does not Budgets Budget depends on the product's stage and its current TACoS, so there is no single right answer. But there is a workable default split: | Campaign group | Share of PPC budget | |---|---| | Manual Sponsored Products | 70–80% | | Product targeting | 10–20% | | Auto | under 10% | Start each campaign at a level that can actually produce data. A campaign on a trivial budget will take months to tell you anything, and you will make decisions on eight clicks. Launching without losing control Launch 1–3 campaigns a week if you want to keep spend under control, or all at once if you are comfortable with the step up. Then check daily that they spend. New campaigns very often do not, because the opening bid is below the going rate. This is the single most common reason a launch appears to do nothing. If yesterday's spend was under 30% of budget, raise all target bids by 10% — and raise placement bids too if you have set any. Repeat daily until it spends. Do your top 3–5 products first. Get the structure right where the money is, then apply the same pattern outward. One check worth doing immediately Look at what share of spend is going to your own brand name. If branded terms are taking more than about 5% of total spend, add your brand as a negative phrase in the broad, phrase and auto campaigns. Branded search converts extremely well, which makes those campaigns look excellent while they buy sales you would very likely have had anyway. See Brand Keywords: Proving Upper-Funnel Spend Works for the fuller argument about when branded spend does earn its place. And look at which match type is working. If broad is consistently outperforming phrase in your account, launch more broad. The account tells you its own preference if you ask. Then start optimising With the structure clean and the missing campaigns live, the routines have something readable to work on. See The Nine-Report PPC Routine. Related - Create a Sponsored Products Campaign — building each one, step by step. - Score Your PPC Performance — scoring the account before you rebuild it. - Campaigns: Budgets, Placement and Audience Adjustments — the screen where structure becomes visible. - Easy-Win Campaigns — the low-bid structures worth adding to a complete account. - How Much to Spend on Branded Keywords — isolating branded terms, and why it matters. - Designing the Offer — what happens before a new product needs campaigns.

The Nine-Report PPC Routine

Advertising accounts do not fail dramatically. They drift — a few non-converting targets here, a placement bid that stopped matching reality there — and six months later the account is expensive and nobody can point to the decision that made it so. The defence is a schedule. Nine reports, on three cadences, each answering one question. Run them and the drift never accumulates. Watch the walkthrough Prefer to watch? One hundred and four seconds on the nine reports and their cadences, why bids get reviewed weekly, and the half of the work that does not need you at all. The schedule | Cadence | Report | Question it answers | |---|---|---| | Weekly | Bleeders | what is taking clicks and returning nothing? | | Weekly | Placement bids | is spend sitting on the placement that performs? | | Weekly | Bid optimisation | are bids pointed at the targets that work? | | Weekly | Main keyword performance | how are we doing on the terms that define the niche? | | Bi-weekly | Search term harvesting | what is converting that we are not yet targeting? | | Bi-weekly | Bleeders 2.0 | what converts, but at a price we cannot afford? | | Monthly | Lifetime bleeders | what has never converted, across the account's whole life? | | Monthly | Placement mix | is the spend split across placements sensible? | | Monthly | Ad type mix | is the spend split across ad types sensible? | Four weekly, two fortnightly, three monthly. That is the whole system. Why the cadences differ Weekly, for anything involving bids. Two reasons, and they are both about risk. Small changes made often beat large changes made rarely — if a bid change was wrong, you find out in seven days and reverse it, rather than discovering it after a quarter. And running weekly means you can see whether last week's change helped, which is the only way judgement improves. Fortnightly, for anything that needs volume to be readable. Harvesting needs enough clicks on a search term to know it converts. Judged weekly, you would promote terms on noise. Monthly, for anything structural. Ad type and placement mix are strategy, not maintenance. They move slowly, and reacting to a week of movement would mean reacting to attribution rather than to reality. The 80/20 that makes it survivable Nine reports sounds like a full-time job. It is not, provided you spend your attention where it matters. Sort by spend, and work downward. Start with the top-spending campaigns on your best-selling product, then the next product, and so on. Give a campaign attention in proportion to what it spends. A campaign spending £200 a day deserves a careful look, possibly opening the campaign itself before deciding. A campaign spending £8 a day deserves a glance. The downside risk on the small one is trivially small, so do not spend twenty minutes on it. Doing nothing is a valid decision. If a campaign has good ACoS and a healthy trend, leaving it alone is often the best available move — every change carries risk, and a campaign performing well does not need saving. Where the time actually goes The work splits cleanly in two, and only one half needs you. Gathering. Downloading bulk reports, filtering them, pasting them into sheets, rebuilding pivots. Entirely mechanical, and it is most of the hours. Deciding. Reading the assembled numbers and choosing what to change. This is the part that needs judgement and product knowledge, and it is a small fraction of the time. The traditional answer is to hand the gathering to someone else. The better answer is for the gathering not to exist. What this looks like in TrackIQ Most of the nine are already screens, which removes the gathering half outright: Bleeders, Bleeders 2.0 and lifetime bleeders — Amazon Ads → Search Terms and Targets, sorted by spend, with the date range set to the window you want. Same screen, three date ranges. See The Bleeder Reports. Search terms, with spend and orders side by side Placement bids and placement mix — Amazon Ads → Placements, which puts performance beside the multipliers. Placements, performance against the bids that produced it Bid optimisation — Amazon Ads → Targets and Keywords, at 7 and 30 days. Harvesting — Amazon Ads → Search Terms, filtered to converting terms not yet targeted. Ad type mix — Amazon Ads → Campaigns, grouped by ad type. Main keyword performance — the keyword and rank views together. The date range holds as you move between screens, which is what makes a nine-report pass a morning rather than a project. What to automate, and what not to Some of this is genuinely mechanical and can run without you. Reasonable to automate: the bleeder rules, once you trust the thresholds. They act only on targets with no sales, so the downside is small and bounded. Keep manual: placement bid changes and bid optimisation on your largest campaigns. These need context — whether a product is launching, whether stock is short, whether a competitor has just moved — that no rule can see. See Rule-Based Optimization, and read Guardrails Before You Automate before switching anything on. Expect it to be slow at first The first pass will take much longer than you expect, and the tenth will take a fraction of the first. That is not a sign it is not working; it is what learning a routine looks like. The value is not in any single pass. It is that nothing gets six months to compound quietly. Related - Score Your PPC Performance — the same reports, used once to grade the account. - Restructure Your Campaigns Before You Optimise — the foundation these routines assume. - Ad Type and Placement Mix — the two monthly structural reports. - Setting Target TACoS per Product — the monthly target the weekly work serves. - Scaling Profit Campaigns — the offensive half, once the routine is holding. - What a Relaunch Is, and When to Run One — the second half of the system, once the routine holds.

Search Term Harvesting: Criteria and Bids

Your auto, broad and phrase campaigns are a research budget. Amazon shows your product against searches you never chose, and some of them convert better than the keywords you picked yourself. Harvesting is the routine that finds those, promotes them into targeted campaigns, and turns a lucky match into a controlled one. Watch the walkthrough Prefer to watch? Ninety-nine seconds on the four criteria, the starting-bid maths for each match type, and the negation step that gets skipped. Keyword and search term are not the same thing Worth being precise, because the whole process depends on the difference. Keyword (or target) — what you bid on. black blanket Search term — what the shopper actually typed. big black blanket for king size bed Under exact match the two are nearly the same. Under auto, broad and phrase they diverge enormously, and that gap is where the opportunity lives. Search terms: what shoppers actually typed The criteria You are looking for search terms that are not already targeted and performed well. Four thresholds define "well", and all four should be adjustable per product: | Criterion | A reasonable default | |---|---| | Minimum orders | 2 | | Maximum ACoS | 35% | | Minimum clicks | 6 | | Minimum CVR | 10% | Tune these to your product, not to the defaults. A 10% conversion rate is poor in some categories and excellent in others. If your product converts at 5% across the board, a 10% floor will return nothing and you will conclude, wrongly, that there is nothing to harvest. Minimum orders matters most. One order is luck. Two is the smallest number that is not obviously noise. Cap the list at fifteen If the pass returns more than about fifteen terms, tighten the criteria rather than promoting all of them. Two reasons. A new target needs budget and time to prove itself, and thirty new targets at once means none of them gets either. And a long list is usually a sign the thresholds are too loose, so you would be promoting mediocre terms alongside good ones and diluting the campaign's average. Fifteen good terms, promoted properly, beats forty promoted carelessly. Where to put them Into an existing campaign if one fits the theme, or into a new campaign built for them. Target each term in exact, plus one of broad or phrase. Exact gives you control over the term you now know works. The broader match keeps the research going, finding the next variation. You do not need both broad and phrase — use whichever your account has historically done better with. Isolate the big ones. If a harvested term is genuinely large — a head term for your niche — give it its own single-keyword campaign. That way its budget and bids are not competing with thirty smaller terms, and you can push it for rank deliberately. The starting bids This is the part most people guess at, and there is a straightforward rule. For exact match: take the term's current CPC and add 10%. Multiply by 1.1. You already know roughly what the click costs, because you have been paying for it. The uplift buys you position on a term you now know converts. For broad or phrase: 15% below your exact bid. Multiply the exact bid by 0.85. The broader match will pull in less qualified traffic, so it should not pay as much. For a new ASIN target, go lower: 0.8× the CPC. Product targeting behaves differently from keyword targeting, and the evidence from a search term does not transfer cleanly to an ASIN. Start conservative and raise it if it works. Then negate in the source This is the step that gets skipped, and skipping it costs money. Once a term is targeted in its own campaign, add it as a negative exact in the campaign it came from. Otherwise your auto or broad campaign continues bidding on the same term, and you are competing with yourself — two of your campaigns in the same auction, pushing your own cost per click up. Promote and negate. Always both. Cadence Fortnightly. Harvesting needs enough clicks accumulated on a term to judge it, and a weekly pass would have you promoting on noise. Monthly leaves good terms unpromoted for too long. Why this compounds Every promoted term does three things at once: it expands impression share on a search you already win, it pushes organic rank for that term because ad sales feed velocity, and it moves spend toward proven traffic and away from speculative traffic. Run consistently, the effect is cumulative. Accounts that do this for a year end up with dozens of campaigns holding fifteen to twenty proven keywords each — a network of terms that were all discovered by the algorithm and then taken under control. That network is what a mature account actually is. Common mistakes Promoting on one order. Not enough evidence. Using the default thresholds on a product whose conversion rate is nothing like the default. Promoting forty terms at once, so none of them gets the budget to prove itself. Forgetting to negate in the source campaign, and then bidding against yourself indefinitely. Starting exact bids at the CPC rather than above it, then wondering why the new target gets no impressions. Harvesting from a campaign that mixes match types, so you cannot tell which match produced the term. See Create a Sponsored Products Campaign. Related - Harvest Winning Search Terms into Exact-Match Campaigns — the same routine on the TrackIQ screen. - Search Terms and Aggregated Search Terms — reading the report this comes from. - The Nine-Report PPC Routine — where harvesting sits in the schedule. - Outlier and Trending Keywords — finding terms your own campaigns will never surface.

Placement Bid Optimisation

Most optimisation moves money between keywords. Placement optimisation moves it between positions, and that makes it the rare routine that improves efficiency and organic rank at the same time. It is also the one where a single number — Top of Search Impression Share — tells you whether a bid increase will do anything at all. Watch the walkthrough Prefer to watch? One hundred and four seconds on the four columns that decide each move, why impression share settles it, and the budget-capped distinction people get wrong. What a placement bid actually does For every Sponsored Products campaign, Amazon lets you raise your bid for a specific placement by a percentage. A 100% Top of Search multiplier on a £1.50 keyword bid means you bid £3.00 in that placement, and £1.50 everywhere else. That is a second lever on top of keyword bids, and used well it reshapes where a campaign's money goes without touching any individual keyword. Placements, with performance beside the multipliers Why it does two jobs Most optimisation chases efficiency alone. This one chases both, because for most products Top of Search converts best — higher CTR, higher CVR, better ACoS — and it is also where the sales velocity that moves organic rank happens. So shifting spend from Rest of Search and Product Pages toward Top of Search tends to lower ACoS and push rank. That combination is why this is worth doing weekly. Run it weekly, and change little Weekly, deliberately. Frequent small adjustments beat occasional large ones: the downside of any single change is small because you will revisit it in seven days, and you get to see whether last week's change helped. Work in spend order. Top-spending campaigns on your best-selling product first. The more a campaign spends, the more thought it deserves. The four things to read For each campaign, four columns decide the move. 1. ACoS trend: 30 days against 7 days Is efficiency improving or worsening in this placement? Improving, and already at a good level — you can raise the bid with reasonable confidence. Improving, on a campaign spending heavily — consider leaving it entirely. It is going the right way; let it. Worsening — reduce. One caveat that catches people out. If the 7-day ACoS is only 2–4% worse than the 30-day, treat that as flat, not as a decline. Recent days always understate because attribution has not landed yet. See Amazon PPC: Core Terms and Metrics. 2. Top of Search Impression Share This is the number that decides whether raising a bid can achieve anything. Low (roughly 5–15%) — there is a lot of Top of Search traffic you are not getting. A bid increase has room to work. High (40–60% or more) — you are already taking most of the available impressions. Raising the bid further will mostly just raise your cost per click, regardless of how well the placement performs. That second case is the useful one. If Top of Search is performing well but impression share is already 55%, the honest answer is that there is little left to buy. And it works the other way: if Top of Search is performing badly and impression share is high, cutting the multiplier is a low-risk move — from 40% to 20%, say — because you will still capture the traffic that was converting. 3. Yesterday's spend, and the day before This sizes the risk of whatever you are about to do. Low daily spend — changes will have a small absolute effect. Be decisive; the downside is trivial. High daily spend — changes matter in both directions. Make small adjustments, and open the campaign itself before deciding. 4. Budget against spend If a campaign is hitting its budget, raising bids is pointless. The budget caps the spend regardless of what you bid; all a higher bid does is spend the same money faster and at a worse cost per click. When a campaign is budget-capped, the choice is: raise the budget if ACoS justifies it, or lower bids to buy more clicks with the same money. The common shape, and the fix The pattern you will meet repeatedly: one placement performs well, the others do not, and the poor ones are taking most of the spend. Say Product Pages runs at a poor ACoS but took 27% of the campaign's spend last month, while Top of Search performs well. If the campaign is budget-capped, raising the Top of Search multiplier is enough. The campaign has a fixed pot, so pulling spend toward Top of Search necessarily starves the others. If it is not budget-capped, raising one placement does not reduce the others — it just increases total spend. Lower the bids on the underperforming placements instead. That distinction is the one most often got wrong. The advanced move: keyword and placement bids together When a campaign spends evenly across placements but only one placement works, raising a single multiplier takes a long time to shift the balance. There is a faster, riskier route. Cut the keyword bids, then raise the Top of Search multiplier to compensate. | Target's ACoS | Bid change | |---|---| | above 40–45% | reduce by 50% | | 30–40% | reduce by 25% | | below 30% | leave alone | Then raise the Top of Search multiplier by 100%. The arithmetic: lower keyword bids largely remove you from Rest of Search and Product Pages, while the doubled multiplier keeps your effective Top of Search bid near where it was. Over the following weeks, spend migrates to Top of Search. This is genuinely risky. Cuts that aggressive can stall a campaign entirely, and it takes weeks to stabilise. Once it has settled, bring it back gradually — raise the Top of Search multiplier 10–20% a week until spend is where you want it. Use it on a campaign that fits the pattern: Top of Search performing well on a low multiplier and low impression share, while the other placements consume the budget. Do not use it on your largest campaign as a first experiment. Common mistakes Raising bids on a budget-capped campaign. Nothing to gain. Raising Top of Search when impression share is already high. You are buying the same impressions at a higher price. Reading a 7-day dip as a decline. Attribution has not landed. Setting multipliers from a hunch rather than from measured placement performance. See Audit Your Placement Multipliers. Making large changes on your biggest campaign because the numbers looked clear in a spreadsheet. Related - Placements: Top of Search, Rest of Search and Product Pages — the screen, in full. - Ad Type and Placement Mix — the monthly structural view of the same question. - The Nine-Report PPC Routine — where this sits in the schedule. - Running a Launch: Stabilise, Scale, Scale Back — forcing spend to Top of Search during a launch.

Bid Optimisation: Six Rules

This is the longest routine in the system and the one worth mastering first. Every week, you look at each campaign's last 7 and 30 days and apply a small number of rules to steer it toward what you want it to do. Not automatic. The rules are mechanical, but choosing which rules to apply to which campaign is the judgement, and that is where the performance comes from. Watch the walkthrough Prefer to watch? One hundred and twenty-three seconds on the six rules with their real thresholds, how to combine them to steer a campaign, and why Sponsored Products earns more patience. The idea: navigate, do not fix You are not trying to make each campaign perfect this week. You are nudging it in a direction. ACoS too high? Apply the rules that lower bids. ACoS comfortably low? Apply the rules that raise bids, and buy more volume. Both fine? Small changes, or none at all. Doing nothing is a real option and often the right one. Every change carries risk, and a campaign that is performing does not need saving. Targets, with spend, orders and ACoS The six rules Four adjust bids, two act on the campaign. 1. Lower bids on targets with clicks and no sales Graduated by how much evidence you have: | Clicks, no orders, last 30 days | Action | |---|---| | 6–7 | reduce bid by 5% | | 8–14 | reduce bid by 10% | | more than 14 | pause the target | This is the safest rule in the set, because it only touches targets that have produced no sales at all. It can reasonably be applied to every campaign. Adjust the thresholds if your product converts below 10%. On a low-CVR product, fourteen clicks without a sale is not yet strong evidence, and you would be pausing targets that were merely unlucky. This graduated approach is the operational version of The Bleeder Reports — taper first, pause when the evidence is strong. 2. Lower bids on targets above an ACoS threshold This one acts on targets that do convert but cost too much. Set a threshold, then reduce each target's bid by the difference between its ACoS and that threshold. With a 40% threshold: | Target's ACoS | Bid reduction | |---|---| | 44% | 4% | | 49% | 9% | | 58% | 18% | The elegance is that the correction scales with the size of the problem. A target slightly over gets a nudge; a target wildly over gets a real cut. Nothing gets an aggressive change it did not earn. Set the threshold near the campaign's average ACoS, not at your break-even. If the campaign averages 39%, a 40% threshold pulls the worst targets toward the middle without disturbing the rest. 3. Raise bids on targets with almost no clicks Targets with 1 or fewer clicks in 30 days: raise the bid by a set percentage — 10% is a sensible default. These targets are not losing you money; they are invisible. Their bid is under the going rate, so they never enter the auction. Nudging them upward, week after week, gradually brings them into play. This is how you keep testing without launching anything. Every campaign accumulates targets that never got a chance, and this rule slowly gives them one. 4. Raise bids on targets with good ACoS Set a "good" ACoS ceiling, then raise the bid on every target below it by a set percentage. If good means under 24% for this campaign, every target beating 24% gets, say, +10%. More impressions on the traffic you already know works. The ceiling is per campaign. A 24% ceiling is right for a high-margin product in a soft category and completely wrong elsewhere. 5. Change the campaign budget Set a new daily budget outright. 6. Pause the campaign For campaigns that cannot be rescued by bidding. Combining rules to steer The rules only mean something in combination: | Situation | Apply | |---|---| | ACoS too high | rules 1 and 2; maybe a small rule 4 on your best targets | | Want more volume, ACoS is healthy | rules 3 and 4 | | Budget-capped, good ACoS | rule 5 — raise the budget | | Budget-capped, bad ACoS | rules 1 and 2 — lower bids, do not raise budget | | Performing well, no concerns | nothing | The budget-capped pair is worth internalising. A campaign hitting its budget every day is telling you something, but what to do depends entirely on ACoS. Good ACoS means you are leaving sales on the table. Bad ACoS means you are buying too expensively, and more budget just buys more of the problem. Two principles that override the rules Sponsored Products deserves more patience SP campaigns build organic rank in a way SB, SBV and SD do not. Their value is partly in a place the ACoS column cannot see. So allow SP a higher ACoS than you would allow the others, and give it more second chances. Judged purely on last-click efficiency, you will cut the campaigns doing your long-term work. Under target ACoS, decide deliberately When a campaign is beating its target, you have two honest options: leave it alone, or spend more and accept a worse ACoS. If ACoS sits at 23% or below over both 7 and 30 days, raise bids. Even if it drifts to 26–29%, that is still strong, and you will have bought real extra volume. If ACoS is already around 29%, be careful. Raising bids may take it to 35%. That might be a trade you want — but make it a decision rather than a side effect. Either way, move in small weekly increments. Focus increases on targets with good ACoS and on targets with almost no clicks, which is where the upside is cheapest. How to work through it Sort by 30-day spend, grouped by portfolio. Best-selling product first. Give each campaign attention proportional to its spend. Big spenders may warrant opening the campaign itself. Small spenders get a glance and a decision. Then stop. The temptation is to keep tuning. The routine works because it runs every week, not because any one pass is exhaustive. Related - Placement Bid Optimisation — the same weekly slot, one level up. - The Nine-Report PPC Routine — where this sits in the schedule. - Keywords and Match Types — the screen these decisions are made on. - Scaling Profit Campaigns — steering campaigns once they are scaling.

Setting Target TACoS per Product

Most sellers set advertising budgets by feel, then discover at the end of the quarter what it cost them. Target TACoS replaces that with a number you choose deliberately, per product, and manage toward. The relationship that makes it work is unusually direct: one percentage point of TACoS is one percentage point of margin. Reduce TACoS from 21% to 18% and you have added three points of profit margin. Nothing else in advertising is that clean. Watch the walkthrough Prefer to watch? One hundred and eleven seconds on why the target belongs to the product rather than the account, what each lifecycle stage can afford, and how fast you can safely move it. Set it per parent ASIN, not per account An account-level TACoS target is an average of products at completely different stages, and averages hide exactly what you need to see. A mature product subsidising a launch looks identical to two mediocre products. Set a target for each parent ASIN. That is the level at which the decision is real, because it is the level at which the product's stage, margin and competition are consistent. The stage decides the number A product's right TACoS changes as it matures, and expecting a launch to hit a mature product's number is the most common way people strangle a new listing. | Stage | Target TACoS | Why | |---|---|---| | Launch | 25–50% | buying the velocity that creates rank and reviews | | Growth | 15–25% | maintaining momentum while efficiency improves | | Mature | 5–15% | organic sales carry the product; ads supplement | The trajectory is the point. A product should be moving down this table over its life. If a two-year-old product is still at 30% TACoS, either it never achieved organic rank or something is wrong with the listing — and no bid change fixes either. Working out your number Start from gross margin before advertising. If a product clears 35% before ad spend, a 10% TACoS target leaves roughly 25%. Then decide the profit margin you actually want, which again depends on stage: Mature products: 17–23%. This is what a healthy, established product should return. Growth phase: accept less, deliberately, to keep pushing. New products: break even, or take a small planned loss. You are buying rank and reviews, and that is what it costs. Competitive niches need higher TACoS, permanently. That is a fact about the category rather than a failure of management, and setting an unrealistically low target in a hard category just means you stop advertising and lose position. Adjusting it: move slowly You now have last month's actual TACoS and profit margin, and a target. The gap tells you which way to move. If margin is below what you want Lower TACoS gradually, a couple of points a month. Last month: 21% TACoS, 14% margin. Set this month's target at 19–20%. Next month, 17–18%. Within two months margin is at 17%+ without anything breaking. Do not drop from 20% to 15% in one month. Cutting spend that hard reduces sales velocity, which reduces organic rank, which reduces organic sales — so total revenue falls and TACoS may not improve at all. That is the flywheel running backwards, and it is much harder to restart than it was to stall. If margin is above what you want You are under-investing, and that is also a mistake — just a quieter one. Last month: 27% margin at 7% TACoS. There is room to buy growth. Set 12–15% and put the difference into campaigns that convert. A high margin with very low TACoS usually means a product is coasting on organic position that a competitor will eventually take. Then manage to it A target is only useful if something changes when you miss it. Review monthly. TACoS is a slow metric — weekly readings are mostly noise and attribution. When a product is above target, the weekly routines are the instrument: bleeders, bid reductions on high-ACoS targets, placement rebalancing. See Bid Optimisation: Six Rules. When a product is below target, raise bids on targets that already convert well and on targets with almost no clicks — the cheapest available growth. Advertising performance against total sales Where the number comes from TACoS needs total sales, not just ad sales, which is why it cannot be read from the advertising console alone. Historically this meant exporting ad spend from one place, total revenue from another, and reconciling them per ASIN in a spreadsheet every month — which is why most sellers track TACoS at account level, if at all. In TrackIQ it is already joined. Ad spend and total sales sit against the same products over the same date range, so TACoS per parent ASIN is a column rather than a monthly reconciliation exercise. See Understanding Business Reports for the total-sales side, and Understanding ROAS and Key Amazon Advertising Metrics for how the metrics relate. Common mistakes One target for the whole account. It hides every problem worth finding. Cutting TACoS too fast, and stalling the flywheel. Treating a competitive category's high TACoS as failure. Compare against the category, not against an ideal. Never raising it. A product with high margin and very low TACoS is usually under-invested. Reading TACoS weekly and reacting to attribution noise. Related - Amazon PPC: Core Terms and Metrics — ACoS, TACoS and how they differ. - Profit Campaigns and Ranking Campaigns — ranking campaigns are meant to exceed target TACoS. - What Amazon PPC Is Actually For — why profit rather than revenue is the measure. - Designing the Offer — the margin the target assumes. - What a Relaunch Is, and When to Run One — why a launch is allowed to exceed its target. - When You Are Going Out of Stock — cutting spend deliberately when supply is short.

Easy-Win Campaigns

Most advertising effort goes into the campaigns that spend the most. Reasonable — but it leaves a category of campaign almost entirely unexploited: cheap, low-bid campaigns that individually generate modest sales at very low ACoS, and collectively pull your account average down. They are not exciting. They are close to free money, and they need almost no maintenance once they are running. Watch the walkthrough Prefer to watch? One hundred and twenty-one seconds on the six structures and their exact bids, why the low bid is the whole strategy, and how to ramp them a cent at a time. The principle Every one of these bids far below your normal rate. The point is not volume — it is cost. You are picking up clicks nobody else wants at the bottom of the auction, and at those prices even a mediocre conversion rate produces a good ACoS. Which means the usual instinct is wrong here. When one of these performs well, resist raising bids much. The low bid is the strategy. Push it up and you have converted a cheap campaign into an ordinary one. The campaigns Auto Catch-All One Sponsored Products campaign containing all your ASINs. Bids: $0.40 across all four targeting groups — close match, loose match, substitutes, complements. Down-only bidding, no placement multipliers. Budget: $25–50/day. Expect ACoS around 20–25%. Adjust bids per targeting group for the first few weeks to land in that range, then largely leave it. This is the highest-priority one to launch. Auto Low Bid One campaign per parent ASIN, with all child ASINs included. | Targeting group | Bid | |---|---| | Close match, substitutes | $0.10 | | Loose match, complements | $0.05 | Down-only. Budget $25–50/day. Genuinely set-and-forget. Each campaign typically produces five to twenty sales a month at under 15% ACoS. Small numbers that add up across a catalogue. Auto Top of Search The same idea aimed specifically at cheap Top of Search clicks. | Targeting group | Bid | |---|---| | Close match, substitutes | $0.03 | | Loose match, complements | $0.02 | Then set the Top of Search placement multiplier to 900%. The arithmetic: a 3-cent bid multiplied by ten is 30 cents at Top of Search, and near-invisible everywhere else. You are buying Top of Search position at a fraction of the usual rate, accepting that you will only win the cheap end of it. Get these bids exactly right. At 30 cents instead of 3, the 900% multiplier means $3.00 a click. This is the campaign most capable of costing you real money through a typo. Low-Bid Display Product Targeting A Sponsored Display campaign using contextual targeting, optimised for conversions. Target your sub-category rather than the broad one — not "Beauty & Personal Care" but the specific shelf your product sits on — plus closely related categories. Bids: $0.10. The same warning applies: at $1.00 this becomes an expensive campaign very quickly. Low-Bid Manual Research Three ad groups — exact, broad and phrase — targeting every term that produced at least one sale in the last 60 days. For a new product, use your keyword research instead. Bid at roughly 15% of your portfolio's average CPC. If your average CPC is $1.14, bid $0.17. Set Top of Search to 200%. Down-only. Low Bid, High-Intent Audience An auto campaign at $0.20, down-only, all four targeting groups — with an audience bid adjustment applied to the high interest based on shopping history audience. Cheap clicks, weighted toward shoppers Amazon already believes are likely to buy. Bid adjustments, by placement and by audience Getting them to spend, without overshooting Low-bid campaigns very often produce nothing at first, because the bid sits below the auction floor. The fix is patience and small steps. Every other day for the first three weeks: look at targets with fewer than 2 clicks the previous day and raise those bids by $0.01. Leave anything with 2 or more clicks alone. Raise by one cent. Not five, not ten. These campaigns work on thin margins between bid and value, and a large jump turns a cheap campaign into an ordinary one in a single step. Stop after about five increases. If a campaign will not spend after five nudges, pause it and try again another time. Name it clearly so you can find it later. And stop raising once clicks are steady, even if ACoS looks excellent. A 12% ACoS is the campaign working exactly as intended. It is not a signal to scale. When to expect more from them Q4 and Prime Day. Conversion rates rise across the board, and these campaigns benefit disproportionately because their economics are already favourable. Raise bids modestly going in, and put them back afterwards. See Prepare for Prime Day. What to expect overall Individually unimpressive: a few sales a month each. Across a catalogue with a campaign per parent ASIN, the aggregate is a meaningful block of low-ACoS revenue that pulls the account average down and needs almost no attention. They also work unevenly. A structure that performs well on one ASIN may do nothing on another. Launch across the catalogue, keep what works, pause what does not, and do not try to diagnose the failures — the whole point is that they are cheap enough not to warrant the analysis. Common mistakes Raising bids because ACoS looks good. The low bid is the entire strategy. Getting the Auto TOS bids wrong, where a 900% multiplier turns a typo into real money. Raising by more than a cent during the ramp. Judging them on volume. They are not volume campaigns. Leaving them out of your negatives. They are auto campaigns, so they still drift — see The Bleeder Reports. Related - Restructure Your Campaigns Before You Optimise — where these sit in a complete structure. - Create a Sponsored Products Campaign — building each one. - Cut Wasted Ad Spend in an Afternoon — keeping auto campaigns honest. - Winning the Frequently Bought Together Placement — turning paid spend into an organic placement.

How Much to Spend on Branded Keywords

Branded campaigns always look like your best campaigns. Someone searches your brand name, sees your ad, buys. ROAS is enormous, ACoS is tiny, and the report says you are doing something clever. The uncomfortable question is how many of those people would have bought anyway. Watch the walkthrough Prefer to watch? One hundred and twelve seconds on why ACoS cannot judge branded spend, how to measure branded market share, and the test that tells you to stop spending. Why ACoS cannot answer this Branded search is the warmest traffic on Amazon. The shopper has already decided which brand they want — they are typing your name. Of course they convert. So a 6% ACoS on branded terms tells you nothing about whether the spend was incremental. Some of those sales are genuinely won by the ad. Some would have arrived through your organic listing seconds later, for free. Both mistakes are common. Some brands pour budget into terms that would have converted organically. Others under-spend and let competitors intercept shoppers who were explicitly looking for them. The metric that separates the two is not ACoS. The metric that works: branded market share Instead of asking "did the ad convert", ask "of everyone searching for my brand, how many bought from me?" Branded market share = your purchases ÷ total market purchases, across branded search terms. This is answerable because Amazon publishes it. Search Query Performance reports, per search term, how many purchases happened in total and how many were yours. How to build it: 1. Get the Brand View Search Query Performance report. 2. Classify the branded terms. Your brand name, common misspellings, abbreviations, and product names unique to you. Misspellings matter — they are often a meaningful share of branded volume. 3. Divide your purchases by total purchases across those terms. Track it weekly. Search Query Performance, with your share against the market One caution about what to watch. The absolute number of branded purchases is not the KPI. As a brand grows, branded searches rise on their own, so the raw number climbs whatever you do with advertising. The share is the signal. Isolate branded traffic first This step is not optional, and skipping it makes everything after it unreliable. Branded terms need to sit in their own exact or phrase campaigns, and be negated from your auto and non-branded campaigns. Without that separation, "branded spend" is a number you cannot control or even measure — it drifts as your auto campaigns wander onto brand terms, and any test you run reflects changes in campaign mix rather than the change you intended. With it, branded spend is a dial you can turn. The test Once branded traffic is isolated and you have a few weeks of market share history, you can calibrate by experiment rather than argument. If branded market share is high You are capturing most of your own demand. Test spending less. Reduce branded allocation — say from 20% of budget to 10% — and watch market share for a few weeks. If share holds, the spend was not buying those sales. You have found free margin, and the freed budget can go to acquiring customers who do not yet know you. If share drops, put it back. You have learned where the floor is, which is exactly what the test is for. If branded market share is low Check for real problems first, because advertising cannot fix any of these: Buy Box suppression — you cannot win a sale you cannot be added to a basket for. Pricing — if a reseller undercuts you on your own brand term, that is the issue. Organic visibility on your own brand — if you are not ranking for your own name, something is wrong with indexing or the listing. Conversion problems on the listing itself. If none of those apply, test spending more. Raise allocation from 10% to 15–20% and watch. If share rises meaningfully, the extra spend is genuinely capturing demand that was going elsewhere. If share barely moves, the constraint is not advertising. More budget will not help, and you should stop and fix the real cause. That second outcome is the valuable one, because it stops you spending against a problem that advertising was never going to solve. The principle The goal is not to maximise branded spend, and not to eliminate it. It is to spend the minimum required to capture the demand your brand has already created, and put everything else toward creating new demand. Branded advertising defends. It does not grow. The growth comes from the non-branded budget you free up by not over-defending. For the upper-funnel side of that argument — whether your other advertising is creating branded demand in the first place — see Brand Keywords: Proving Upper-Funnel Spend Works. Where the data lives The manual version means downloading Search Query Performance weekly, classifying branded terms by hand in a spreadsheet, and maintaining the history yourself — which is why most brands do this once, if ever, rather than weekly. In TrackIQ, Search Query Performance is a screen with history already behind it. Your purchases and market purchases sit side by side per term, over a date range that holds, so branded market share is a filter and a calculation rather than a weekly export ritual. See Understanding Search Query Performance and Brand Keywords. A useful rule of thumb If branded terms are taking more than about 5% of total ad spend, look closely. That is not automatically wrong — a brand under active attack from competitors may justify more — but it is the level at which the question deserves an answer rather than an assumption. Related - Understanding Search Query Performance — the report this method runs on. - Restructure Your Campaigns Before You Optimise — isolating branded terms into their own campaigns. - Setting Target TACoS per Product — where freed branded budget should go. - Tracking Branded Market Share — measuring whether that spend is incremental.

Scaling Profit Campaigns

Most advertising work is defensive: find the waste, cut it, keep the account tidy. Necessary, but it has a ceiling — you can only save what you were already spending. Scaling is the offensive half. If you have campaigns running comfortably below break-even ACoS, every additional pound you can put through them at that efficiency is direct profit. The job becomes finding more places to put it. Watch the walkthrough Prefer to watch? One hundred and twenty-two seconds on the arithmetic that justifies scaling, where campaign ideas come from, and the doubling-down step people skip. The arithmetic that justifies it Two numbers, and you need both before starting. Break-even ACoS — the point where ad spend exactly equals the profit on the sale. Set by your margin. Target ACoS — the efficiency you actually want. If break-even is 40%, then a campaign running at 30% is returning roughly 10% profit on that spend, and one at 25% is returning about 15%. Which leads to the conclusion that drives the whole exercise: if you could spend £100,000 a day at 25% ACoS, you would want to. Every pound is profitable. The only real limit is how much traffic you can find at that efficiency. So scaling is not "spend more". It is find more places that hold the efficiency, then spend more there. When this is the right move Premium products priced above the competition, where efficiency matters more than volume. Products with weak conversion rates that are poor relaunch candidates — if you cannot win rank, win profit instead. Q4, when conversion rates and demand rise and campaigns that were marginal become profitable. Brand new products that need sales velocity to accumulate reviews. After a successful relaunch, once organic position is established and the ranking work is done. The method: launch, measure, double down This is a volume game played deliberately, not a single clever campaign. 1. Confirm your break-even and target ACoS. Everything downstream is judged against them. 2. Read your own data first. Which ad types, match types and placements already perform for these products? You are looking to build more of what already works, not to guess. 3. Shortlist the next 10–12 campaigns you would launch for your main ASINs. 4. Launch three a week, on the same day every week. The consistency matters more than the number. Three a week is 150 a year, which is enough to find the structures that work for your catalogue. 5. Start with medium bids and check daily. Early on you are hunting cheap clicks, so let the campaign find its level before pushing. If it is not spending, raise gradually. 6. When one works, build five to ten more like it. This is the step people skip. A campaign structure that performs at 25% ACoS on one ASIN will often do the same on others. Having found a winner, replicate the shape rather than moving on to a new idea. Campaigns, ranked by spend and efficiency Where the campaign ideas come from You do not need a new concept each week. The variations that produce a large campaign list are combinations of things you already have: Match type — the same keyword set in exact, phrase and broad as separate campaigns. Placement focus — a Top of Search-only variant and a Product Pages-only variant behave like different campaigns. Ad type — the same targets as Sponsored Products, Sponsored Brands and Sponsored Display. Targeting basis — keyword campaigns beside ASIN campaigns aimed at competitors where you compare well. Bid level — a low-bid version of a campaign is a genuinely different campaign, as Easy-Win Campaigns covers. Audience — the same campaign with an audience bid adjustment applied. Combine those and a dozen structures becomes a queue that lasts months. Scale hard, but keep the routine A campaign that produces profitable sales should be scaled aggressively — raise budgets, raise bids, and keep going while efficiency holds. But keep running the weekly routines on them. Scaling and optimising are not alternatives. A scaled campaign accumulates bleeders like any other, and the faster it spends the faster it accumulates them. See The Nine-Report PPC Routine. Watch TACoS, not just ACoS. Scaling profitable campaigns raises total ad spend, which raises TACoS even while every individual campaign stays efficient. That may be exactly what you want during a growth phase — but it should be a decision. See Setting Target TACoS per Product. What stops it working Not knowing break-even ACoS. Without it, "profitable" is a guess and you may be scaling a loss. Launching one campaign and waiting. The method depends on volume and iteration. Abandoning a winner to try something new, instead of replicating it. Scaling into a stock problem. Driving velocity into a product that runs out is worse than not scaling at all — you lose the rank you bought. See Catch a Stockout Early. Raising bids too fast at launch, so the campaign never gets the cheap clicks that made it viable. Related - Easy-Win Campaigns — the low-bid structures worth replicating first. - Bid Optimisation: Six Rules — steering campaigns once they are scaling. - Shift Budget from Harvesting to Acquisition — when efficiency has crowded out growth. - Sponsored Brands Video at Scale — testing creative, then scaling only the winner.

Outlier and Trending Keywords

Two keyword tactics that most accounts never run, both aimed at the same thing: finding profitable traffic your competitors are not bidding on. One looks sideways, at searches adjacent to your product rather than describing it. The other looks forward, at searches that are growing quickly and are not yet crowded. Watch the walkthrough Prefer to watch? One hundred and twenty seconds on what an outlier keyword is and why it converts, the filter that finds them, and why trending terms need the opposite bid approach. Outlier keywords What they are Outlier keywords are broader terms related to your product without describing it. Sell a car wash mitt, and your obvious keywords are car wash mitt, car wash glove. Your outlier keywords are car cleaning accessories, gifts for car lovers, car stuff. Nobody typing "gifts for car lovers" is looking for a wash mitt specifically. But some of them will buy one. Why they work High search volume. Broad terms are searched far more than specific ones. Low cost per click. Few sellers bid on terms that do not describe their product, so the auction is soft. Effectively total share. If you are the only car wash mitt on the results page for gifts for car lovers, you capture essentially all of the wash-mitt demand there. It is like being the only one of your product in a shop where nobody came looking for it — the few who want one have no alternative. The sales are incremental. These are not shoppers you would have reached otherwise. Five extra sales a day is about 1,800 a year that would not have happened. Finding them Work from your competitors rather than your imagination. 1. Identify your top 3–5 competitors by best-seller rank within your niche. 2. Pull the keywords those ASINs rank for, then filter to find the adjacent ones: | Filter | Setting | Why | |---|---|---| | Minimum search volume | 1,000 (or 500 for a wider net) | worth the effort | | Number of competitors ranking | between 1 and 2 | few rivals means a soft auction | | Competitor rank | between 1 and 20 | someone is ranking, so the term converts for something | That "1 to 2 competitors" filter is the one doing the work. A term ten competitors rank for is a main keyword, already contested. A term one competitor ranks for is an opening. 3. Add them to your back-end search terms, then build the campaign. The campaign Sponsored Products, manual, exact match. Down-only. No placement multipliers. Budget $25–50. All bids at $0.50. You are casting a wide net cheaply to see which terms have anything in them. Optimising it weekly Set the target ACoS at your gross margin minus 5 points. On a 35% margin, the ceiling is 30%. For targets above the ceiling, cut the bid by the difference: | Target's ACoS | Bid change | |---|---| | 45% | reduce 15% ($1.00 → $0.85) | | 70% | reduce 40% ($1.00 → $0.60) | Kill non-converters quickly: more than 15 clicks and no sales, pause it. This is a discovery campaign, and most terms will fail. Cut them faster than you would in a core campaign. When it works, double the budget. If ACoS is inside the ceiling and the campaign is spending its budget, double it, and keep doubling while efficiency holds. These sales are entirely incremental, so there is no cannibalisation to worry about. When an outlier becomes a main keyword Occasionally one converts unusually well. When that happens, treat it like a keyword you would want to rank for organically: Break it into its own single-keyword campaign. Push Top of Search to buy the velocity that moves rank. Consider adding it to the title. At that point you are running a ranking campaign rather than a profit campaign, and the rules change — see Profit Campaigns and Ranking Campaigns. Trending keywords The opportunity Trending keywords are terms growing more than 50% month over month. Get onto one early and you buy top placements cheaply, before competitors notice, and convert that into organic rank. Once you hold the organic position, dislodging you is expensive. A common cause is Amazon's own autocomplete. When Amazon changes what it suggests in the search bar, shopper behaviour shifts quickly — and there is a window before the rest of the category reacts. Finding them Same starting point, different filter: minimum search volume 1,000, and search-volume trend of at least +50%. That list is your trending set. The campaign Very close to the outlier campaign, with two deliberate differences: Sponsored Products, manual, exact, down-only. Budget $25–50. Bid at Amazon's suggested bid, not $0.50. These terms are growing, so you need to actually compete for them. Set Top of Search to 25%. You are trying to establish rank, which means winning the placement that drives velocity. Target ACoS: your gross margin — a little more generous than the outlier ceiling, because you are buying a position rather than just profitable clicks. If you normally run above your margin, adjust accordingly. Same delta-based bid cuts. Same 15-clicks-no-sales rule. Same doubling when it works — and here the urgency is real, because you are racing competitors to a term that is still cheap. Cadence Check for new trending keywords every one to two months. More often in high-volume niches, where terms emerge and get crowded faster. Run consistently, this is one of the more reliable ways to grow market share: you are repeatedly first onto terms that later become contested. Keywords, with rank and performance together Doing this without a separate research tool Both tactics need the same underlying data: what your competitors rank for, at what volume, and which way that volume is moving. Historically that meant a third-party keyword tool and a manual export each round. TrackIQ carries the same signals. Search Query Performance gives search volume and your share of it per term; organic rank tracking shows what you and competitors hold; and Search Terms shows what is already converting for you. See Understanding Search Query Performance and SQP and Keywords: Finding Unbid Queries, which is the same "we are not bidding on this yet" question applied to your own data. Common mistakes Bidding normally on outliers. The low bid is what makes them profitable. Being slow to cut. Most outlier terms fail; that is expected. Cut at 15 clicks. Not doubling down when one works. Incremental profitable sales should be scaled hard. Using the outlier bid approach on trending terms. Trending terms are contested and need a real bid plus Top of Search. Checking for trends once and never again. The whole advantage is being early, repeatedly. Related - Search Term Harvesting: Criteria and Bids — promoting terms your own campaigns already found. - Scaling Profit Campaigns — what to do once one of these works. - Keywords and Match Types — reading them once live.

Reviving a Sleeping Product

The relaunch framework depends on data you already have — your conversion rate against the keyword average, your click-through rate, your current rank. A product selling 200 units a day has all of it. A product selling 10 units a day has almost none. Its Search Query Performance is thin and its advertising history is negligible, so the comparison that decides which keywords to push simply cannot be made. So you buy the data first, deliberately and cheaply, and only then run the relaunch. Watch the walkthrough Prefer to watch? One hundred and twenty seconds on why a low-volume product cannot be relaunched directly, how to choose which one to revive, and why a negative result is worth paying for. Do your successful products first This matters more than the process that follows. A product outselling your sleeping ones has demonstrated product-market fit. Maximising it is higher-probability, faster and more profitable than reviving something that has never worked. Only when your successful products are genuinely optimised does reviving a sleeping one become the best available use of your money. That is the same precondition as Before You Add a Product, and it is skipped for the same reason: reviving something feels more interesting than improving something. Choose which sleeping product to work on If you have several dormant products, do not pick by affection. Build a comparison table across your candidates, judging each against the leading offers in its market. You are answering two questions: Which has the most competitive offer relative to what already wins in that niche? Which has the highest upside if it did work? Then take the top one or two. Not five. The whole point of this process is that it costs money to run, and spreading it produces nothing anywhere — the same arithmetic as Grouping and Prioritising Keywords. Which products carry, and which do not The three added steps The sleeping process is the relaunch process with three steps bolted on the front. 1. Manual keyword research, because your own data cannot supply the keywords. 2. Group the keywords, exactly as in a normal relaunch. 3. Buy the missing data with a deliberately targeted test campaign. Then you rejoin the standard relaunch at phase one, with real numbers instead of guesses. See Keyword Research from Search Query Performance. Why not skip straight to relaunching? Because a relaunch is expensive, and without data you would be choosing keywords by instinct. You would be spending heavily to rank for terms you have no evidence you convert on — and the single most useful thing the relaunch framework does is refuse to push where the data says you do not deserve to rank. The test campaign costs a fraction of a relaunch. Buying 50–100 clicks per root keyword is cheap relative to driving enough velocity to move organic position, and it tells you which keywords are worth that spend. Flying blind here is the expensive option, not the fast one. What you will have at the end A ranked keyword list built from competitor analysis rather than your own thin history. Keyword groups, each with a root and its long tails. Real conversion data on each root keyword — where you convert best, and where you beat the category average. Which is exactly the input the relaunch framework needs. From there the process is identical to a successful product's, and everything in What a Relaunch Is applies. When to stop instead The test campaign can tell you the answer is no, and that is a useful outcome rather than a failure. If you buy 100 clicks on every root keyword and convert below the category average on all of them, the product is not currently worthy of the positions you were going to buy. Relaunching would spend a great deal to arrive at the same conclusion more slowly. At that point the work is the offer, the listing or the product — see Answer the Complaints in Your Content and Fix the Product by Studying Competitors. Learning that for the price of a test campaign is the process working. Related - Keyword Research for a Product With No Data — building the list from competitors. - Buying the Data You Are Missing — the test campaign that decides it. - What a Relaunch Is — where this process rejoins.

Keyword Research for a Product With No Data

When your own product has no meaningful search history, you borrow someone else's. Your competitors rank for the keywords in your niche, and Amazon will tell you which ones. Done properly, this recovers 90%+ of the relevant keywords for your product in an afternoon. Watch the walkthrough Prefer to watch? One hundred and twelve seconds on choosing competitors whose keywords are genuinely yours, the check that validates the list, and what the list still cannot tell you. Step 1: pick the right competitors List 5–10 direct competitors — products genuinely like yours, not merely in the same category. Choose the biggest sellers, identified by Best Seller Rank. They rank for the widest set of terms, which is what makes their keyword footprint useful. Match on the things that determine shared search terms: features, quality level, price band, and customer type. A premium version of your product attracts different searches from a budget one, and mixing them dilutes the list. If a competitor has variations, take the child ASIN that sells most. Parent-level selection blurs the keyword picture across variants that may target different searches. Getting this step wrong invalidates everything downstream. The whole method assumes their keywords are your keywords. Step 2: pull their search terms Brand Analytics → Top Search Terms, with the date range set to the last month. Enter your competitor ASINs into the ASIN field, space-separated. Amazon returns the search terms where those products appear, along with which products receive the most clicks on each. Step 3: sanity-check the output For each keyword returned, at least one of the top three clicked products should be one of your chosen competitors. That check is how you know your competitor selection was right. If your competitors are not among the top-clicked products for most of the terms, you picked products that do not actually compete for the same shoppers — go back to step 1 and reselect. When it passes, the list should cover 90%+ of the relevant keywords for your product. Step 4: capture search frequency rank alongside each keyword Download the report and keep two columns: the keyword, and its Search Frequency Rank. SFR is a rank, so lower means more searched. It is how you will size the keywords when you group them — a term at SFR 8,000 is a completely different proposition from one at SFR 400,000. Search terms and their volume Step 5: fill in whatever data you do have You have thin data, not zero data. Add whatever exists for these keywords from your own Search Query Performance and advertising reports. On a sleeping product most rows will be empty, but the ones that are not are the most valuable rows in the sheet — those are keywords where you have already converted at least once. Any keyword where you have real conversion data goes to the top of your list, because it is the closest thing to evidence you have. See Understanding Search Query Performance and Search Terms and Aggregated Search Terms. What you have now — and what you do not You have: the keyword universe for your niche, sized by search frequency. You do not have: any idea which of them you convert on. That distinction is the whole reason for the next step. This list tells you what people search. It does not tell you whether shoppers respond to your product when they search it, and that is the question a relaunch depends on. So group these keywords, then buy the data — see Grouping and Prioritising Keywords and Buying the Data You Are Missing. A use beyond sleeping products This method is worth running even on healthy products, because your own data only shows keywords you already appear for. Competitor keyword footprints reveal terms you have never ranked for and never bid on — which is the same gap Outlier and Trending Keywords attacks from a different angle. Common mistakes Choosing competitors by category rather than by genuine similarity. Using parent ASINs where the variants target different searches. Skipping the top-three check, and building a plan on the wrong keyword set. Discarding SFR, then having no way to size the groups. Treating the list as a plan. It is an input; the test campaign is what turns it into one. Related - Reviving a Sleeping Product — why this step exists. - Buying the Data You Are Missing — turning the list into evidence. - Grouping and Prioritising Keywords — the grouping step, unchanged.

Buying the Data You Are Missing

You have a keyword list and no evidence about any of it. This campaign exists to buy that evidence — cheaply, and in a form the relaunch framework can use. The goal is 50–100 clicks per root keyword. Not sales, not rank. Clicks, so that conversion rate becomes measurable. Watch the walkthrough Prefer to watch? One hundred and twenty-seven seconds on the campaign settings and why Top of Search matters, the three daily checks, and the benchmark that only exists once you buy it. First, pick the keywords to test From your grouped list, take the top 5–10 root keywords. Root keywords are the low-SFR terms — generally under 150,000 — that differ from each other by at least one word. Each anchors a group; the long tails hang off it. Then filter that shortlist: Highly relevant only. If a term is not clearly about your product, testing it wastes clicks. Highest search volume first, but exclude anything above SFR 250,000 — too thin to produce a readable result in reasonable time. Drop anything you already rank in the top 10 for organically. You have data there, and paying for clicks on a term you already win teaches you nothing. The campaign | Setting | Value | |---|---| | Ad type | Sponsored Products | | Match type | Exact | | Bidding | Dynamic bids, down only | | Keywords | your top 5–10 roots | | Starting bid | 75% of suggested (suggested × 0.75) | | Bid cap | $1.50 initially, lower in expensive niches | | Top of Search placement | 200% | | Daily budget | $50–100 | | Name | Product — SP — Exact — TOS — Top KWs | Why Top of Search at 200%: you need clicks concentrated in one placement so the conversion rate you measure means something. Clicks scattered across product pages and rest-of-search produce a blended number you cannot compare against a keyword average. Why a bid cap: this is a data-buying exercise, not a ranking push. The cap stops a competitive term consuming the whole budget. Why the higher budget is better: spending more finishes the test sooner. You are buying a fixed number of clicks; the only variable is how long it takes. Daily optimisation while it runs Three checks, the same shape as a real launch — see Running a Launch. 1. Is it spending? If under 75% of budget, raise the Top of Search bid and the keyword bids by 10%. Repeat daily until it clears 75%. 2. Is the spend at Top of Search? If more than 20% is landing on rest-of-search or product pages, lower keyword bids 10% and raise the Top of Search bid 25%, daily, until 80%+ sits at Top of Search. Unless the other placements are converting well — in which case the data is still useful and you can leave it. 3. Is it exhausting the budget early? Check by early afternoon. If it is already spent, lower the Top of Search bid or raise the budget, so you are not dark for the rest of the day. Placements, where the clicks are landing What the campaign is actually producing Three things, and only the first is obvious. 1. Conversion rate per keyword. Which roots convert well and which do not — so you know which group to launch first, and which to leave alone. 2. Search Query Performance data. Driving sponsored sales generates organic impressions, which starts populating SQP for these terms. You are seeding the report that was empty. 3. A conversion benchmark. SQP gives you the keyword's average conversion rate — which is the number your product has to beat. Without it you have your own conversion rate and nothing to compare it against. That third output is the one that matters most, and it is why this cannot be skipped. The relaunch framework's central gate is your CVR against the keyword average, and on a sleeping product neither figure exists until you buy them. Reading the result Roots where you beat the keyword average — these are your launch candidates. Start with the strongest and run the standard relaunch. See Keyword Research from Search Query Performance. Roots where you trail badly — do not launch on them. No amount of velocity fixes a conversion problem. Roots with too few clicks to judge — either keep the campaign running on those specifically, or accept they are too thin to be worth a launch. All roots below average — the honest answer is that the product is not ready. That is a listing or offer problem, not a ranking one, and finding it out here costs a fraction of finding it out mid-relaunch. Then rejoin the main process With conversion data and a benchmark, the sleeping product is now indistinguishable from a successful one as far as the framework is concerned. Go to phase one and run it normally — see What a Relaunch Is. Common mistakes Testing too many keywords, so none accumulates enough clicks. Skipping the Top of Search concentration, producing a blended conversion rate. Judging on 15 clicks. The 50–100 range is what makes the number readable. Bidding at full suggested, spending more than the exercise requires. Treating it as a launch and pushing for rank before the data says you should. Ignoring a bad result and relaunching anyway. Related - Reviving a Sleeping Product — where this sits. - Keyword Research for a Product With No Data — producing the list to test. - Running a Launch — the same daily checks, at full scale.

Optimisation as a Calculated Bet

Every bid change you make is a bet. You might be right and improve the campaign, or wrong and make it worse — and no amount of care removes that. So the goal is not to be certain. It is to place as many high-probability bets as possible, and to size each one against what it would cost to be wrong. Read this before your first few optimisation rounds. It is the mindset the bid, placement, bleeder and harvesting rules all sit on top of. Watch the walkthrough Prefer to watch? One hundred and fifty-two seconds on sizing a change to the cost of being wrong, steering campaigns toward target ACOS, and leaving the sweet-spot campaigns alone. The core rule If a change has a high chance of helping, make it — even if it is small. A 3% keyword bid rise in the right place compounds across months. If it has a low chance of helping, either skip it or shrink it. Not 7% on the keyword — 3%. Not 25% on the placement — 10%. And if it has a serious chance of backfiring, do not make it at all. That is the whole discipline. Small positive changes stack up. One large negative change on a major campaign undoes a quarter of them. Size the change to the size of the campaign Large, established campaigns are a car at 100 mph. A sharp turn crashes it. Keep placement changes to 5–10% and keyword bids to 3–5%. Small campaigns are a car at 20 mph. You can make sharp turns and even U-turns safely, because there is very little at stake. | Campaign | Placement bid change | Keyword bid change | |---|---|---| | Large / 80-20 campaigns | 5–10% | 3–5% | | Mid-size | 10–15% | 5–7% | | Small / new | larger changes are fine | larger changes are fine | The asymmetry is the point. On your biggest campaign, being wrong is expensive and being right is only slightly better than doing nothing. On a small one, being wrong costs almost nothing. Steering, not solving Optimisation is not a maths problem with a correct answer. Treating it as one is the most common reason people's changes backfire. Think of each campaign as a ship on a course, where the course is your target ACOS. On course — 30-day and 7-day ACOS both sitting near target. Do nothing. It is sailing correctly. Well below target — say 20% against a 35% target. That is a promising campaign, not a finished one. Raise bids and placement adjustments gradually and let ACOS drift up toward target while spend grows. You are buying more volume at a price you can afford. Above target — 39%, 43%, 47%. Correct downward, gradually, over two to four weeks. And let the gap set the aggression. A campaign at 39% against a 35% target needs a nudge. One at 55% needs real correction. Campaign ACOS against target, at a glance Reading the gap between each campaign's ACOS and its target across a whole portfolio is the input to every decision here — see Campaigns: Budgets, Placement and Audience Adjustments. Rebuilding that view by exporting reports and pivoting them each week is hours of assembly before you have made a single decision. Nurture the young, leave the mature alone A keyword or campaign that launched recently and immediately produced low ACOS and strong conversion is a young player with obvious potential. Invest in it. Raise bids, raise placement adjustments, give it budget, and give it attention. You are trying to turn it into one of your main campaigns. A campaign that has run for weeks or months at an ACOS you are happy with is at its peak. Leave it alone. Frequent tweaking of something already performing has more downside than upside — the realistic outcomes are unchanged or worse. The bands, against a 35% breakeven ACOS 15–24% — exceptional. Push it. Grow spend deliberately and let ACOS rise toward the high twenties. ACOS 25–33% — the sweet spot. Stop optimising and let it run. A campaign spending $150 a day at 28% is worth far more than the same campaign spending $50 a day at 20%. ACOS at or above target — correct it, at the pace the gap justifies. Over-optimising a campaign in the sweet spot is how you push it to 40%. Which is no longer excellent — merely acceptable. When one backfires It will happen. Not if, when. Document changes on your biggest campaigns — the ones carrying a meaningful share of account spend. A short log of what you changed and when. Watch those campaigns for the following few days. A big campaign going from 30% to 39% shows up quickly. When it does, undo the change. The campaign and ad group history in the console shows exactly what you did. Then check it daily until it recovers. Put the bids, placement adjustments and keywords back where they were when it was working. Re-enable what you paused; remove what you added. Do not panic, and do not compound it with a second correction. Most campaigns recover fully. Common mistakes Treating optimisation as arithmetic, and expecting a formula to produce the right bid. Making large changes to large campaigns, where the downside dwarfs the upside. Optimising campaigns that are already in the sweet spot, and pushing them out of it. Making every change the same size, regardless of confidence or campaign size. Not documenting changes on major campaigns, so you cannot undo what you cannot remember. Panicking after a backfire and making a second aggressive change on top of the first. Related - Bid Optimisation: Six Rules — the rules this mindset governs. - Placement Bid Optimisation — the same logic applied to placements. - Setting Target TACOS per Product — where the target you steer toward comes from.

Lifetime Bleeder Targets

Your bleeder report catches targets with 10+ clicks and no sales in the last 60 days — because 60 days is the most historical data Amazon's reports give you. Which leaves a gap. A target that takes nine clicks every 60 days never trips that threshold, and because it never converts it never appears in the profitability report either. So it just keeps spending, quietly, forever. Watch the walkthrough Prefer to watch? One hundred and forty-one seconds on the gap between the 60-day and profitability reports, the four filters that close it, and why the first pass is where the money is. The gap, precisely The 60-day bleeder report catches high-click, zero-sale targets inside a rolling window. The profitability report catches targets that convert badly — but it only sees targets that have converted at least once. A target with nine clicks per 60 days and no sales appears in neither. It is under the threshold of one and invisible to the other. Run this for the first time on an account nobody has cleaned and you routinely find targets with 120, 150, even 180 lifetime clicks and zero orders. Hundreds of dollars each, accumulated a click at a time. The fix: filter on lifetime, not 60 days The campaign manager's targeting tab will filter on Lifetime — and it reports lifetime performance for each target, not just the 60-day slice. Go to the Targeting tab in the campaign manager. Set the date range to Lifetime. Then apply four filters: | Filter | Value | |---|---| | Target status | Enabled | | Campaign status | Enabled | | Clicks | greater than 10 | | Orders | 0 | Export the result. The two status filters matter. Without them you spend your review on targets and campaigns that are already paused, which is wasted effort and inflates the list. Then decide, then pause Review the export and mark what to pause. In practice you will pause almost all of it. A target with 40 lifetime clicks and no order has already had its chance. But the review is worth doing, because occasionally one is a keyword you deliberately want to keep bidding on for ranking reasons. Then pause the marked targets in the campaign manager. The outcome is straightforward: wasted spend removed, ACOS and TACOS down, with no effect on anything that was working. Monthly, not weekly This is a monthly process, unlike the 60-day bleeders. Because the whole point is that these targets accumulate slowly. Once you have cleaned an account, new lifetime bleeders appear at the same slow rate that created them — a weekly run would find nothing most weeks. Expect a large first pass and small ones after that. The first run on an uncleaned account is where nearly all of the money is. Where it sits in the routine This is a third bleeder pass, not a replacement for the other two: 60-day bleeders — high clicks, no sales, recent. Weekly or fortnightly. Profitability bleeders — converting, but above your target. Weekly or fortnightly. Lifetime bleeders — low volume, no sales, accumulated over months. Monthly. Each catches something the others structurally cannot — see The Bleeder Reports and The Nine-Report PPC Routine. Targets and their spend, with no orders against them Surfacing zero-order targets by lifetime spend across every campaign at once is what turns this from an export-and-filter exercise into a decision — see Rule-Based Optimization: Automating Bids and Budgets. The manual version means downloading the targeting export, filtering it, and re-reading it every month. Common mistakes Assuming the 60-day report covers it. By construction it cannot. Forgetting the status filters, and reviewing targets that are already paused. Running it weekly, and concluding it is not worth doing because it finds nothing. Skipping the review and pausing everything blindly, including terms you hold for ranking. Doing it once. The first pass finds the accumulated waste; the monthly rhythm stops it rebuilding. Related - The Bleeder Reports — the two passes this one completes. - The Nine-Report PPC Routine — the cadence it fits into. - Optimisation as a Calculated Bet — why this one is a rare low-risk change.

Winning the Frequently Bought Together Placement

The Frequently Bought Together module sits on the product detail page and shows what shoppers buy alongside the product they are looking at. It is organic, it is trusted, and it costs nothing once you are in it. Amazon decides who appears there, based on real co-purchase behaviour. Which means you cannot buy the placement — but you can manufacture the behaviour that earns it. That is what this campaign does: paid spend, converted into a permanent organic placement. Watch the walkthrough Prefer to watch? One hundred and sixty-six seconds on sourcing target ASINs from five places, forcing the ads onto product pages, and why this campaign runs on a 30 to 60 day clock. The idea Product Targeting ads placed on other brands' detail pages create co-purchase events between your product and theirs. Enough of those events, and the algorithm starts pairing you organically. Two things follow from that. You are targeting competitor and complementary ASINs, not keywords. And you are playing a 30–60 day game, because the algorithm needs accumulated data, not a good week. Step 1: build a target list of 10–30 ASINs Pull candidates from five sources, aiming for 5–10 from each. 1. Your own search term report. Download 60 days, filter the customer search term column for entries starting with B0 — those are ASINs your auto campaigns already served against. Sort by orders and take the top 5–10. These are proven converters. 2. Existing FBT modules on top sellers. Search your main keyword, open the top 3–5 organic listings, and copy every ASIN in their FBT modules. These are products the algorithm has already paired with high-traffic listings. 3. Amazon's shopping assistant. On those same listings, ask it what customers often buy alongside the product. Another 3–5 ASINs, showing what shoppers cross-shop. 4. Competitors you beat on price or rating. Products priced 15–30% above yours, or rated below yours. When your ad appears on their page, the comparison does the work. 5. Market Basket Analysis. Brand Analytics → Consumer Behavior Analytics → Market Basket Analysis, searched on your ASIN. The #1 Purchase Combination data is real co-purchase behaviour rather than inference. Step 2: one campaign, one ad group | Setting | Value | |---|---| | Type | Sponsored Products | | Targeting | Manual → Product targeting (not keyword) | | Name | SP \| PT \| FBT \| [your ASIN] \| [month] | | End date | none | | Daily budget | $30–75, roughly $2–3 per target | | Bidding | Down only | | Base bid | $0.30–0.50 | Add your one promoted ASIN under Products. Then paste the whole target list into Product targeting → Individual products → Enter list. Exact, not Expanded. Leave the refine filters — rating, price, brand — blank. You are targeting specific ASINs, not a category. Each ASIN gets its own editable bid, which is the point of putting them all in one ad group. Step 3: force the ads onto product pages This is the step that decides whether the campaign works at all. Low base bid, high placement modifier on Product pages. Set the Product pages adjustment high — start around 100% — and leave Top of search and Rest of search at 0%. Because the FBT module lives on the detail page, and a co-purchase signal only forms when your ad is seen next to the product you want to be paired with. Spend that lands in search results buys clicks but generates none of the behaviour you are trying to create. Verify it rather than assuming it. After a week, pull the placement report and confirm 80%+ of impressions are on product pages. If not, raise the Product pages modifier further. Placement split, where the spend is actually going That placement check is the campaign's health metric — see Placement Bid Optimisation. A campaign with 70% of its spend in search is not a slightly worse version of this strategy; it is a different campaign that will never produce the outcome. Step 4: run it, and judge it on the right things Week 1, daily: confirm every target is getting impressions. Zero impressions after three days → raise the base bid $0.05. Budget maxing out is a good sign; consider doubling it. Weekly: check the placement split. Then sort the Targeting tab by total cost and pause any single target that has spent 3× your product price with zero conversions. Leave the rest running. Monthly: open the FBT module on each target ASIN in an incognito window and look. Expect a higher ACOS than your search campaigns, and accept it. You are buying a permanent organic placement, not this month's ROAS. Judging this campaign on ACOS alone will make you kill it two weeks before it works. Step 5: when you win a placement Reduce the bid on that specific target ASIN. You now hold the position organically; the goal from here is to keep it cheaply. Keep bidding normally on the targets you have not won yet. Do not pause the campaign. The remaining targets are still accumulating, and the ones you have won benefit from continued reinforcement. Common mistakes Raising the Top of Search modifier, which spends the budget in exactly the place that cannot produce FBT signals. Using keyword targeting, which cannot place you on a specific competitor's page. Judging it at two weeks. Thirty to sixty days is the honest minimum. Killing it on ACOS, when the return is an organic placement rather than a paid conversion. Spreading the budget too thin across 40 targets at $30 a day, so none accumulates enough signal. Never checking the actual module, and so not knowing whether you have already won. Related - Placement Bid Optimisation — the mechanic this campaign depends on. - Easy-Win Campaigns — other high-leverage campaign structures. - Optimisation as a Calculated Bet — why this one runs on a longer clock.

Sponsored Brands Video at Scale

Most sellers use Sponsored Brands Video badly or not at all, which is exactly why it is worth using well. Less competition on the format, and a creative that carries far more information than a static ad. The method has three parts: make three videos, spend a fixed amount finding out which one wins, then scale only the winner across a structured set of keyword groups. The result you are aiming for is a low-ACOS profit campaign — not rank. Watch the walkthrough Prefer to watch? One hundred and seventy-two seconds on testing three videos on identical traffic, grouping keywords by order count, and optimising each group against breakeven. Step A: make three videos, not one Three is the number because the whole method rests on having something to compare. Brief them as separate jobs, not one job asking for three. Separate briefs give you genuinely different creative; a single brief produces three variations of the same idea, which tells you nothing. Landscape, always. Amazon's video placements display landscape properly and portrait poorly. Subtitles are mandatory. These ads play on mute. A video whose message depends on audio is a video with no message. Check the current ad specs before commissioning — dimensions, length and file requirements change. Four concepts worth briefing: Social-style clip — fast, native, informal. Unboxing — shows scale and what actually arrives. Product in use — answers the "how does this work" question. Review — a person explaining why it is good. Match the presenter to your actual buyer. Brand Analytics → Consumer Behavior Analytics → Demographics tells you who is buying, so you are not guessing at age or audience. Step B: test them against 200 clicks each Pick your 5–10 core keywords — the terms that already drive most of your sales from other campaigns. Then build three identical campaigns, one per video, changing nothing except the creative. | Setting | Value | |---|---| | Type | Sponsored Brands | | Goal | Drive page visits | | Ad format | Video | | Landing page | Product detail page | | Targeting | Keyword, Exact only | | Keywords | your 5–10 core terms | | Bid | 85% of suggested | | Daily budget | $10–20 | | Name | [Brand] [Product] Test Video 1 | Uncheck Broad and Phrase, and remove the keyword themes. Identical, controlled traffic is the entire point — a difference in match types makes the comparison meaningless. Check daily that each is spending. If one is not, raise its bids 7%. A video that never gets clicks is not a losing video; it is an untested one. Set the date range to Lifetime, and pause each campaign at 200 clicks. Then compare: lowest ACOS and highest CTR wins. 200 clicks is the point at which the difference between videos is a real difference rather than noise. Calling it at 50 clicks is how people scale the wrong creative. Step C: scale the winner across ordered keyword groups Now build five campaigns using only the winning video, split by how well each keyword already converts. Download a 60-day bulk report — Sponsored Products data and Sponsored Products search term data only. Open the SP Search Term Report tab and group the search terms by order count: | Group | Orders | |---|---| | 1 | 2–3 | | 2 | 4–5 | | 3 | 6–7 | | 4 | 8–9 | | 5 | 10+ | Filter at the campaign or portfolio level for the product you are working on, otherwise you pull in terms belonging to other products. Then create one campaign per group: Phrase match only. Maximum 15 keywords per campaign. Bid 55% below suggested. Name them so the group is visible — [Brand] [Product] [SBV] [Phrase 2-3]. The grouping is what makes this work. Keywords with similar proven conversion behave similarly, so each campaign can be optimised as a unit instead of averaging strong and weak terms together. Optimising the scaled campaigns First, calculate your breakeven ACOS, because every rule below is relative to it: Gross profit before PPC = price − COGS − FBA fee − referral fee. Breakeven ACOS = gross profit before PPC ÷ price. Then review every three days, on a 30-day window: | Condition | Action | |---|---| | ACOS 10% below breakeven | Raise bid 5% | | ACOS above breakeven | Lower bid by the difference | | ACOS more than double breakeven | Pause the keyword | | Under 3 clicks | Raise bid 5% | | 15 clicks, no sales | Pause the keyword | "Lower by the difference" is literal: ACOS 47% against a 39% breakeven → lower the bid by 8%. Raise budgets on the low-ACOS campaigns that are hitting their cap — $20, then $50, then $100, as long as ACOS holds. And keep feeding them. Re-check the search term report regularly, and add newly qualifying terms into their matching group, still capped at 15 per campaign. Keyword performance against your breakeven Grouping search terms by order count and checking each one against breakeven ACOS every three days is the whole workload here — see Bid Optimisation: Six Rules and Rule-Based Optimization: Automating Bids and Budgets. Done by hand it is a bulk export, five filters and a rebuilt sheet, twice a week, per product. What this is for Profit, not rank. These campaigns should make money — see Profit Campaigns and Ranking Campaigns. If the ACOS will not come down, the answer is to stop, not to push through it. A ranking campaign can justify a bad ACOS; this one cannot. Common mistakes Testing one video, which gives you nothing to compare. Briefing three videos as one job, and getting three near-identical results. Calling the winner early, before 200 clicks. Letting the test campaigns differ in match type, bid or budget, so the comparison measures the settings rather than the creative. Skipping subtitles, on ads that play silently. Putting all keywords in one scaled campaign, which averages away the grouping that makes the optimisation rules work. Running it as a ranking play, and tolerating an ACOS that defeats the purpose. Related - Create a Sponsored Brands Video Campaign — the mechanics of the campaign type. - Profit Campaigns and Ranking Campaigns — which of the two this is. - Bid Optimisation: Six Rules — the wider version of the rules above.

Bulk Auto Campaigns for Peak Season

In peak season there is far more traffic than your existing campaigns are structured to capture, and it does not behave the way it does in March. Running five auto campaigns per product at different bids is a way of buying that traffic without having to predict which bid is right. You are not optimising. You are sampling — then killing the losers fast and tripling down on the winners. Watch the walkthrough Prefer to watch? One hundred and seventy seconds on building the bid ladder from your average CPC, the daily kill-and-scale rules, and watching TACOS rather than ACOS alone. Who this is for Your best-selling products only, and specifically those that do at least 1.5× their normal sales in peak. Not the whole catalogue. This tactic needs daily attention, and spreading it across twenty products means none of them gets the daily attention it depends on. Step 1: launch five auto campaigns per product Five campaigns, identical except the bid. Set the bid ladder from your current average CPC. The bids sit below it, spaced evenly: | Your average CPC | Bid ladder | |---|---| | $1.80 | $0.75 / $0.80 / $0.85 / $0.90 / $0.95 | | $1.00 | $0.35 / $0.40 / $0.45 / $0.50 / $0.55 | All four targeting groups on: close match, loose match, substitutes, complements. Close match bid 15–20% higher than the other three. Top of Search modifier 30%. Budget $10–30 per campaign. The ladder is the mechanism. You do not know which bid clears in a market where everyone's CPC is inflated, so you buy five answers and let the data pick. Step 2: the first seven days Checked daily. Not weekly — daily. Peak season compresses everything, and a campaign wasting money for six days has wasted most of its season. Three checks: 1. Is it spending? Under 75% of budget → raise under-clicked target bids 7%, or raise Top of Search 15%. 2. What is bleeding? A target with 15+ clicks and no sales → pause the target. A campaign with bad ACOS or heavy clicks and no sales → pause the campaign. 3. What is working? A campaign at good ACOS → triple the budget. Raise bids on strong targets 5–7%. Raise placement bids where placements perform. Define good and bad against breakeven, not by feel Bad ACOS = 15% above your breakeven. Good ACOS = at least 5% below your breakeven. Anything between the two is neither — leave it alone and let it run another few days. One exception on placements: do not raise the Product pages bid, even when it looks like it is performing. Detail-page traffic in peak converts differently and scales badly. Campaigns against target, sorted by what needs attention Ranking every campaign against its breakeven daily is what makes a five-campaign-per-product structure manageable — see Rule-Based Optimization: Automating Bids and Budgets. Twenty-five auto campaigns across five products, checked by hand every morning through Q4, is a full-time job. Step 3: expand what is working Once the first five show real performance, launch five more at different bids. Same optimisation rules. Then launch a third set with dynamic bids up-and-down, to test whether letting Amazon raise bids improves things. In peak it often does — but test it rather than assuming it. Then roll the whole tactic out to additional products in the portfolio. The sequence matters. Expand only after the first set proves itself on this product, in this season. Peak behaves differently every year. Step 4: keep reallocating Profitable and hitting its budget cap → raise the budget. High ACOS → cut the budget or pause. Move the money from the second group to the first, continuously. And if product TACOS is where you want it and budget is available, create more auto campaigns. There is no natural limit here other than TACOS and attention. Watch TACOS daily, not ACOS alone. ACOS tells you whether a campaign is efficient; TACOS tells you whether the whole product is still profitable while you scale — see Setting Target TACOS per Product. Why it works Auto campaigns find search terms your manual campaigns do not have, and peak season brings shoppers using language your keyword research never saw. Five bids per product means you are not betting on one guess about what a click is worth in an inflated market. And killing losers on a 15-click rule means the failures are cheap while the winners get tripled. It is deliberately a volume strategy. Most of the campaigns will be cut. That is the design, not a failure of it. Common mistakes Running it on average products instead of the ones with genuine peak uplift. Checking weekly. The whole method depends on a daily cycle. Defining good and bad ACOS by feel rather than against breakeven. Scaling budgets without watching TACOS, and growing revenue into a loss. Raising Product pages placement bids because they look good in week one. Expanding to more products before the first set has proven itself. Leaving it running after peak. These bids and budgets are built for inflated traffic; in January they are simply overspending. Related - Easy-Win Campaigns — the year-round version of buying cheap traffic. - Prepare for Prime Day: Budget Rules and Inventory Cover — the same discipline at a shorter event. - Setting Target TACOS per Product — the ceiling all this scaling runs against.

Tracking Branded Market Share

Branded PPC is the hardest spend in the account to justify, because the counterfactual is invisible. If you stop bidding on your own brand name, do you lose those sales — or would they have come anyway? Branded market share is the metric that answers it, and the answer only appears as a trend, not as a snapshot. Watch the walkthrough Prefer to watch? One hundred and fifty-six seconds on defining brand terms correctly, building four weeks of baseline, and testing branded spend against the share trend. What branded market share means here Of all the purchases made on search terms containing your brand, what share went to you? If it sits near 100%, you already own that traffic — and paying for it may be buying sales you would have had for free. If it drops when you reduce branded spend, the spend was defending something real. That is the feedback loop. Without it, branded budget is argued about rather than decided. Step 1: define what counts as branded List the brand terms — your brand name and its common misspellings. Nothing else. Do not list product phrases. If your brand is Northwind, you list northwind. You do not list northwind blanket, northwind sheets, or any other product-modified variant. Because the rule is containment: any search term containing your brand token is flagged as branded, and the modified phrases are caught automatically. Listing them separately only creates gaps and double-counting. Add misspellings, which are common in branded search and easy to miss. Set this up once, and revisit it only when the brand itself changes. Step 2: pull the branded view weekly Brand Analytics → Search Query Performance → Brand view, weekly. Weekly, not monthly. The whole point is to see a trend responding to a change you made, and a monthly cadence gives you four data points a year. Every week, you want the same three things: Top branded search terms and their purchases. Top non-branded terms and their purchases. Branded versus non-branded totals, in both absolute and percentage terms. Branded and non-branded queries, week by week Classifying every query as branded or non-branded and tracking the split week over week is the entire measurement — see Generic, Branded and Competitor Terms and Understanding Search Query Performance. The manual version is a fresh export, a fresh sheet tab and a manual paste every single week — which is exactly the kind of task that gets done for six weeks and then quietly stops. Step 3: establish a baseline before changing anything Four weeks minimum, with branded spend unchanged. One week tells you nothing. You need to know what normal variation looks like before you can recognise a response. Most established brands sit somewhere between 80% and 100% on branded market share. Where you sit is less important than how stable it is. Step 4: then test Change branded spend deliberately, and watch the chart. If your branded share is already very high Reduce branded PPC spend and watch. Share stays flat → the spend was buying sales you were getting anyway. Move that budget to non-branded terms, where it buys incremental customers. Share falls → the spend was defending real ground. Put it back. If your branded share is lower than you expected Someone else is converting your brand searches — usually competitors bidding on your name. Then branded spend is genuinely defensive, and the question becomes how much it takes to hold the position rather than whether to spend at all. See How Much to Spend on Branded Keywords. Read it as a trend, not a number Change one thing at a time, or you cannot attribute the movement. Give each test three or four weeks. Branded search volume is seasonal and lumpy; a single week's dip is usually noise. And watch non-branded purchases at the same time. Branded share rising while total non-branded purchases fall is not a win — it usually means you have simply got smaller. Why this is worth the effort Branded terms convert extremely well, which makes branded campaigns look excellent on ACOS and makes them very easy to over-fund. A campaign with a 5% ACOS is not automatically a good campaign. If those buyers were coming anyway, its true incremental ACOS is infinite. This tracker is the only practical way to find out — see Profit Campaigns and Ranking Campaigns. Common mistakes Listing product phrases as brand terms, which breaks the classification. Forgetting misspellings, and under-counting branded traffic. Judging on one week, before a baseline exists. Changing branded spend and something else at the same time. Reading branded share alone, without watching non-branded purchases beside it. Assuming a low branded ACOS proves the spend is working. Related - How Much to Spend on Branded Keywords — the decision this measurement informs. - Generic, Branded and Competitor Terms — the classification underneath it. - Understanding Search Query Performance — the report this all runs on.

In-House PPC or an Agency

Who runs your advertising is one of the more consequential decisions you will make about the brand, and it is usually decided on cost — which is the wrong axis entirely. Here is the honest case on both sides. Watch the walkthrough Prefer to watch? One hundred and sixty-four seconds on why speed of execution compounds, how many hours an agency account really gets, and the performance gap that is the real cost. The case for in-house Speed of execution This is the largest single advantage, and it compounds. Stock changes. Running low or back in stock — you adjust spend the same hour instead of waiting. Deals and promotions. Budget adjusted in real time to protect profit through a Lightning or Best Deal. Big sales days. Prime Day and Black Friday reward hourly attention. Raising budgets on the campaigns that are working, while they are working, is how record days actually happen. Target TACOS changes applied the same day. New campaign types tested when you think of them. Performance problems investigated the moment you notice them. Each of these is small on its own. Across a year they are most of the gap between adequate advertising and excellent advertising. An agency can do all of it. It just happens more slowly — and speed here is not a convenience, it is the return. Alignment with the actual business Your team knows the brand, the products, the customers, the competitors, the keywords and what has already been tried. That context is not transferable in an onboarding call. And an agency, structurally, optimises advertising performance. You optimise business profit. Those are not the same objective. Most agencies now track profit, but their processes are not built to maximise it — see Setting Target TACOS per Product. In-house also means tailored process. An agency runs a standardised system across every client, by design. That is how it stays economic — and it is why the last increment of performance is hard to buy. The arithmetic of attention An account manager typically handles 6–12 accounts. Take 8 accounts across a 45-hour week — a conservative assumption. That is 5.6 hours per account, and closer to 5 after breaks. Five hours a week is not enough for full advertising management on a serious account. It produces something; it does not produce excellence. And the limited time forces automation — standard rules applied across every account. That is genuinely good at getting you to about 7 out of 10. It is not how you get past it. Then ask who is actually doing the work. Best case a Western-based manager; often an offshore assistant. Consider that alongside five hours a week when you judge what you are buying. The case for an agency Expertise you do not have yet Advertising rewards depth. Overbidding, poor targeting, ignoring new ad formats — these are expensive mistakes and easy to make. Reaching a genuinely expert level takes real time, and Amazon keeps changing, so it takes ongoing attention rather than a one-off effort. Time — the strongest argument Managing this in-house consumes a meaningful part of every week, even with a clean system. And it grows. More products, more campaigns, more time. That time comes out of inventory planning, product development, and new marketplaces — which are not small things to sacrifice. If you genuinely do not have the time to learn this and run it, an agency is a real solution to a real problem. How to decide The honest answer is that it depends on you — your time, your comfort with numbers, and what else is competing for your attention. But be clear about what the cost actually is. Not the fee. An agency at $2,500–3,000 a month is $30–36k a year, and that is the number everyone fixates on. The real cost is the performance gap. If you spend $50k a month on ads and your ACOS is two points worse than it could be, that is $12k a year on its own — before counting the growth you did not get. Which cuts both ways. An agency that outperforms what you would do yourself is worth well above its fee. One that underperforms costs far more than it charges. What the spend is actually returning Knowing which side of that line you are on requires measuring performance rather than assuming it — see Score Your PPC Performance. That assessment is the same whether the work is done in-house or bought in. The pattern worth knowing Among successful Amazon brands, the large majority run advertising in-house — and they treat it as a skill they have deliberately built. There are real exceptions, including very large and very well-run brands using agencies successfully. But the pattern is strong enough to be worth weighing. And most of the argument for in-house rests on time and expertise being scarce. Both are addressable: a defined process removes most of the expertise problem, and the right tooling removes most of the time problem — which is what the rest of this training is for. See The Brand Manager's Calendar. If you take it in-house Give it six to nine months. Three months in you will be markedly better than on day one; by month seven the difference is substantial. Follow a defined process from the start rather than improvising — see The Nine-Report PPC Routine. And measure the handover honestly. Score performance before you change anything, so you can tell whether the switch worked. Common mistakes Deciding on the fee rather than on the performance gap. Taking it in-house without allocating the time, and doing it badly. Assuming an agency's standard process will beat a well-run in-house one on a mature account. Switching without a baseline, so you never learn whether it helped. Underestimating what in-house costs you in inventory and product work. Related - Score Your PPC Performance — the baseline this decision needs. - The Brand Manager's Calendar — what in-house actually costs in hours. - What Amazon PPC Is Actually For — the objective both options are serving.

Amazon's AI Video Generator

The barrier to Sponsored Brands Video has always been the video. Commissioning creative takes weeks and costs real money, which is why most sellers test one video and scale it regardless of whether it was any good. Amazon's own generator removes that constraint. It builds videos from the images, title and description already on your listing — which means you can produce tens of them and keep only the ones that work. Watch the walkthrough Prefer to watch? One hundred and seventy-one seconds on generating from listing assets, the edits every output needs, and testing twenty variations instead of three. Where it lives Campaign Manager → Creative tools → Creative Studio. Three tools sit there: an image generator, a video generator, and an inspiration library. Open the video generator and choose Generate video. Generating Pick a product and generate. Amazon produces three different videos from your existing listing assets — images, title, description. Then watch them and judge. Some will be usable, some will not. That is expected, and it is why three come back rather than one. Editing — where the value actually is The raw output is a starting point, not a finished ad. Every generated video needs a pass. What is worth fixing: Text placement. The generator routinely puts copy over the product itself, which defeats the point of the shot. Move it. The copy. The generated line is generic. Replace it with something specific to what your product actually is — see Answer the Complaints in Your Content for what specificity buys you. Your logo, added where it belongs. Which images appear, and in what order. A call to action at the end. Scenes, added or removed as the story needs. Narration and on-screen text. The text edit is the highest-value change. These ads play on mute, so the on-screen copy carries the message entirely — see Sponsored Brands Video at Scale. Straight into a campaign Create Sponsored Brand Campaign generates the campaign from the video directly, and you carry on through the normal setup. Convenient — but do not let it replace the testing structure. A campaign built in two clicks is still a campaign that needs the right keywords, match types, bids and budget. What this actually changes It changes the economics of creative testing. Commissioned video means three videos and a real cost, so the test is expensive and you run it once. Generated video means ten or twenty variations at effectively no cost — so you can test properly, discard most of them, and keep the few that genuinely perform. The method does not change: Test on identical traffic — same keywords, same match type, same bid, same budget, differing only in the creative. Judge at 200 clicks each. Lowest ACOS and highest CTR wins. Scale only the winner — see Sponsored Brands Video at Scale. Campaign performance, one creative against another Comparing several near-identical campaigns on CTR and ACOS is the part that decides which creative survives — see Campaigns: Budgets, Placement and Audience Adjustments. With twenty generated videos in flight, doing that from exports is where the method quietly breaks down. What it will not do It cannot show your product being used in a way your listing images do not already show, because it only has your listing images to work with. It is not a substitute for genuine demonstration or review-style creative where that is what the product needs. And it inherits your image quality. Weak listing images produce weak videos — see Why the Main Image Decides Your CTR. So treat it as a way to test many angles cheaply, and commission real creative once you know which angle works. Common mistakes Running the raw output unedited, with copy sitting over the product. Generating one video and skipping the comparison entirely. Letting the one-click campaign button stand in for a properly structured test. Ignoring on-screen text, on ads that play silently. Expecting it to fix poor listing images, which are its only input. Related - Sponsored Brands Video at Scale — the testing and scaling method these videos feed. - Create a Sponsored Brands Video Campaign — the campaign mechanics. - Why the Main Image Decides Your CTR — the assets the generator is built from.

The PPC Management Workflow

The reason a nine-report routine is workable at all is that you do not do most of it. The system splits into two roles. One person produces the reports and executes the changes; you make the decisions. That is roughly an 80/20 split of the work, deliberately arranged so that your 20% is the part that actually moves performance. Without that split, the routine is not efficient — at which point you would be better off optimising campaigns by hand. Watch the walkthrough Prefer to watch? One hundred and seventy-five seconds on the four-step workflow per report, why the report categories have to be complete, and the week-5 to week-25 ramp. The structure Nine recurring reports: Four weekly. Two fortnightly. Three monthly. Together they cover every aspect of campaign optimisation — bids and placements, cleaning up waste, adding new targets, and the periodic view of overall performance. See The Nine-Report PPC Routine and The Brand Manager's Calendar. Run all nine on their cadence and you can be confident nothing is being missed. That completeness is the point of having a fixed set rather than doing whatever seems urgent. The workflow, per report 1. The assistant builds the report — pulls the export, filters it, formats it. 2. You review the data and make the optimisation decisions. 3. You document each decision in the report. 4. The assistant implements them in the campaign manager. Your entire role is step 2 and 3. Reading, deciding, and writing the decision down. Steps 1 and 4 are the bulk of the hours and have almost no direct effect on performance. They are assembly and execution. What you review, and what gets executed Steps 1 and 4 are exactly the parts that shrink when the data assembles itself and the changes apply through rules — see Rule-Based Optimization: Automating Bids and Budgets. The judgement in step 2 is yours either way; the question is how many hours sit either side of it. What each report is for They are deliberately different in purpose, which is why the set works as a whole: Bids and placements — steering existing spend toward what performs. Cleaning up waste — cutting targets that spend without returning, in three different windows. Growth — harvesting new targets and adding campaigns. Overview — the periodic 360° read on where performance actually stands. Miss a category and you develop a systematic blind spot, which is precisely what these routines exist to prevent. Resourcing it The assistant is not optional in this model. Finding and training one takes two to three weeks, so start before you need them — see Onboarding a VA. And train on the processes themselves, not on general Amazon knowledge. What you need is someone who can reliably produce the same report the same way every week. Expect a real learning curve This is the part people are not warned about, and it causes more abandonment than any of the tactics. At first it takes longer than you expect. That is normal. By week 5 — around 25–30% faster than week 1. By week 10 — noticeably faster, and the decisions are better. By week 25 — seamless, and completed with minimal effort. So judge the system at week 10, not week 2. The first few cycles are you learning what each report is telling you, and that cost is paid once. If you abandon it in month one because it is slow, you paid the entire learning cost and collected none of the benefit. Common mistakes Running the system without an assistant, and spending all your time on assembly. Doing the reports yourself "just this once", repeatedly. Making decisions without documenting them, so nothing can be executed or reviewed later. Dropping reports from the set because a week is busy, and creating a blind spot. Judging the system in week two, at the peak of the learning curve. Delegating the decisions as well as the assembly. Step 2 is the part that is actually yours. Related - The Nine-Report PPC Routine — the reports themselves. - Onboarding a VA — resourcing the other 80%. - Optimisation as a Calculated Bet — how to make the decisions in step 2.

Acting on Your PPC Score

You have scored your advertising across eight reports, one to ten each — a maximum of 80. See Score Your PPC Performance. Above 50 is a good sign. Below 50 is a bad one. What you do next depends on that score and on who currently runs your advertising. Four paths, and picking the wrong one wastes months. Watch the walkthrough Prefer to watch? One hundred and fifty-nine seconds on patching gaps versus rebuilding, the three options when an agency underperforms, and baselining so the next review is not another guess. If you run advertising in-house Score 7+ per report — good performance Add only the specific processes that scored badly. If two of eight reports scored low, add those two. Your existing system is producing good results, so a wholesale replacement risks more than it gains. This is a small lift with a clear return — you are patching identified gaps rather than rebuilding. Score 6 or lower — poor performance Implement the full system. All the reports, on their cadences, with the weekly, fortnightly and monthly rhythm — see The Nine-Report PPC Routine and The PPC Management Workflow. Because a low score across several reports usually means several critical processes are missing entirely, and adding one at a time to a structurally incomplete system does not compound. If an agency runs your advertising Score 7+ — good performance Same answer: add only the low-scoring processes. Your agency is doing well. Ask them to add the specific missing reports, and give them the criteria rather than a complaint. Score 6 or lower — poor performance Three options, and there is no single right answer. It depends on your size, product count, capability and available time. Option 1: bring it in-house. Right if you want to build the skill and can commit roughly seven hours a week. It is the strongest long-term position — see In-House PPC or an Agency — and it is the option most people underestimate the time cost of. Option 2: switch agencies. Gather free audits and quotes from several, compare them, and pick the best fit. The audits are useful on their own, because several independent reads of your account will agree on the real problems. Option 3: keep the agency and have them add the missing processes. Honestly the weakest of the three, and worth knowing why before you choose it. A low score usually means many critical processes are missing entirely, which is a large amount of new work for an agency operating a standardised system. In practice this works sometimes and fails often — not through bad faith, but because the volume of additional process is beyond what the engagement was priced for. Choose it when the relationship is genuinely good and you would rather fix than replace — with clear expectations about how much you are asking for. What the score is measuring, per campaign Whichever path you take, re-score at the same eight reports in a few months — see Campaigns: Budgets, Placement and Audience Adjustments. Without a repeatable score you cannot tell whether the change worked, and every one of these paths is expensive to get wrong twice. The decision in one table | | In-house | Agency | |---|---|---| | Score 7+ | add the low-scoring processes | ask the agency to add them | | Score ≤6 | implement the full system | in-house, switch agency, or add processes | Whatever you choose, baseline first Record the score before you change anything. Then re-score after a defined period. That is the only way to know whether the decision was right, and it costs one afternoon. Common mistakes Rebuilding everything on a good score, and breaking a system that worked. Adding one process at a time to a score of 4, where the gaps are structural. Bringing it in-house without the seven hours a week, and doing it worse than the agency did. Skipping the free audits when considering a switch. Keeping a failing agency and asking them to absorb a system's worth of new process. Not baselining, so the next review is another guess. Related - Score Your PPC Performance — producing the score this decision runs on. - In-House PPC or an Agency — the fuller case on both sides. - The PPC Management Workflow — what implementing the full system involves.