Playbooks & Use Cases

Protect the Products That Bring You New Customers

Most catalogs have one or two products that do the recruiting. They bring in people who have never bought from the brand before, and everything else sells to the customers those products earned. Those products are worth defending differently from the rest of the catalog — and most accounts do not know which ones they are. Time: about 30 minutes. Do it monthly. Watch the walkthrough Prefer to watch? Eighty-three seconds on finding the gateway ASINs and funding them differently. Step 1 — Find the gateway products AMC Cloud → NTB ASINs Analysis. Top ASINs by new-to-brand users, and the share each one drives Read the two charts in the right order, because they answer different questions: The doughnut first. It shows what share of all your new customers each ASIN brought. This is volume, and it is what actually matters to the business. Write down the products above roughly 10%. Then the bars. These show each ASIN's rate — what proportion of its buyers were new. A product can be 80% new-to-brand and irrelevant, because almost nobody buys it. Check the legend. On this dashboard the NTB segment is the lighter one, on the left. The other NTB dashboards do it differently. If your per-ASIN rates do not roughly average to the headline NTB Share tile, you have read them backwards. What you are looking for: the shortlist of ASINs that between them account for most of your new customers. It is usually two or three, and it is usually more concentrated than anyone expects. Step 2 — Check they cannot run out Inventory. Days of cover for each product For every product on your shortlist, look at days of cover and the inbound quantity. A stock-out on a gateway ASIN is not a sales problem for one product. It is an acquisition outage for the whole brand — for as long as it lasts, the thing that recruits your customers is switched off, and the effect shows up months later as a thin repeat cohort. Set a higher bar for these products than for the rest of the catalog. If your general reorder point is three weeks of cover, make it five or six here. The extra holding cost is cheap next to a month of not acquiring anybody. Step 3 — Alert on them specifically Insights & Alerts. The stock alerts fire on the whole catalog at fixed thresholds — Low Stock below three weeks of cover, Critical Stock below one week, Stockout Forecast at fourteen days' projected runway. See Alert Categories, Priorities and What Actually Triggers Them. Those are the right rules and the wrong priority. A Critical Stock alert on a long-tail ASIN is routine; the same alert on a gateway product is an emergency, and the alert screen cannot tell the difference. So keep the shortlist somewhere you will see it — a rule group name, a saved view, a note in the report you read every Monday. When an inventory alert fires, the first question is "is it one of the three?" Step 4 — Fund them differently Gateway products deserve a worse ACoS target than the rest of the catalog, and that is not a typo. You are not buying that sale. You are buying a customer, and the return arrives on their second, third and fourth orders — none of which will ever be attributed to the ad you paid for. Judging a recruiting product on the efficiency of a single transaction guarantees you underfund it. In practice: - Give gateway ASINs their own rule group, so bid rules can treat them separately. See Rule Groups. - Set their Target ACoS above your catalog average — how far above depends on how much a customer is worth to you over a year. - Never let an automated pause rule touch them. Add them to a group your pause rules do not run on. Step 5 — Watch their rank, not just their sales Rank Tracker. The tracked projects, with visibility and Amazon's Choice per project Put the shortlist in their own Rank Tracker project. Two things to watch: Visibility and Top 10 — a gateway product sliding out of the first page stops being discovered by people who do not know you, which is precisely its job. Amazon's Choice — the badge is worth more here than anywhere else in the catalog, because it moves the shopper who is not comparing. That shopper is, by definition, new to you. What good looks like After a month of this you should be able to answer four questions without opening anything: - Which two or three ASINs bring most of our new customers? - What is their stock cover today? - What ACoS are we willing to pay on them? - Has their rank moved this month? If any answer is "I would have to check", that product is not yet protected. What not to do Do not confuse a gateway product with a best seller. They are often different, and the best seller usually sells to people who already know you. Do not act on one month. NTB share moves with seasonality — gifting periods inflate it. Confirm the shortlist across two or three months before you rebuild budgets around it. Do not assume it is permanent. A launch can become the gateway within two months. Re-run step 1 monthly; it takes five minutes once you know where to look. Related - NTB ASINs Analysis: Finding Your Gateway Products — the dashboard in step 1. - Inventory: On Hand, Inbound, Days of Cover and Velocity — step 2. - Rule Groups — how to fund them separately. - Catch a stockout before it costs you rank — the companion playbook. - Why Inventory Is the Second Pillar — why a stockout on a gateway product costs most.

Cut Wasted Ad Spend in an Afternoon

Every account carries spend that is not buying anything. Not underperforming spend — spend with nothing on the other side of it. It accumulates quietly, and it is the easiest money you will ever find. This is a single sitting, in order, from the biggest wins to the smallest. Time: about two hours. Do it quarterly. Watch the walkthrough Prefer to watch? Seventy-seven seconds on the order to work in, and the rule setting to get right. Before you start Set your date range to the last 30 days, and know your break-even ACoS — the point where an extra sale stops making you money. Everything below is judged against that number, not against a number you like. Step 1 — Take the recommendations first Search Query Performance → Recommended Actions. Recommended actions, sorted by money impact This is the shortest path to the largest wins, because the work of finding them is already done. Filter to Cut bids and sort by wasted spend. Each row names the query, the money at stake, a Why and a Suggestion. The two shapes worth acting on immediately: High ACoS, low conversion — "this term does not convert for this catalog". Negative-match it, or pause the keywords carrying it. High ACoS, decent conversion — "the term works but the clicks are overpriced". Lower the bid. Do not pause it; you would be throwing away a term that converts. The distinction is the whole point of the screen. Pausing an overpriced-but-converting term is the most common self-inflicted wound in this exercise. Competitor-brand queries are deliberately absent from these recommendations, because a better page cannot win a shopper who wants somebody else's brand. If a rival is missing from your competitor list, its queries will keep appearing here with advice that cannot work. See Generic, Branded and Competitor Terms. Step 2 — Find the terms that never convert Amazon Ads → Search Terms, with Aggregated Search Terms switched on. Search terms, aggregated across the account Aggregation matters here. In the default view the same customer search appears once per keyword per ad group, so a term wasting real money looks small in every row and large in none. Sort by spend, descending. Then look for: Clicks above ~25, orders zero. Twenty-five clicks is roughly where "not enough data" becomes "this does not convert" for most catalogs. Below that you are guessing. Spend above your threshold, orders zero. Use spend rather than clicks when your click costs vary a lot — a $0.20 click and a $5 click are not the same evidence. Add these as negative exact matches in the ad groups they came from. Not campaign-wide unless you are certain: a term that fails in one ad group sometimes works in another with a different product behind it. Step 3 — The overpriced converters Same screen, different filter: terms that do convert but at an ACoS well above break-even. These are bid problems, not relevance problems. Lower the bid on the keywords feeding them by 15–25% and let them run. If the term still converts at the lower bid, you have just taken a loss-maker into profit. If it stops converting entirely, it was only ever buying position. Do not cut more than about 25% in one move. You will not learn anything from a change large enough to kill the term outright. Step 4 — The capped campaigns nobody noticed Check for Campaign Daily Budget Capped in your alerts. This is the opposite of waste and belongs in the same sitting, because you are already in the numbers. A campaign hitting its budget before the day ends is turning away traffic it could have converted — and if its ACoS is below break-even, that is money you are choosing not to make. Move the budget you freed in steps 2 and 3 here first, before anywhere else. Step 5 — Stop it coming back Automations → Create Rule. A pause branch: enough clicks to prove it, and no orders Everything above is manual work you should not have to repeat. Two rules will hold most of it: A pause rule for dead weight. Clicks greater than 25 and orders equal to 0 → pause targeting. The built-in Pause dead keywords recipe is exactly this. A bid rule for the overpriced. ACoS above your ceiling → lower the bid, capped at 10–15% per run. Two things to get right before you enable them: Set the look-back to 14 days and ignore the freshest day. Yesterday's spend is complete but yesterday's sales are not — a pause rule that reads right up to yesterday will pause things that converted and have not been credited yet. See Look-back Windows and Adjustment Frequency. Dry-run it first. The preview computes every change and applies none, and it shows which branch claimed each target. See Dry-Run Preview. What good looks like A first pass on an account that has never had one typically finds 5–15% of spend with nothing behind it. A second pass three months later should find much less — if it finds just as much, your rules are not doing their job. Write down what you cut and what you expected to happen. The check in a month is not "did spend fall" — it is "did sales hold". What not to do Do not pause on thin evidence. Four clicks and no orders is not failure; it is barely a test. Do not negate broadly on the first pass. Exact negatives in the specific ad group. You can widen later; you cannot easily discover what a broad negative silently killed. Do not do this and a bid-strategy change in the same week. You will not know which one moved the numbers. Do not forget the freed budget. Cutting waste and leaving the money unspent is a smaller account, not a better one. Related - Recommended Actions: The Queries That Need a Decision — step 1. - Search Terms and Aggregated Search Terms — step 2. - Building a Rule from Scratch — step 5. - Guardrails to Set Before You Automate Anything — before you enable anything. - The Bleeder Reports — the same job across three time windows, including the lifetime pass. - Bid Optimisation: Six Rules — tapering bids before pausing outright. - Easy-Win Campaigns — where low-bid auto campaigns fit alongside the cleanup.

Catch a Stockout Before It Costs You Rank

A stock-out costs you more than the sales you miss while it lasts. Best Seller Rank decays with no sales, organic position follows, and when the stock lands you restart from a worse place than you left. The damage is recoverable. It is just slow, and it is entirely avoidable with about ten minutes a week. Time: 10 minutes weekly, plus whatever the reorder takes. Watch the walkthrough Prefer to watch? Eighty-three seconds on the alert that arrives in time, and slowing yourself down. Step 1 — Let the alerts find it Insights & Alerts. Alert priorities and the category split Three inventory rules fire before you run out, and they are the only alerts in the product that measure against an absolute rather than against last week: | Alert | Fires when | Read it as | |---|---|---| | Stockout Forecast | projected to stock out within 14 days at current velocity | the early warning | | Low Stock Alert | weeks of cover below 3 weeks | act now | | Critical Stock Alert | weeks of cover below 1 week | too late to reorder normally | Stockout Forecast is the one that saves you, because it is the only one that arrives while a reorder can still land. By the time Critical Stock fires you are managing a stock-out, not preventing one. Open Alert History and check whether the same ASIN has fired before. An ASIN that goes low every few weeks is describing a reorder point that is set too tight, not a run of bad luck. See Alert History and Priority. Step 2 — Check the cover yourself Inventory. Days of cover, inbound and velocity per product Alerts use fixed thresholds. Your lead times are not fixed, so do the arithmetic that actually matters: days of cover must exceed supplier lead time + inbound transit + Amazon receiving Amazon receiving is the part people forget. Stock that has arrived at the warehouse but is not yet checked in cannot be sold, and on a busy week that can be several days on its own. Sort by days of cover, ascending. Anything below your true replenishment time is already late, whatever the alert says. Check inbound quantity beside it. Low cover with a large inbound is fine. Low cover with nothing inbound is the actual problem. Step 3 — Protect the rank while you are short This is the step almost nobody takes, and it is where the lasting damage is either done or avoided. If you are going to be tight for a week or two, slow your own sales down deliberately rather than selling out and going to zero: Reduce bids rather than pausing campaigns. A paused campaign stops accumulating history; a reduced bid keeps you present at lower volume. When stock lands you resume from a slowed campaign, not a cold one. Raise the price slightly. Unpopular, and effective. Fewer units at a higher margin stretches the same stock across the gap and keeps some velocity, which is what BSR is measuring. Never let it hit zero if you can help it. A listing that goes out of stock loses its position outright, and the Buy Box with it. Trickling out at a slower rate is materially better than a hard stop. Step 4 — Watch what it costs you Rank Tracker → the product → BSR Tracker. Best Seller Rank over time, with the axis inverted BSR reacts to sales velocity within days, so it is the fastest read on how much damage a shortage is doing. Remember the chart's axis is inverted — a line going up is good — and the arrows track the outcome, so green means the rank improved. Watch the 7-day and 30-day columns on the tracked products table. A cliff with no marketing change behind it is very often availability, and this is where you see it first. Then check the Rank Heatmap for the same product. Organic position follows BSR with a lag, so a rank slide that starts a week after a shortage is the shortage arriving. Step 5 — Come back properly Recovery is not automatic. When stock lands: Restore bids gradually, not in one jump. A sudden return to full spend on a listing with degraded rank buys expensive clicks at a worse conversion rate. Expect BSR to lag. It is computed over a rolling window, so it improves after sales do, not with them. Watch Amazon's Choice. The badge is sensitive to availability and does not always come straight back. If the count fell, that is the last thing to recover. See Tracking the Amazon's Choice Badge. What good looks like Stockout Forecast fires and you have already ordered. That is the whole goal. The alert should be a confirmation, not news. Over a quarter, the number of Critical Stock alerts should fall while Low Stock alerts stay roughly constant — that pattern means you are catching things earlier, not eliminating variability. What not to do Do not use the 3-week threshold as your reorder point. It is a general default. Yours depends on your lead times. Do not pause advertising outright unless you are genuinely at zero. See step 3. Do not let a rule pause it for you. Rules do not know your stock position, and a pause rule that fires during a shortage will do exactly the wrong thing. See Guardrails to Set Before You Automate Anything. Do not judge recovery on sales alone. Sales come back before rank does; rank is what determines whether they stay. Related - Alert Categories, Priorities and What Actually Triggers Them — the three inventory rules. - Inventory: On Hand, Inbound, Days of Cover and Velocity — step 2. - Best Seller Rank Tracking — step 4. - Protect the products that bring you new customers — which ASINs matter most here. - What a Relaunch Is, and When to Run One — why 98% in stock is a precondition for launching. - Monitoring Inventory Against Sales — the weekly routine behind the alert. - When You Are Going Out of Stock — what to do when it is already unavoidable.

Diagnose an ACoS Spike in Ten Minutes

ACoS is a ratio, so it moves when either half moves. Spend up and sales down look identical on the headline number and need opposite responses — which is why the first step is never "look at the campaigns". This is a fixed order. It takes ten minutes and it ends with a cause, not a suspicion. Watch the walkthrough Prefer to watch? Eighty-six seconds on splitting the ratio before you open a single screen. Step 0 — Split the ratio Before opening anything, answer one question: did spend rise, or did sales fall? Any screen with the Values / % change toggle will tell you. Set the range to the spike, compare against the previous period, and look at spend and ad sales separately. | What moved | Where to go | |---|---| | Spend rose, sales roughly flat | step 1 — a bidding or budget change | | Sales fell, spend roughly flat | step 3 — a conversion or availability problem | | Both moved | do step 1 first, then step 3 | Skipping this step is how people spend an afternoon auditing keywords when the actual cause was a listing that went out of stock. Step 1 — Check what you changed Automations → the rules list. Scroll right to Last Run and Changed. A rule that touched several hundred entities the night before the spike is your first suspect, and it takes fifteen seconds to check. Open its Run History, find the run covering that date, and View it. Every changed entity is listed with the branch that fired and the value before and after. If bids across a group went up 15% that morning, you have your answer without leaving the screen. See Scheduling: Last Run, Next Run and Run History. Also check the Change Log for anything a person did. Between the two, most spend-side spikes are explained inside two minutes. Step 2 — Find where the spend went Amazon Ads → Campaigns. Campaign performance with the comparison toggle Set vs previous and switch to % change. Sort by spend change, descending. You are looking for concentration. Almost always, one or two campaigns account for most of the increase. If the rise is spread evenly across everything, that is a market-level change — competitors bidding up, or a seasonal shift — and it is a different conversation. Once you have the campaign, go one level down to its ad groups, then to keywords and targets. Keep asking the same question: is this concentrated or spread? Step 3 — If sales fell instead Spend-side causes are easy. Sales-side causes are the ones that matter more, and they are rarely in the advertising at all. Check availability first. An out-of-stock or suppressed listing stops converting while the ads keep spending. This is the single most common cause of a sales-side ACoS spike. See Catch a stockout before it costs you rank. Then check conversion. In Search Query Performance, look at your click share against your purchase share on your main terms. Clicks holding while purchases fall means people are arriving and leaving — a price change, a review slide, a competitor promotion. See Search Query Cart Analysis. Then check rank. A slide in organic position means you are paying for traffic you used to get free. Rank Tracker dates it. Step 4 — Look at the query mix Amazon Ads → Search Terms, aggregated. Search terms aggregated across the account A spike with no obvious campaign cause is often a mix change: broad or auto targeting started matching new queries, and those queries convert worse than the ones you were getting. Sort by spend change. New high-spend terms that did not exist last period are the tell. Negative-match the ones that clearly do not fit, and check whether a match-type change or a bid rise widened the net. Step 5 — Confirm before you act Two checks that stop you fixing the wrong thing: Attribution lag. Recent days under-report sales — up to 7 days for Sponsored Products, 14 for the other formats. A "spike" in the last two days is very often incomplete data. Re-check with the freshest days excluded before touching anything. See Why SP, SB and SD Numbers Don't Add Up. Is it actually a spike? Compare to the same period last month and, if you have it, last year. Some of what looks like a spike is a weekday pattern or a seasonal one. What good looks like You finish with a sentence, not a feeling: "spend rose 40% because the bid rule raised bids on the generic group on Tuesday, and those terms convert below break-even." If you cannot write that sentence, you have not finished — and changing bids before you can is guessing with money. What not to do Do not cut budgets first. It lowers spend and ACoS while lowering sales too. It hides the cause rather than fixing it. Do not blame the rule without checking it. Open the run. The rule may have done exactly what you asked; the problem may be the threshold you set. Do not react to one day. Daily ACoS is noisy on any account below serious volume. Do not fix two things at once. You will not learn which one worked. Related - Understanding ROAS and Key Amazon Advertising Metrics — what ACoS is measuring. - Scheduling: Last Run, Next Run and Run History — step 1. - Values vs % Change: The Comparison Toggle — step 0. - Cut wasted ad spend in an afternoon — what to do once you have found it. - The Bleeder Reports — when the cause is a target bleeding rather than a market shift. - Placement Bid Optimisation — when the cause is spend sitting on the wrong placement.

Triage a Rank Drop Across Your Top Keywords

A rank drop is rarely one thing going wrong across the board. It is usually one product, or one cluster of related terms, and the rest of the noise is people looking in the wrong place. Triage in this order and you will have the shape of it in about fifteen minutes. Time: 15 minutes. Do it whenever visibility moves, and monthly regardless. Watch the walkthrough Prefer to watch? Eighty-eight seconds on narrowing it down, and the causality check people skip. Step 1 — Is it everything, or one thing? Rank Tracker → the portfolio. Projects, with visibility, average position and Top 10 Look across your projects and read three columns together: Visibility — higher is better. Avg Position — lower is better. Top 10 — a count, and the one that cannot be dragged by outliers. Two of those move in opposite directions, which is the most common misreading of this screen. A project whose Avg Position went from 14 to 20 got worse. If one project moved and the others did not, you have narrowed it already. If everything moved together, look for a cause that touches everything — an account-level issue, a seasonal shift in what people search, or an Amazon change. Remember visibility is a mean of means: a project's figure is the average of its products'. One product going to zero drags the project down without anything else changing. Open the project before believing its headline. Step 2 — Which product? Open the project. The per-product table gives Visibility, Avg Position, Top 10, Current BSR and Amazon's Choice per ASIN. Find the row that moved away from its siblings. If four products held and one fell, this is a product problem, and steps 3 to 5 are about that product. If all of them fell together, it is a brand or category problem — skip to step 6. A row at 0.00% visibility is not a rank drop. That product is returning nothing at all, which is a different failure: a suppressed listing, a duplicate ASIN, or a product that never ranked for those terms. See Rank Data Shows Blank. Step 3 — Which keywords, and when? Open the product and take the Rank Heatmap. Daily rank per keyword, with the rank distribution This is the screen that dates the problem. Every tracked keyword gets a row and every day gets a column, so you can read right to left and see exactly when each term started to slide. Two shapes, two different causes: All keywords fell on the same day. Something happened to the listing — it went out of stock, the price moved, the Buy Box was lost, or the content changed. Listing-level causes hit every term at once. Some keywords fell and others held. That is competitive. Somebody moved on those specific terms, or the terms themselves changed in what they return. Check Rank Distribution too. An average that got worse while the 1–3 and 4–10 bands held means you lost ground in the tail, which matters far less than the headline suggests. Step 4 — Did sales cause it, or did it cause sales? BSR Tracker, same product. This is the causality check, and it is worth the thirty seconds. Best Seller Rank moves with sales velocity; organic rank follows it with a lag. If BSR fell first and organic rank followed, the cause is upstream of search — you sold less, for a reason that has nothing to do with ranking. Availability, price, a review slide, a competitor promotion. If organic rank fell first and BSR followed, you lost visibility and the sales decline is the consequence. Now it is a search problem. Getting this backwards is how teams spend a month optimising listings when the real cause was a two-week stock-out. Step 5 — Is the traffic still arriving? Search Query Performance. SQP analysis, week by week For the affected terms, compare your impression share, click share and purchase share week over week. SQP reports weekly cohorts, so use week-on-week rather than an arbitrary range. See SQP Is Weekly: Reading Week Cohorts Correctly. | Pattern | Reading | |---|---| | Impression share fell | you lost position — a genuine rank problem | | Impression share held, click share fell | you are still shown and no longer chosen — image, title, price, rating | | Click share held, purchase share fell | they arrive and do not buy — a detail-page problem | Only the first row is actually about rank. The other two are listing problems wearing a rank problem's clothes. Step 6 — Check who took it If you track competitors on the same keywords, the Compare tab settles it: your rank and theirs on the same terms, on the same day. A competitor climbing into the positions you lost is a competitive answer, and it changes what you do next — you are now deciding whether to fight for that term, not why you slipped. Also check Amazon's Choice. A badge count that fell on those terms usually moved to somebody, and losing it costs conversion on top of position. See Tracking the Amazon's Choice Badge. What good looks like You end with a scoped statement: "one product, six of its twenty-five keywords, starting on the 14th, and BSR fell first." That points straight at the fortnight when something changed in supply or price. What not to do Do not react to a single day. Amazon results are personalised and volatile. The heatmap exists because the trend is the signal — use the median column, not yesterday's cell. Do not optimise the listing before step 4. If BSR fell first, the listing was never the problem. Do not raise bids to compensate. Paid position does not restore organic position; it just costs more to be in the same place. See Organic vs Sponsored Rank. Do not forget there is no backfill. Rank history starts the day tracking was enabled, so a project created last week cannot tell you what happened last month. Related - Reading Visibility, Average Position and Top 10 — step 1. - Best Seller Rank Tracking — step 4, the causality check. - Search Query Cart Analysis: Why a Query Doesn't Convert — when step 5 points at the page. - Tracking Competitor ASINs Alongside Your Own — step 6. - Profit Campaigns and Ranking Campaigns — the campaign type built to move rank back. - What a Relaunch Is, and When to Run One — the deliberate way to move rank back.

Decide Whether Your Upper-Funnel DSP Is Worth Keeping

DSP is the line item most often defended on faith and cut on impulse. It rarely looks good on ACoS, the case for it is always "brand building", and nobody can produce a number. There is a number. It takes about an hour to build, and it either justifies the spend or it does not. Time: about an hour. Do it quarterly, or before any renewal. Watch the walkthrough Prefer to watch? Eighty-one seconds on what you actually bought, and the harder comparison. Step 0 — Ask the right question Not "is DSP profitable?" — measured on last-click it almost never is, and that tells you nothing. The question is: is DSP doing something the rest of the account cannot do? There are only three honest candidates: 1. Bringing customers who are new to the brand. 2. Creating demand that converts later somewhere cheaper. 3. Re-reaching people at a cost below what Sponsored Products pays for the same outcome. If it is doing none of the three, it is an expensive way to buy sales you already had. Step 1 — Find out what you actually bought Amazon DSP → Campaigns. DSP campaigns by funnel stage Read the campaign names and sort them into two piles: prospecting (audiences who have not met you) and remarketing / retargeting (audiences built from people who have). This is the single most important step, and it is usually where the argument ends. A DSP buy that is 80% remarketing is not upper-funnel advertising. It is a re-engagement channel wearing upper-funnel language, and it should be judged as one. Write down the split. Everything below depends on it. Step 2 — Check whether it brings anyone new AMC Cloud → NTB Purchases. New-to-brand share by ad product type Compare DSP's new-to-brand share against your other formats. If DSP is at the bottom, that is consistent with a remarketing-weighted buy and it is not automatically damning — but it does mean claim (1) is dead. You are not buying new customers. If DSP is at the top, claim (1) is alive and you have your justification. Confirm it on NTB Customers, which counts people rather than orders and will tell you whether the volume is meaningful or just the rate. The trap here is a good rate on a tiny volume. A format that is 70% new-to-brand and delivers 0.1% of your new customers is not acquiring anybody. Step 3 — Check where it sits in the journey AMC Cloud → Path to Purchase. Conversion paths, ranked, with the NTB columns This is the step that settles it, because it shows order, not just presence. Compare paths that begin with DSP against paths that begin with Sponsored Products, and read the NTB purchases column rather than sales: DSP-first paths with a high NTB rate — DSP is genuinely finding new people and handing them to search. That is the upper funnel working exactly as advertised, and it is worth paying for. DSP-first paths with a low NTB rate — DSP appears first because it re-reached somebody you already had. It is not opening the journey; it is intercepting one already in progress. Then the harder comparison: does adding DSP to a path improve it? Compare a path with DSP against the same path without. If the single-touch Sponsored Products path produces more purchases and more new customers than the DSP-assisted one, the DSP exposure is a cost added to a sale that was already coming. Step 4 — Check whether it created demand AMC Cloud → Brand Keywords. If DSP is building awareness, it should produce brand searches — people who saw the ad and later typed your name into Amazon. That is claim (2), and this is the only place it is measurable. Look at Campaigns Driving Searches. If your DSP campaigns appear and their branded search volume moves with DSP spend, you have evidence of created demand that will convert later through cheaper channels. If DSP spend rises and branded searches do not move at all, claim (2) is dead too. Step 5 — Price the remarketing honestly If DSP turns out to be mostly remarketing, that is claim (3), and it is still a legitimate case — but only if the price is right. The comparison is: what does DSP pay to re-reach a past visitor, versus what Sponsored Products pays to reach the same person again through search? If DSP is cheaper per outcome, keep it and stop calling it upper-funnel. If it is dearer, you are paying a premium for display inventory to do a job search already does. Making the call | What you found | Decision | |---|---| | High NTB, DSP-first paths acquire | Keep and fund. This is what it is for. | | Low NTB, but branded searches track spend | Keep, smaller. It is creating demand, not customers. | | Low NTB, no search lift, cheap remarketing | Keep as remarketing. Rename it in your own reporting. | | Low NTB, no search lift, dear remarketing | Cut. It is buying sales you already had. | | Prospecting exists but delivers nothing | Fix before cutting. The audiences are wrong, not the channel. | That last row matters. If your prospecting campaigns are funded and still deliver no new customers, the problem is audience selection, and cutting the whole channel throws away the only part that could have worked. What not to do Do not judge it on ACoS. Prospecting always looks worse than remarketing — that is the definition of the two strategies, not a finding. Do not cut prospecting to improve the average. It feeds the remarketing pools that look so good. Cutting it improves this month and starves next quarter. Do not decide on one month. AMC reports monthly cohorts. Look at three. Do not forget DSP is read-only here. TrackIQ reports it; changes happen in Amazon's DSP console. Related - DSP Campaigns — step 1, reading the funnel. - NTB Purchases: New Business by Campaign — step 2. - Path to Purchase: Which Ad Sequences Actually Convert — step 3, the decisive one. - Brand Keywords: Proving Upper-Funnel Spend Works — step 4.

Harvest Winning Search Terms into Exact-Match Campaigns

Broad and auto campaigns are discovery engines. They find the queries you did not think of, and then keep paying broad-match prices for them forever. Harvesting is moving a proven query out of discovery and into a keyword you control. It is the highest-return hour in Amazon advertising, and it is almost entirely mechanical. Time: about an hour. Do it weekly, or fortnightly at minimum. Watch the walkthrough Prefer to watch? Eighty-six seconds on aggregation, exact match, and the step people skip. Step 1 — Find the queries that converted Amazon Ads → Search Terms, with Aggregated Search Terms switched on. Search terms, aggregated across the account Aggregation is not optional here. In the default view a single customer search appears once per keyword per ad group, so a query that produced twelve orders across three ad groups looks like three unremarkable rows and never surfaces. Filter to terms from broad, phrase and auto targeting, then sort by orders, descending. What qualifies: enough orders to be more than luck, at an ACoS you would accept. Two orders is a coincidence; five or more at a workable ACoS is a pattern. Set the bar where you would act by hand. Step 2 — Cross-check against SQP Search Query Performance → Recommended Actions, filtered to Add as keyword. Recommended actions, with the add-as-keyword filter This finds a different population and is worth the two minutes: queries where you have impression and click share but no keyword bidding on them. Some are terms you rank on organically and have never advertised — and those are the best harvests of all, because demand is already proven and you are only choosing to defend it. The Why column tells you which is which. A row saying you are shown and clicked but not bidding is a straightforward add. Step 3 — Check you are not already bidding on it Amazon Ads → Keywords. Existing keywords and match types Search each candidate before you add it. Duplicating a keyword you already run means two of your own keywords competing for the same search, which raises your cost and teaches you nothing. If the term already exists as broad and you are adding exact, that is fine and intended — that is the harvest. If it already exists as exact, you are done; there is nothing to move. Step 4 — Add it as exact, in the right place Add the winning query as an exact-match keyword, and put it where you can see it: A dedicated harvest campaign or ad group is the usual structure. Keeping harvested exacts separate from your discovery campaigns means you can budget and bid them differently, and you can tell at a glance how much of your performance comes from proven terms. Set the opening bid from what you already paid. You know the CPC the term achieved in broad; start near it rather than guessing. Exact match usually costs a little more per click and converts better, so a small premium is normal. Keep the product the same. Harvest the term into an ad group advertising the ASIN that actually converted for it. A term that worked for one product will not necessarily work for its sibling. Step 5 — Negate it in the source This is the step people skip, and skipping it wastes the entire exercise. Add the harvested term as a negative exact in the broad, phrase or auto campaign it came from. Without this you now have two campaigns bidding on the same search: your new exact keyword and the original broad one. You compete with yourself, your costs rise, and the exact keyword never accumulates the clean history that made harvesting worthwhile. Negative exact, not phrase or broad — you want the broad campaign to keep discovering variants, just not this specific one. Step 6 — Watch it for two weeks A harvested keyword needs a fortnight before you judge it, and the failure modes are recognisable: | What happened | Reading | |---|---| | Converts as well or better, similar CPC | working — leave it | | Converts as well, CPC much higher | you bid too aggressively; ease it down | | Volume collapsed | the negative is too broad, or exact match is genuinely narrower than the broad variants it was catching | | Stopped converting entirely | the broad match was converting on neighbouring queries, not this one | That last row is the interesting failure. It means the term you harvested was never the one doing the work — the broad keyword was matching something adjacent. Go back to the aggregated search terms and look at what else that keyword was catching. What good looks like Over a few months, a growing share of your spend sits on exact-match terms with known performance, and your broad and auto campaigns shrink toward what they are actually for: finding new queries, not serving old ones. A discovery campaign whose spend never falls is a discovery campaign that is not being harvested. What not to do Do not harvest on one order. You will fill the account with exact keywords that never fire again. Do not harvest competitor-brand queries into exact match without deciding that deliberately. They are expensive and they convert differently. See Generic, Branded and Competitor Terms. Do not forget the negative. Step 5 is the one that pays for the exercise. Do not do this and a bid change in the same week. You will not know which moved the numbers. Coming soon A Harvest Keywords rule type is visible in the rule chooser with a Coming Soon badge — "promote converting search terms into exact keywords automatically" — along with a Negative Targeting rule that would handle step 5. Until they ship, this stays a manual weekly job. See Rule Types Coming Soon. Related - Search Terms and Aggregated Search Terms — step 1, and why aggregation matters. - Keywords and Match Types — step 3 and 4. - Find converting search queries you aren't bidding on — the SQP-first version of the same hunt. - Cut Wasted Ad Spend in an Afternoon — the other half of search-term hygiene. - Score Your PPC Performance — where harvesting sits among the eight reports. - Search Term Harvesting: Criteria and Bids — the criteria, the fifteen-term cap and the starting bids.

Find Your Next Bundle from What Customers Already Buy Together

You already know which products sell. You probably do not know which ones get bought together — and that pairing is the cheapest new product you will ever launch, because it requires no new product. The data is in AMC. The discipline is in not over-reading it. Time: about 45 minutes. Do it quarterly. Watch the walkthrough Prefer to watch? Eighty seconds on reading the base before the ratio, and testing it cheaply. Step 1 — Check whether you have enough data AMC Cloud → ASIN Purchase Upsell. Lead ASINs, overlap ASINs, overlap users and the average ratio Before looking at a single pair, read the two tiles that decide whether this exercise is worth doing: Overlap Users — how many customers bought more than one of your ASINs at all. Avg Overlap Ratio — on average, what share of a lead product's buyers also bought the companion. Cross-purchase is rare. An average ratio under 1% and overlap users in the low hundreds is entirely normal, and it means every pair you are about to look at rests on a handful of people. That does not make the exercise pointless — it makes it a hypothesis generator rather than a source of findings. If overlap users is in single or low double digits, stop. There is nothing here yet. Come back after a quarter with more volume. Step 2 — Rank by volume, not by ratio Use Overlap User Count by Lead ASIN first. It ranks lead products by how many customers bought them and at least one other ASIN. Then read the ratios against that. The four-way split: | Overlap users | Ratio | Reading | |---|---|---| | high | high | a real pair — act on it | | high | low | popular lead, weak companion pull | | low | high | noise — a big percentage of very few people | | low | low | nothing here | The bottom-left row is the trap, because a 40% overlap ratio catches the eye and looks like a discovery. On three customers it is not. Step 3 — Read the pairs properly The pair table, with overlap users beside lead distinct users The table gives both products with Overlap Users, Lead Distinct Users, Overlap Ratio and Lead Purchases. Seeing the denominator beside the ratio is the whole point. Overlap ratio = overlap users ÷ lead distinct users, so three overlap users against 165 lead buyers is 1.82% — a number you should treat as a whisper, not a signal. Sort your shortlist by overlap users and take the top three or four pairs. That is your candidate list. Step 4 — Sort the candidates into three kinds Not every pair means the same thing, and they lead to different actions: The obvious pair. A product and its refill, or a set and its accessory. You did not need AMC to find this, but it confirms the demand is real. Bundle it. The complementary pair. Two products used together that you had not connected — different categories, same occasion. This is the valuable one. It is a genuine discovery about how customers use what you sell. The substitute pair. Two variants of the same thing, bought by someone who tried one and switched, or bought both to compare. Do not bundle these. Nobody wants two of the same product in a box. This is a signal about your variant strategy, not a bundle. Telling them apart takes product knowledge, not data. Look at the two listings and ask whether a customer would want both at the same time. Step 5 — Remarket before you bundle Bundling costs inventory, a new listing, and photography. Remarketing costs a campaign. So test the hypothesis the cheap way first. The audience is customers who bought the lead product and not the overlap. That is a DSP remarketing campaign with an unusually specific brief: they own one thing, show them the other. If that campaign converts, you have validated the pair with real money and you can bundle with confidence. If it does not, you have learned the same thing for the price of a test. See DSP Campaigns. Step 6 — Then build the bundle If the remarketing worked, a listed bundle removes a step for the customer and usually lifts average order value. Price it below the sum. The point is convenience plus a small saving; if it costs the same as buying both, it is only a convenience. Watch which one is the gateway. If one of the pair is a product that recruits new customers, bundling can reduce acquisition — you have raised the entry price into your brand. Check it against NTB ASINs Analysis before you make the bundle your headline offer. Keep both singles listed. A bundle should be an addition, not a replacement. What good looks like One or two validated pairs a year is a good return from this. It is not a high-frequency exercise, and the value is concentrated in the complementary pairs you would never have guessed. What not to do Do not treat a ratio as a finding without its denominator. Step 3 exists for this reason. Do not assume order. The dashboard pairs a lead and an overlap ASIN; it does not tell you which was bought first. Do not read causation into it. Both purchases may follow the same need rather than one leading to the other. Do not forget it is ad-attributed. Purely organic cross-purchase is not represented, so your real overlap is higher than the dashboard shows. Small numbers move violently. A pair going from 2 to 4 customers is a 100% increase and means nothing. Related - ASIN Purchase Upsell: Finding Bundle and Remarketing Pairs — the dashboard in full. - NTB ASINs Analysis: Finding Your Gateway Products — check before bundling a recruiter. - DSP Campaigns — where the cheap test gets built. - Sales Performance: Product-Level Reporting — what each product does on its own. - Six Ways to Find Product Candidates — five more ways to generate candidates. - Before You Add a Product — whether a new product is the right move at all.

Set a Target ACoS Across a Whole Portfolio

One ACoS target across a whole portfolio is the most common bad idea in Amazon advertising. Products have different margins, different lifecycles and different jobs, and a single number holds all of them to the standard of the average one. This is how to set targets that mean something — and then let rules hold them without you. Time: half a day to set up, then it maintains itself. Watch the walkthrough Prefer to watch? Eighty-six seconds on grouping by job, and letting the formula size the correction. Step 1 — Work out your break-even, once Before any target, you need the ceiling: Break-even ACoS = your contribution margin before advertising If a product sells for $30 with $19 of product, fee and shipping cost, you have $11 — roughly 37%. At 37% ACoS that product makes nothing. Above it you are buying revenue with your own money. Everything else in this playbook is expressed as a position relative to that number, per product. Do this per product family, not per SKU, unless your margins genuinely differ by variant. Step 2 — Group by the job, not the product Automations → Rule Groups. Rule groups, with the campaign count on each A portfolio target only works if the portfolio is divided by what you want each part to do. Four groups cover most accounts: | Group | Job | Target, relative to break-even | |---|---|---| | Defend | branded terms, hold what is yours | far below — these convert cheaply | | Harvest | proven exact terms, make money | comfortably below | | Grow | generic terms, win new customers | at or slightly below | | Launch | new products, buy rank and reviews | above — deliberately | Launch above break-even is not a mistake. You are buying velocity and reviews for a period you have decided in advance, and you should decide the end date at the same time as the target. Gateway products belong in their own group too, with a deliberately looser target — you are buying a customer, not a transaction. See Protect the Products That Bring You New Customers. The campaign count on each rule group is your blast radius. Check it before attaching a rule. Step 3 — Set the target where the rules can read it Automations → the rule → Cadence & Guardrails. Target ACOS, the change cap and bid bounds Target ACOS is a field on the rule, and the Set bid to hit Target ACOS action reads it: new bid = current bid × (Target ACOS ÷ actual ACOS) This is why goal-seeking beats a fixed percentage nudge across a portfolio. A target at 100% ACoS against a 50% goal halves; one at 55% barely moves. The size of the correction matches the size of the problem, which is exactly what you want when one rule is managing hundreds of targets at once. The panel's own guidance is worth heeding: stay under your pre-ad margin for profit; go above it only while launching. Step 4 — Bound what the rule may do Same panel, Bounds & Change Cap: Max change per run — 10–15% for a daily rule. The formula can want a large correction; this stops it taking one in a single step. Min bid and Max bid — absolute floors and ceilings. These are the backstop against a bad calculation on thin data. A target with two clicks and a freak conversion will produce an enormous suggested bid, and the cap is what stops it being applied. Look-back 14 days, ignore the freshest day. Fourteen days holds two complete weekly cycles; ignoring the newest day keeps the rule from reacting to sales Amazon has not attributed yet. See Look-back Windows and Adjustment Frequency. Step 5 — Build the rule as a branch list One rule per group, and the branch order does the work. A shape that holds up: | # | When | Then | |---|---|---| | IF 1 | Clicks > 50 and Orders = 0 | Pause targeting | | ELSE IF 2 | Spend ≥ $50 and Orders = 0 | Pause targeting | | ELSE IF 3 | Orders > 1 | Set bid to hit Target ACoS | | ELSE IF 4 | Clicks < 50 or Orders ≤ 1 | Increase bid by 15% | Read it as a decision list: kill what has proved it does not work, steer what has proved it does, and give everything else room because it has not been tested yet. Branch order is the logic — branches are walked in order and only the first match fires. The catch-all belongs last, or it swallows the entities the specific branches were meant to catch. See Building a Rule from Scratch. Step 6 — Dry-run every group before enabling Dry-run preview computes every change and applies none, and it shows which branch claimed each target. Two things to check per group: Is the changed count plausible? Half of everything, every day, means the thresholds are too loose. Zero means the rule is decoration. Is the branch distribution what you expected? If every row says branch 4, branches 1–3 are unreachable and the order is wrong. See Dry-Run Preview. Step 7 — Review monthly, not daily Once running, the review is short: Changed count per rule. A rule changing nothing is not protecting you. Actual ACoS per group against its target. Consistently above means the target is unreachable at your conversion rate, not that the rule is failing. Launch groups against their end date. A launch target that nobody switched off is just an expensive group. What not to do Do not set one target for everything. That is the problem this playbook exists to fix. Do not set a target above break-even outside a launch, and write the end date down when you set it. Do not run daily on a 30-day window with a large cap. Every run re-reacts to the same month and the changes compound on evidence that has not moved. Do not let a pause rule touch gateway products or launches. Put them in groups your pause rules do not run on. Do not change target and cadence in the same week. Move one, watch, then the other. Related - Rule Groups: Applying One Rule Across Many Campaigns — step 2. - Building a Rule from Scratch — steps 3 to 5. - Guardrails to Set Before You Automate Anything — step 4 in full. - Understanding ROAS and Key Amazon Advertising Metrics — the break-even maths. - Spot Cannibalisation Between Sponsored Brands and Sponsored Products — before you blame one ad type for another.

Prove Your Display Campaigns Created Demand

Display advertising is judged by a system built to measure search. Somebody sees your ad, does not click, and buys a fortnight later through a branded search — and the display campaign gets nothing. You cannot fix the attribution. You can, however, produce evidence that the demand exists and that display created it. Time: about 45 minutes. Do it before any budget conversation. Watch the walkthrough Prefer to watch? Eighty-five seconds on three claims, in descending order of strength. What "created demand" actually means Three claims, in descending order of how convincing they are: 1. Display brought people who had never bought from the brand. 2. Display caused people to go and search for the brand. 3. Display appears early in journeys that end in a purchase. Claim 1 is the strongest and the hardest to earn. Claim 2 is the one most people never think to measure. Claim 3 is the weakest, because presence is not influence — but it is where the sequence is visible. Take them in order. Step 1 — Did it bring new customers? AMC Cloud → NTB Customers. New-to-brand share and volume by ad product type This dashboard counts people, which is what you want when the claim is about acquisition. Read both charts, because they disagree and the disagreement is the argument: The bars give each format's NTB rate — what proportion of its buyers were new. The doughnut gives each format's share of all new customers — the volume. A strong result is a decent rate and a meaningful share. A rate of 70% on 0.1% of your new customers is not evidence of acquisition; it is a rounding error with a good percentage. Check the legend before reading the bars. On this dashboard NTB is the lighter segment; on NTB Purchases it is the darker one. They do not agree with each other. If display shows a real share here, stop — you have your case, and it is the best one available. Step 2 — Did it cause brand searches? AMC Cloud → Brand Keywords. Branded searches generated after ad exposure This is the measurement most accounts never run, and it is the cleanest evidence of created demand that Amazon makes available. Brand keywords are searches customers ran on Amazon for your brand after being exposed to your ads. Somebody who saw a display ad and later typed your brand name has changed their behaviour because of it. They may then convert through Sponsored Brands, or organically, and neither of those will credit display — but the search itself is the fingerprint. Look at Campaigns Driving Searches. If your display campaigns appear there, they are generating brand-search intent. Then look at the trend. Compare branded search volume across months against display spend across the same months. Moving together is your evidence. Display spend rising while branded searches stay flat is the null result, and it is worth knowing honestly. Read the keyword list too. Brand + a specific product means display built awareness of a thing, which converts far better than brand alone. A product appearing there that you did not advertise is display doing discovery work. Step 3 — Where does it sit in the journey? AMC Cloud → Path to Purchase. Conversion paths with their NTB columns Now the sequence. Find the paths that begin with display and read the NTB purchases column, not sales. DSP-first paths with a high NTB rate — display opened the journey and handed a genuinely new customer to search. That is created demand, visible in order. DSP-first paths with a low NTB rate — display went first because it re-reached somebody the brand already had. It intercepted a journey rather than starting one. This is where the argument is usually won or lost, and it is worth being honest about the outcome. A remarketing-weighted display buy will show low NTB here, and that is a description of what was purchased rather than a failure of the channel. Step 4 — Assemble the case Three numbers, in one place, compared against the same period last quarter: | Evidence | Where | What convinces | |---|---|---| | New customers from display | NTB Customers | a share of acquisition, not just a rate | | Branded searches from display campaigns | Brand Keywords | volume that moves with display spend | | Display-first paths | Path to Purchase | a high NTB rate on those paths | If all three point the same way, the case is made. If they disagree, trust them in the order given — acquisition beats search lift beats sequence. What you cannot prove here Be straight about the limits, because overclaiming is how a good case gets dismissed: None of this is incrementality. It shows the customer was new to the brand and that a search followed exposure. It does not prove the purchase would not have happened anyway. A true incrementality test needs a holdout, which is outside TrackIQ. Exposure is not influence. Somebody who scrolled past a display ad appears in the path. Organic is invisible. Paths contain only ad exposures, so a customer who found you through organic search and later saw a display ad looks display-influenced. Monthly cohorts. Every AMC dashboard reports one calendar month. Use three before drawing a trend. If the evidence is thin A null result is a finding, and it usually points at what you bought rather than at display as a channel. Open DSP Campaigns and check the prospecting-versus-remarketing split. A buy that is overwhelmingly remarketing was never going to create demand — it was purchased to re-engage. In that case the honest move is to rename it in your own reporting and judge it on remarketing efficiency instead. See Decide whether your upper-funnel DSP is worth keeping. Related - Brand Keywords: Proving Upper-Funnel Spend Works — step 2, the key measurement. - NTB Customers: New vs Repeat, Counted as People — step 1. - Path to Purchase: Which Ad Sequences Actually Convert — step 3. - DSP Campaigns — what your display spend actually bought. - Create a Sponsored Display Campaign — the targeting families this method judges.

Get Your Daily Number in Slack Without Opening TrackIQ

Most people do not need a dashboard every morning. They need one number, and they need it where they already are. Aria in Slack answers questions in a DM or a channel with the same reasoning as the web app — so the daily check-in stops being a login and becomes a message. Time: five minutes to set up. Saves a few minutes every day. Watch the walkthrough Prefer to watch? Seventy-four seconds on setup, brand pins, and asking one question properly. Step 1 — Connect the workspace Settings → Aria in Slack. Connect in three steps One click — no admin install required. Works with any Slack workspace you already belong to. That is the part worth knowing before you ask anyone for permission. This is not a workspace-wide app install that needs sign-off; it connects for you, in workspaces you are already in. Step 2 — Pin the brand Aria asks which brand to work on with your first message. If you manage several, you get a picker, and the choice sticks until you switch. What you get: DMs, channels, per-channel brand pins The detail that makes this useful for teams: each channel keeps its own brand pin. Brand A in #marketing, Brand B in #ops, and neither team has to remember which account they are asking about. Type switch to repick at any time. Step 3 — Put it where the question gets asked Two placements, and they serve different purposes: A DM for your own morning check. Private, no noise for anybody else. A channel for the number the team argues about. /invite @TrackIQ to the channel, then @TrackIQ it in a message. The answer lands in the thread where the discussion is already happening, which is worth more than the answer itself — nobody has to screenshot a dashboard into Slack any more. Step 4 — Work out your actual daily question This is where most people go wrong: they ask for a dashboard in text form and get a wall of numbers. Ask one question with a decision behind it. Good daily questions are short and comparative: - "How did yesterday compare to the same day last week?" - "Which campaigns spent more than $100 yesterday with no sales?" - "Any ASINs below three weeks of cover?" - "What was our ACoS yesterday and what was it the day before?" Bad daily questions are open-ended — "how are we doing?" — because the answer has to guess what you meant. The same principles as the web app apply. See Asking Aria a Good Question. Step 5 — Decide what belongs in Slack and what does not Slack is for the check, not the analysis. | Ask in Slack | Open the app | |---|---| | yesterday's number | a month-over-month trend | | a threshold check | anything you need to sort and filter | | "did the thing I changed work?" | building a report to send onward | | a fast answer during a discussion | comparing more than a handful of rows | If the answer would be a table with twenty rows, you wanted a screen, not a message. Alerts are separate, and complementary Aria in Slack is pull — you ask, it answers. Alerts are push — something crosses a threshold and you are told without asking. Use both, for different things: Alerts for anything with a hard rule: stock cover, spend without sales, an ACoS spike. Those should reach you whether or not you thought to ask. See Alert Categories, Priorities and What Actually Triggers Them. Aria for the questions a rule cannot anticipate — which is most of them. The failure mode is asking Aria every morning for something an alert would have told you unprompted. What good looks like After a fortnight you should be opening TrackIQ less often and for longer. The two-minute "just checking" logins collapse into a Slack message; the sessions that remain are real analysis. If you are still logging in daily to look at the same tile, that tile is a question you have not yet asked Aria. What not to expect It is the same Aria. Identical reasoning to the web app — which also means the same limits. It cannot answer what the data does not contain. See What Aria Can and Cannot Answer. It does not change anything. Asking Aria is a read. Bid and budget changes still happen in rules or in Amazon. Answers are for the pinned brand. In a multi-brand account, a surprising answer is worth a switch check before it becomes a surprising decision. Check the reasoning on anything you will act on. Aria shows its plan and its queries; the discipline is the same as in the app. See Checking Aria's Work When an Answer Looks Wrong. Related - Chatting With Aria in Slack — the feature in full. - Connecting TrackIQ to Slack — the connection itself. - Asking Aria a Good Question — step 4. - Setting Up Smart Alerts and Real-Time Notifications — the push half.

Build a Weekly Executive Summary You Never Have to Write

Somebody spends a morning every week assembling the same numbers into the same deck, and by the time it lands it is already a week old. That job should not exist. The goal is not a report that writes itself — it is a structure that stays the same, so producing it takes minutes and reading it takes seconds. Time: an hour to set up. Ten minutes a week after that. Watch the walkthrough Prefer to watch? Eighty-one seconds on describing a report, and why boring structure is the goal. Step 1 — Decide who it is for, before anything else An executive summary that tries to serve everybody serves nobody. Pick one audience and write down the decision they make with it: | Audience | The decision | What belongs in it | |---|---|---| | Founder / GM | where to put money next | total sales, TACoS, category movement | | Ads manager | what to change this week | ACoS by group, wasted spend, what moved | | Board | is this business working | trend, margin, acquisition | If you cannot name the decision, you are producing a status update, and nobody reads those twice. Step 2 — Generate the first one Business Reports → AI Report Generator. Quick-start shapes, a description box, and the comparison toggle Start from Executive Review or Weekly Ops, then describe it properly. The instruction on the screen is the whole skill: Mention the audience, the metrics, and the decision this should support. A weak description — "weekly performance" — produces a generic deck. A useful one names all three: Weekly review for the founder. Total sales and ad sales by category, TACoS per category, the three campaigns whose spend moved most, and anything below three weeks of stock cover. The decision is where next month's budget goes. Choose Slides for people, Dashboard for screens. Slides are for sending; dashboards are for coming back to. Turn on the comparison range. A number without a comparison is not information. This is the single most common omission. Step 3 — Fix it once, properly The first generated report will be roughly right and specifically wrong. Edit it until it is the report you actually want — the right slides, the right order, nothing you would skip past. Then spend two minutes on the thing that decides whether anybody reads it: the order. Put the number that drives the decision on the first slide. Detail goes at the back, where it can be ignored without cost. Step 4 — Save it as a template Templates, with slides, widgets and linked reports The ⋮ menu on the report offers Save as Template. This is the step that converts a one-off into a routine. A template stores the structure, not the data. Next week you produce the same layout against a new range, and the reader gets the same shape they learned last week — which is most of what makes a recurring report readable. People stop reading the report and start reading the change. Watch the Linked Reports column over time. A template with several linked reports is doing real work; one sitting at zero was a layout nobody chose to reuse. Step 5 — Add the interpretation, not just the numbers The numbers are the easy half. What makes a summary worth its slot is the sentence that says what changed and why. Two ways to get it: An Insight Report alongside it. Insight Reports ask the AI to find what is interesting in the data sources you tick, rather than presenting metrics you chose. They also keep Revisions, so you can see how the story changed week to week. See How to Create an AI-Powered Insight Report. Aria, for the one-line why. Ask "what changed most vs last week and what caused it?" and paste the answer under the headline number. See Asking Aria a Good Question. Step 6 — Make the weekly job small Once the template exists, the routine is: 1. Produce the report from the template against last week's range. 2. Read it before anybody else does. 3. Add or correct the one-line interpretation. 4. Send it. Step 2 is not optional. A generated report read by its recipients before its author is how a wrong number becomes a decision. What good looks like The same slides, every week, in the same order. Boring is the goal. A reader who knows where to look reads it in ninety seconds and asks a better question than they used to. The interpretation is the only part you write. If you are still assembling numbers by hand, the template is not finished. What not to do Do not redesign it monthly. Consistency is the product. Change the structure only when the decision it supports changes. Do not include everything you can measure. A slide nobody acts on is a slide that teaches people to skim. Do not skip the comparison range. Twice as important as any individual metric. Do not send a generated report unread. See step 6. A note on delivery There is no email scheduling for reports in the product today — neither report type offers a schedule or recipient field. Producing from the template and sending it yourself is the current path, which is why step 6 is a short manual routine rather than an automation. For a genuinely hands-off daily number, Aria in Slack is the closer fit. See Get your daily number in Slack without opening TrackIQ. Related - Business Report Templates and the AI Report Generator — steps 2 to 4. - Understanding Business Reports — the reports themselves. - Weekly Business Reports and AI-Generated Insights — the built-in weekly view. - How to Create an AI-Powered Insight Report — step 5. - Build a Monthly Board Pack — the monthly version, for a different room.

Benchmark Yourself Against Your Own Category

"Is a 32% ACoS good?" has no answer. Good against what — a launch, a mature line, a clearance? The industry benchmarks you can find online are averages of businesses that are not yours. The only benchmark that means anything is your own category, last quarter. Category Insights is where you build it. Time: an hour to set up. Twenty minutes a month after. Watch the walkthrough Prefer to watch? Eighty-seven seconds on building your own benchmark, and the two patterns to catch. Step 1 — Define categories you would actually fund Analytics → Category Insights → + Add Category. Category Insights, with Total Sales beside Ad Sales These are your groupings, not Amazon's browse nodes, which means they are only as useful as the thinking behind them. Group by something you would make a budget decision about: - A product line you fund as a unit. - A price tier, where the margin maths is similar. - A use case, if that is how your customers shop. Do not group by something you would never act on separately. And do not mix a hero product with a long tail — the hero's numbers get averaged away and the category tells you nothing about either. A good test: if two products would never compete for the same budget decision, they should not share a category. Step 2 — Use the column no advertising screen has Category Insights reports Total Sales beside Ad Sales. Every Amazon Ads screen shows only ad sales — what advertising was credited with. This one shows the whole category. That difference lets you compute the thing advertising reporting cannot: TACoS = ad spend ÷ total sales ACoS asks "was the ad efficient?". TACoS asks "how much of this business is paid?", which is the question a benchmark should be built on. Step 3 — Establish the baseline Take a stable quarter — no launch, no stock-out, no promotion — and record per category: | Measure | Why it belongs in the baseline | |---|---| | TACoS | how much of the category is bought | | ACoS | efficiency of the advertised part | | Total Sales | the size of the thing | | Total Units | so you can separate price from demand | Write them down somewhere outside the product. This is now your benchmark, and it is worth more than any published industry figure because it already contains your margins, your category dynamics and your competitive set. Step 4 — Compare categories against each other, not a single target This is the habit the whole playbook exists to build. A launch category should look worse than a mature one. If they have the same target, one of them is being managed wrongly. Compare like against like: two mature categories, or this category against itself last quarter. Comparing a launch against a mature line and concluding the launch is inefficient is a category error, not a finding. The four readings that matter, taken from ad sales and total sales moving together or apart: | Pattern | Reading | |---|---| | Ad sales rising, total sales flat | advertising is replacing organic, not adding | | Both rising together | advertising is genuinely growing the category | | Total sales strong, ad sales small | a healthy organic category — do not over-invest | | Total sales falling, ad sales holding | organic erosion hidden behind steady ad performance | The first and last rows are the ones worth catching, and neither is visible on any advertising screen. That is the whole argument for using this one. Step 5 — Watch units beside sales Total Units next to Total Sales separates two very different stories that look identical on revenue alone: Units rising, sales flat — you are selling more for less. A price cut, a mix shift toward cheaper items, or discounting. Units flat, sales rising — price or mix moved up. Fine, unless it is masking softening demand. Both falling — a real demand problem, and the only one of the three that is unambiguous. Step 6 — Turn the benchmark into targets Once you have a stable baseline per category, it becomes the ACoS target for that category's rule group — not a number from a blog post. Set the target where the category actually performs, then move it deliberately when you want the category to do something different. See Set a Target ACoS Across a Whole Portfolio. Review monthly. A category whose TACoS has drifted up while total sales held flat is quietly becoming more expensive to run, and that is exactly the slow change a fixed industry benchmark would never surface. How this differs from Optimization Groups Both group things and report per group. They group different things: | | Category Insights | Optimization Groups | |---|---|---| | Groups | products | campaigns | | Shows | total and ad sales | ad performance | | Answers | how is this part of the business? | how is this part of the advertising? | Most accounts want both, and they will not line up perfectly — one campaign often advertises products from several categories. That mismatch is normal. What not to do Do not benchmark against published averages. They are averages of businesses with different margins, prices and competition. Do not use a quarter with a promotion in it as your baseline. You will benchmark against an event. Do not compare a launch to a mature line. See step 4. Do not let categories drift. A category that made sense two years ago and has quietly absorbed six unrelated SKUs is producing an average of nothing. Availability starts at PRO. On Starter the section is not in the sidebar. See What's Included in Each Plan. Related - Category Insights: Benchmarking Your Categories — the screen in full. - Understanding ROAS and Key Amazon Advertising Metrics — ACoS against TACoS. - Set a Target ACoS Across a Whole Portfolio — step 6. - Optimization Groups — the advertising-side equivalent. - Scoring Product Opportunities — comparing a candidate market against what you already run.

Find Zero-Sales ASINs and Decide What to Do With Them

Every catalog accumulates them: ASINs that sell nothing, month after month, quietly consuming storage fees, attention and occasionally ad spend. Deleting them is not automatically right. Some are dead, some are broken, and some are doing a job that has nothing to do with their own sales. The work is telling them apart. Time: about an hour. Do it quarterly. Watch the walkthrough Prefer to watch? Eighty-nine seconds on confirming it could have sold before concluding it did not. Step 1 — Build the list Analytics → Sales Performance. Product-level performance, sortable by units and sales Set the range to 90 days — long enough that a slow seller is genuinely slow rather than merely unlucky — and sort by units, ascending. Take everything at zero units, plus everything in low single digits. That is your candidate list. On most catalogs it is longer than expected. Step 2 — Eliminate the broken ones first Before concluding a product does not sell, confirm it could have sold. A large share of zero-sales ASINs are not demand failures at all. Check each candidate for: Out of stock. Days of cover at zero for most of the window. It did not fail; it was not available. See Inventory: On Hand, Inbound, Days of Cover and Velocity. Suppressed or inactive listing. A listing that is not buyable sells nothing, and nothing about that is the product's fault. Zero visibility in Rank Tracker. A product returning 0.00% visibility is not ranking within the tracked depth for its terms. Customers are not choosing it over rivals; they are never seeing it. See Rank Data Shows Blank. No advertising, ever. A product that has never been advertised and does not rank organically has not been tested. It is not a failure — it is an unopened box. Anything in this group is a fix, not a cut. Set it aside. Step 3 — Work out what the survivors are for The remainder are genuinely available, visible, and not selling. Now ask what job each one does, because several jobs do not show up in its own sales line: Is it a variant that feeds a parent? Size and colour variants share a listing and reviews. Killing a slow variant can weaken the parent's selection. Is it a gateway product? Check NTB ASINs Analysis. A product with modest sales that brings a meaningful share of your new customers is not underperforming — it is recruiting, and its value lands on other products' revenue. See Protect the products that bring you new customers. Is it half of a pair? Check ASIN Purchase Upsell. A product bought alongside a hero may be part of why the hero converts. Is it defensive? Some ASINs exist to occupy a search result or a price point rather than to sell in volume. If none of those apply, it is what it looks like. Step 4 — Decide, per product Five outcomes, and the honest one is usually not "delete": | Finding | Do | |---|---| | Out of stock most of the window | Restock and re-test. No decision yet. | | Listing suppressed or inactive | Fix the listing. No decision yet. | | Never advertised, no organic rank | Test it properly — a small budget for 30 days | | Recruits customers or supports a pair | Keep, and stop judging it on its own sales | | Available, visible, advertised, still nothing | Discontinue | Only the last row is a genuine failure, and it is usually the smallest group. Step 5 — Test the untested ones properly A product that has never had traffic deserves one honest test before you write it off: Give it a real budget for 30 days. A token amount buys a handful of clicks and tells you nothing. Advertise it on terms that fit it, not on your hero's keywords. A product tested on the wrong queries fails for the wrong reason. Set the bar in advance. Write down what would count as success — a conversion rate, a number of orders — before you start. Otherwise you will find a reason to keep testing. At the end of 30 days it either earned a place or it did not, and you have an answer rather than a suspicion. Step 6 — Clean up what you cut Discontinuing an ASIN is not just a catalog change: Remove it from rule groups, or your rules keep evaluating a product that no longer matters. Remove it from Rank Tracker projects. A discontinued product at 0.00% visibility drags the project's average down and makes a healthy project look mediocre — see Reading Visibility, Average Position and Top 10. Check nothing points at it. A bundle or a variant relationship that assumes it exists will break. Watch its category for a month. If total sales in that category fall by more than the product was selling, it was doing something you had not accounted for. What good looks like A quarterly pass that moves most candidates into "fix" or "test" and only a handful into "discontinue" — and a catalog where every remaining ASIN has a reason to be there that somebody could state out loud. What not to do Do not cut on sales alone. Steps 2 and 3 exist because the sales line is the least informative thing about a zero-sales ASIN. Do not cut a gateway product because its own revenue is unremarkable. That is precisely the mistake this playbook is designed to prevent. Do not cut during or just after a stock-out. You would be judging a product that was not on sale. Do not do it monthly. A quarter is the shortest window that separates slow from dead. Related - Sales Performance: Product-Level Reporting — step 1. - NTB ASINs Analysis: Finding Your Gateway Products — step 3, the check that saves good products. - Rank Data Shows Blank — when zero visibility is the cause. - Inventory: On Hand, Inbound, Days of Cover and Velocity — step 2.

Shift Budget from Harvesting to Acquisition

Every account drifts the same way. Remarketing and branded terms look efficient, so they get more money; prospecting looks expensive, so it gets less. Repeat quarterly and you end up with an advertising programme that is very good at selling to people you already have. It looks healthy right up until growth stops, because the pool it harvests from has stopped being refilled. Time: about two hours. Do it quarterly. Watch the walkthrough Prefer to watch? Seventy-nine seconds on measuring the drift, and judging the fix properly. Step 1 — Measure the drift Two numbers, taken from the same month: NTB share of purchases — AMC Cloud → NTB Purchases. New-to-brand share across ad product types NTB users — AMC Cloud → NTB Customers, which counts people rather than orders. Now compare against the same month last quarter, and last year if you have it. The pattern to look for: | What you see | Reading | |---|---| | NTB share falling, total sales holding | the drift — you are converting advertising into a remarketing programme | | NTB share holding, NTB users falling | the business is shrinking; the ratio is flattering it | | Both holding | no problem to solve; stop here | | NTB share rising | acquisition is working — check it is not just a smaller base | Track the count, not just the share. A percentage can improve while the absolute number of new customers falls, and that is the worst outcome disguised as the best. Step 2 — Find where the money actually is Sort your spend into the two jobs. The split is usually not what people assume: Harvesting — branded terms, remarketing and retargeting audiences, exact-match keywords on proven queries. These convert well because the customer had largely decided. Acquisition — generic terms, broad and auto discovery, prospecting audiences, upper-funnel display. Two screens do most of this: Generic, Branded and Competitor Terms classifies your search queries into the three buckets and tells you what proportion of your search spend is chasing people who already know your name. See that article for how the classification works — and note that a single generic word in your brand list will make the split useless. DSP Campaigns splits display into prospecting versus remarketing by campaign name. A display buy that is overwhelmingly remarketing is harvesting, whatever it is called internally. Step 3 — Work out what the harvesting is really earning Harvesting spend has a specific weakness: it is the spend most likely to have happened anyway. Somebody searching your brand name was going to find you. A remarketing ad shown to a customer who already has the product in their basket is charged for a sale it did not cause. You cannot measure this precisely without a holdout, and TrackIQ does not run holdouts. But Path to Purchase gives you the strongest available proxy: compare a single-touch path against the same path with a harvesting touch added. If the multi-touch version produces no more purchases and fewer new customers, that extra exposure is a cost, not a contribution. See Prove your display campaigns created demand for how far this evidence can honestly be pushed. Step 4 — Move the money, in a bounded way Do not rebalance the whole account at once. You will not be able to attribute the result. Take 10–15% out of harvesting, from the places most likely to be self-cannibalising: branded terms above the volume you need to defend, and remarketing audiences with the shortest lookback. Put it into acquisition with a defined target and a defined end date. Generic terms with real search volume, or prospecting audiences you have not funded properly before. Set the acquisition target where acquisition can actually run. Prospecting will always look worse on ACoS than harvesting — that is the definition of the two strategies, not a finding. Give it its own rule group and its own target. See Set a Target ACoS Across a Whole Portfolio. Step 5 — Judge it on the right measure, after the right interval The measure is NTB users, not ACoS. If you judge an acquisition shift on efficiency, you will reverse it within a month, because efficiency is exactly what you traded away on purpose. The interval is a quarter. New customers take time to arrive and longer to repeat. A month tells you the ACoS went up, which you already knew. At the end of the quarter: | Result | Do | |---|---| | NTB users up, blended ACoS up a little | working — hold, consider another 10% | | NTB users up, blended ACoS up sharply | working but expensive — tighten the acquisition target | | NTB users flat, ACoS up | the acquisition spend is not reaching new people — the audiences or terms are wrong | | NTB users down | reverse it, and look for a cause outside advertising | The third row is the common one, and the fix is targeting rather than budget. Money moved into prospecting that reaches the same people is not acquisition; it is harvesting with a different label. Step 6 — Protect what recruits While you are here, make sure the products that do the acquiring are not being managed like everything else. A gateway ASIN starved of budget or allowed to stock out will undo this whole exercise. See Protect the products that bring you new customers. What not to do Do not cut harvesting to zero. Branded defence is cheap and losing it is expensive. This is a rebalance, not a purge. Do not move more than 15% in one go. You will not learn anything from a change that big. Do not do this and a bid-strategy change in the same quarter. One variable at a time. Do not read one month. AMC reports monthly cohorts and seasonality moves NTB share hard. Do not confuse NTB with incremental. New to brand means the customer had not bought from you in 365 days. It does not prove the ad caused it. Related - NTB Customers: New vs Repeat, Counted as People — the measure that matters in step 5. - Generic, Branded and Competitor Terms — splitting search spend by job. - Decide whether your upper-funnel DSP is worth keeping — the display half of the same question. - Set a Target ACoS Across a Whole Portfolio — giving acquisition its own standard. - Scaling Profit Campaigns — where the freed budget should go.

Prepare for Prime Day: Budget Rules and Inventory Cover

Prime Day rewards preparation and punishes improvisation. Demand arrives faster than you can react, budgets cap out at nine in the morning, and the products that sell out on day one stop selling for the fortnight that follows. Almost all of the work happens before the event. Time: a day, spread over the four weeks before. Then very little on the day. Watch the walkthrough Prefer to watch? Eighty-six seconds on the four weeks before, and the three days after. Four weeks out — inventory decides everything else Inventory. Days of cover, inbound and velocity per product Nothing else on this list matters if you run out. Work the arithmetic properly: cover needed = normal velocity × event multiplier × (lead time + transit + Amazon receiving) Assume the multiplier is higher than last year on your best products and lower than you hope on the rest. Event demand concentrates: a handful of ASINs take most of the lift. Protect the gateway products hardest. A stock-out on the product that recruits your new customers costs you the acquisition, not just the sales — and Prime Day is when the most first-time buyers are looking. See Protect the products that bring you new customers. Get stock in early. Receiving times stretch before a major event; stock that arrives in the week before may not be checked in for it. Two weeks out — decide the budget shape Not just how much, but what happens when it runs out. A capped campaign at 9am is the single most expensive Prime Day mistake. The traffic keeps arriving and you are not in it, and the auction is at its most valuable precisely then. Raise daily budgets on the campaigns you actually want to win, and be honest about which those are — usually your best-converting exact terms and your hero products, not everything. Check Campaign Daily Budget Capped alerts in the run-up. Anything routinely capping on a normal day will cap by breakfast on the day itself. See Alert Categories, Priorities and What Actually Triggers Them. One week out — put the rules in a safe state Automations. Look-back, attribution lag and the change cap This is the step teams forget, and it does real damage. Rules do not know an event is happening. A pause rule reading a 14-day look-back through Prime Day sees an enormous spike in clicks and spend, and days later it sees the post-event slump. Both make it behave badly. Three specific adjustments: Pause rules — turn them off, or raise their thresholds. During the event, spend and clicks climb far faster than orders are attributed, which is exactly the shape that triggers a pause on something that is working. Bid rules — tighten the change cap. Cut max change per run to something small. A goal-seeking bid rule reacting to event-week ACoS will make large corrections based on a week that will not repeat. Check "ignore most recent" is at least 1–2 days. Attribution lags hardest when volume is highest, so fresh days under-report sales more than usual. See Look-back Windows and Adjustment Frequency. Dry-run anything you leave enabled. The preview computes every change and applies none. See Dry-Run Preview. The day — watch three things, change almost nothing Resist the urge to optimise hourly. You do not have enough data to be right, and you will be reacting to attribution that has not landed. Budget caps. The only genuinely urgent thing. A capped campaign is lost traffic, and it is the one problem that gets worse every hour you leave it. Stock cover. Sell-through rates that looked comfortable can halve in an afternoon. Nothing else. ACoS on the day is meaningless — spend is complete and sales are not. Anything you conclude from it will be wrong in three days' time. Aria in Slack is the right tool for the day. Ask "which campaigns are capped?" and "anything below a week of cover?" rather than sitting in dashboards. See Get your daily number in Slack without opening TrackIQ. The three days after — do nothing The most important instruction in this playbook. Attribution is still landing. Sponsored Products credits sales for up to 7 days after the click, and the other formats for 14. Event-day ACoS looks catastrophic on the day and normal by the end of the week. Rules stay off or capped until the numbers settle. Turning them back on into a distorted look-back window is how a good event becomes a bad fortnight. Expect a slump. Demand pulled forward into the event is not there afterwards. That is normal, it is not a problem to fix, and cutting budgets in response to it just deepens it. Then teardown Once attribution has settled — a week or so — the numbers are worth reading properly. That is a separate exercise, and it is where the value of the event actually gets banked. See Run a post-event teardown. What not to do Do not raise every budget. Money spread evenly across everything buys nothing in particular. Do not leave pause rules running. They will pause things that are working. Do not judge anything on event-day ACoS. Do not launch new products for the event. A listing with no history and no reviews will not win an auction against established ones, and you will conclude the product is bad when the timing was. Do not forget to turn the rules back on. Put a reminder somewhere, because a month of unmanaged bids costs more than the event made. Related - Catch a stockout before it costs you rank — the inventory half, in detail. - Guardrails to Set Before You Automate Anything — the settings to check in week three. - Run a post-event teardown — what to do when the dust settles. - Why SP, SB and SD Numbers Don't Add Up — why the day-after numbers lie. - Designing the Offer — the offer mechanics deals depend on. - Preparing a Launch: CTR and CVR Tactics — deals as a launch instrument rather than a promotion. - Inventory Forecasting and Purchase Orders — ordering and capacity ahead of the event. - Reimbursements, COGS and FBA Capacity — requesting storage before it is refused. - Bulk Auto Campaigns for Peak Season — five auto campaigns per product, checked daily.

Run a Post-Event Teardown

Most teams read event numbers on the day, feel something, and move on. A week later the attribution has landed, the picture has changed, and nobody goes back to look. The teardown is where an event stops being a spike on a chart and becomes something you can repeat deliberately. Time: two hours, about a week after the event ends. Watch the walkthrough Prefer to watch? Eighty-four seconds on reading it a week late, and the four numbers to record. Step 0 — Wait Do not start this until attribution has settled. Sponsored Products credits sales up to 7 days after the click; the other formats up to 14. A teardown run on day two will systematically undervalue everything and overvalue nothing. A week after the event is the earliest honest read. Ten days is better if you ran Sponsored Brands or DSP. Step 1 — Did it actually grow the business? Analytics → Category Insights. Total Sales beside Ad Sales, per category Start here rather than in the ad screens, because this is the only place with Total Sales beside Ad Sales. The question is not "did we sell more" — you did, that is what events do. It is how much of it was paid, and did the whole category grow? | Pattern across the event window | Reading | |---|---| | Total sales up, ad sales up proportionally | ordinary event lift | | Total sales up much more than ad sales | the event created organic demand — the best outcome | | Ad sales up, total sales barely moved | you bought sales you would have had anyway | | Total sales up, ad sales up far more | you paid a premium for the same demand | The third row is the one worth catching, and it is invisible on any advertising screen. Step 2 — Compare against the right baseline Not against the week before — that is depressed by people waiting for the event. Not against the week after — that is depressed by demand already pulled forward. Compare against the same event last year if you have it, and against a normal month otherwise. Write down the multiplier you actually achieved per category, because that is what you will plan the next event with. Step 3 — Where did the money go, and did it work? Amazon Ads → Campaigns, with vs previous and % change. Campaign performance with the comparison toggle Sort by spend change and answer three questions per campaign: Did it cap? A campaign that hit its budget is a campaign you underfunded, and you now know by how much. This is the single most actionable output of the teardown. Did the extra spend convert? Compare its ACoS across the event against its normal ACoS. Events usually cost more per sale; the question is whether the premium was worth the volume. Was it the right campaign? Concentration matters. If your spend increase was spread evenly across everything, you did not make a choice — you just scaled. Step 4 — What did the traffic actually ask for? Amazon Ads → Search Terms, aggregated, for the event window. Events change the query mix, not just the volume. New terms appear, gift-shaped queries spike, and broad campaigns match things they never normally see. New high-spend terms that converted — harvest them. They are proven, and they will come back next event. See Harvest Winning Search Terms into Exact-Match Campaigns. New high-spend terms that did not — negative-match them before the next event, not after it. This step is where a teardown pays for itself, because the query mix repeats far more reliably than the numbers do. Step 5 — Did you acquire anybody? AMC Cloud → NTB Customers, for the event month. Events are one of the few moments when large numbers of people who have never bought from you are actively shopping. Whether you converted any of them is a different question from whether you sold a lot. NTB users up sharply — the event did the job events are best at. Those customers are the return, and they arrive over the following months. Sales up, NTB users flat — you discounted to your existing customers. Profitable perhaps, but it did not grow the base, and that changes how you value the discount next time. Remember AMC reports monthly cohorts, so this is the event month as a whole rather than the event days. See Why AMC Data Is Monthly and Sometimes Empty. Step 6 — The inventory post-mortem Two failures, both expensive, and both only visible afterwards: What sold out? Note how long it was unavailable and what its rank did afterwards. That is the real cost of under-ordering, and it is usually larger than the missed sales. Check BSR Tracker for the shape of the recovery. What did not sell? Event stock that did not move is now carrying cost on a product with proven weak demand at a discount. Those ASINs belong in the next zero-sales review. See Find zero-sales ASINs and decide what to do with them. Step 7 — Write down four numbers The teardown is worthless unless next year's version of you can find it. Record, per category: 1. The sales multiplier you achieved. 2. The budget each key campaign needed to avoid capping. 3. The ACoS premium the event cost. 4. The NTB lift, if any. Those four turn the next event from a guess into a plan. Step 8 — Put the rules back If you disabled or capped rules for the event, turn them back on — but only once the look-back window has cleared the event. A 14-day window still containing event days will produce large, wrong corrections. Dry-run each one first and check the changed count looks normal. See Dry-Run Preview. What not to do Do not tear down on day two. Step 0 exists for a reason. Do not compare against the adjacent weeks. Both are distorted, in opposite directions. Do not treat the post-event slump as a problem. It is demand you already collected. Do not skip step 7. An unwritten teardown is an afternoon spent learning something you will have forgotten by the next event. Related - Prepare for Prime Day: budget rules and inventory cover — the other half. - Category Insights: Benchmarking Your Categories — step 1. - NTB Customers: New vs Repeat, Counted as People — step 5. - Why SP, SB and SD Numbers Don't Add Up — why step 0 matters.

Spot Cannibalisation Between Sponsored Brands and Sponsored Products

Sponsored Brands and Sponsored Products often bid on the same queries for the same products. When they do, you can end up paying twice to reach one shopper, or paying SB prices for a sale SP would have won anyway. The word "cannibalisation" gets used loosely. This playbook is about establishing whether it is actually happening, because the answer is frequently no. Time: about 90 minutes. Do it twice a year, or after any SB expansion. Watch the walkthrough Prefer to watch? Seventy-eight seconds on what cannot be measured, and the test that can. Step 0 — Know what you cannot measure Be clear about the limit before you start, because it shapes every conclusion below. Amazon does not tell you when two of your ads appeared on the same search for the same shopper. There is no report for it, in TrackIQ or anywhere else. What you can do is find the overlap in queries and products, measure the two formats against each other, and reason carefully about the difference. Anyone claiming a precise cannibalisation figure is estimating. So will you be — just knowingly. Step 1 — Find the overlap Amazon Ads → Search Terms, with Aggregated Search Terms on. Search terms, aggregated across the account Aggregation matters here for a specific reason: it collapses the same customer search across every keyword and ad group it came from, which is exactly the view you need to see one query served by two formats. Sort by spend and look for queries that appear in both an SB and an SP campaign. Those are your candidates. Everything else is not a cannibalisation question at all. A large SB budget with almost no query overlap is not cannibalising anything — it is reaching different searches, and this exercise ends here for those campaigns. Step 2 — Check whether they are chasing the same products Overlapping queries only matter if they lead to the same ASINs. Amazon Ads → Product Ads shows ASIN-level performance across formats. If your SB campaign drives a category page or a range of products while SP drives one hero ASIN, they are doing different jobs on the same query, and that is usually fine. If both formats point at the same single ASIN on the same query, you have a genuine candidate. Step 3 — Compare the two on that query Amazon Ads → Campaigns, filtered to the overlapping campaigns. Campaign performance with the comparison toggle For each overlapping query, compare the formats on: CPC. Sponsored Brands typically costs more per click. If SB is paying materially more for the same query and converting no better, that is the clearest signal in this whole exercise. Conversion rate. If SB converts better, it may be earning its premium — brand-led creative can qualify the click before it happens. New-to-brand share. This is the one that settles most arguments, and it needs AMC. Step 4 — Read the NTB split AMC Cloud → NTB Purchases. New-to-brand share by ad product type Sponsored Brands frequently shows the highest NTB rate of any format, and that changes the conclusion entirely. If SB's NTB share is well above SP's, the two formats are not competing for the same customer — SB is reaching people who do not know you and SP is converting people who do. Cutting SB to save money on overlap would remove your best recruiter to save a small amount on duplicate exposure. If SB's NTB share matches SP's, they are reaching the same population, and the overlap is real. Now the CPC comparison from step 3 decides which one should hold the query. Watch the volume trap: SB often has the best NTB rate on a tiny share of your new customers. Both facts matter. See NTB Customers: New vs Repeat, Counted as People. Step 5 — Test it, do not assume it The only honest way to size the effect is to change one thing and watch. Pick your three or four clearest overlaps. Not the whole account. Reduce, do not remove. Lower the SB bid on those queries by 25–30% rather than pausing. Pausing tells you what happens with none of it; reducing tells you about the margin, which is what you are actually deciding. Watch total sales for that product, not the campaign's sales. This is the crucial instrument. If SB spend falls and SP sales rise by roughly the same amount, the traffic simply moved — that is cannibalisation, and you have just saved the difference. If SB spend falls and total sales fall too, SB was bringing something SP was not. Give it three weeks. Attribution lags 7–14 days depending on format, so a two-week read is mostly noise. See Why SP, SB and SD Numbers Don't Add Up. Step 6 — Decide | What the test showed | Do | |---|---| | SP sales rose to match SB's fall | genuine overlap — keep SB reduced on those queries | | Total sales fell with SB spend | SB was incremental — restore it | | Nothing moved either way | the spend was small; leave it and move on | | SB fell, NTB users fell | restore it — you cut acquisition, not waste | The last row is the one to watch for, and it is the reason step 4 exists. What not to do Do not pause Sponsored Brands wholesale to "stop cannibalisation". You will remove a discovery format and attribute the resulting decline to something else three months later. Do not test across the whole account. You will not be able to read the result. Do not judge on campaign-level sales. Both campaigns' numbers will move; the product's total is the honest measure. Do not conclude from one week. Two formats with different attribution windows need three. Do not run this alongside a bid-strategy change. One variable. Related - Search Terms and Aggregated Search Terms — step 1, and why aggregation matters. - NTB Purchases: New Business by Campaign — step 4, the deciding evidence. - Why SP, SB and SD Numbers Don't Add Up — the attribution differences behind the test. - Product Ads: ASIN-Level Ad Performance — step 2. - Create a Sponsored Brands Campaign — how the brand banner is set up in the first place.

Feed TrackIQ Data into Your Own AI Workflow

TrackIQ's screens answer the questions TrackIQ anticipated. Sooner or later you have one it did not — a question that joins your Amazon data to your own margins, your own forecast, or your own spreadsheet. MCP is the answer. It connects your TrackIQ data to the AI tools you already use, so the analysis happens wherever you work rather than only inside the product. Time: ten minutes to connect. Then it is just how you work. Watch the walkthrough Prefer to watch? Eighty-five seconds on connecting it, and the questions worth asking it. Step 1 — Connect it Settings → MCP & API Keys. The MCP connection panel The setup is deliberately small: paste one URL into Claude Desktop, Claude Code or Cursor and it registers itself. No JSON config file, no credentials to copy, no key management. Cursor is the exception worth knowing about — it does not yet support Dynamic Client Registration, so there is a manual OAuth client path for it. The Advanced section has per-client setup instructions. See Installing TrackIQ MCP in Claude, ChatGPT and Cursor. Step 2 — Understand what it can reach MCP exposes your TrackIQ data as tools an AI can call — campaigns, ad groups, keywords, targets, search terms, product performance, inventory, rank, SQP, DSP and AMC. Two things follow from that: It reads the same data the screens do. If a number is not in TrackIQ, MCP cannot invent it. What each plan unlocks differs. See MCP Tiers: What Each Plan Unlocks — and note the counter-intuitive one: the MCP Only plan includes Search Query Performance, Amazon DSP and AMC Cloud, which Starter does not. Step 3 — Ask the questions the screens cannot This is the point of the exercise. The screens are better than MCP for anything they already do — they are faster, sorted, filtered and visual. MCP earns its place on questions that join TrackIQ data to something else: Your own margins. TrackIQ knows spend and sales; it does not know your COGS. "Pull last month's product performance and work out contribution after ads using the cost column in this spreadsheet." Your own forecast. "Compare actual units against the forecast in this file and list the ten biggest misses." Cross-source reasoning. "Take the queries with the worst cart share, look up which ASINs they land on, and check those ASINs' stock cover." One-off shapes. Anything you would otherwise export to CSV and pivot by hand. Step 4 — Build the repeatable ones properly Once a question turns out to be worth asking every week, stop asking it conversationally. Write it down as a prompt you reuse. The wording matters more than people expect — name the date range, the metrics, and what a good answer looks like. The same discipline as Asking Aria a Good Question. Keep the source data current. If your margin spreadsheet is three months stale, so is every answer derived from it. Decide where the output goes. An analysis nobody reads is worse than no analysis, because it cost something. Step 5 — Watch who is querying what Settings → MCP & API Keys → Call Logs. Call logs, showing the question each tool asked The call log records the natural-language question the AI asked on your behalf, not just an endpoint — with Question, Tool Calls and Posted At, its own date range, and Search, Filters, Manage Columns and Export. That makes it a genuine audit trail rather than a debug log, and it is worth checking for two reasons: Verification. If an answer looked wrong, the log shows what was actually asked and which tools ran. That is usually enough to see whether the question or the data was at fault. Access hygiene. Per-brand API keys mean you can see which integration is pulling what. An unfamiliar pattern is worth a look. See MCP Call Logs: Seeing Which AI Tools Queried Your Data. When to use which | Question | Use | |---|---| | Something a screen already shows | the screen — faster and sorted | | A daily number, on your phone | Aria in Slack | | Joins TrackIQ data to your own data | MCP | | A recurring report for other people | a report template | | A one-off shape you would pivot by hand | MCP | The mistake is using MCP for things the product already does well. It is not a faster dashboard; it is a way to ask questions the dashboard was never going to answer. What not to expect It cannot change anything. MCP is read access. Bids, budgets and rules are changed in TrackIQ or in Amazon. It does not know your business. It knows your Amazon data. Context about discontinued lines, promotions or supplier problems has to come from you. Check anything you will act on. An AI reading your data can still reason badly about it. The same caution as Checking Aria's Work When an Answer Looks Wrong. Data freshness is TrackIQ's freshness. MCP does not reach past TrackIQ into Amazon. See Where Your Data Comes From and How Fresh It Is. Related - What Is MCP, and Why It Matters — the concept. - Installing TrackIQ MCP in Claude, ChatGPT and Cursor — step 1 in detail. - The Full MCP Tool Reference — everything it can reach. - MCP Tiers: What Each Plan Unlocks — what your plan includes.

Work Out Which Queries Lose You the Sale at the Cart

A query that spends money and does not convert is not one problem. It is four possible problems, and they need four different fixes — only one of which is anything to do with bidding. This is how to find which one you have, query by query, in about an hour. Time: an hour. Do it monthly, or whenever ACoS drifts on a term that used to work. Watch the walkthrough Prefer to watch? Eighty-two seconds on the four places a query breaks, and why only one of them is bidding. Step 1 — Let the product pick the queries Search Query Performance → Recommended Actions. Recommended actions, sorted by money at stake Filter to Fix conversion and Fix listing. Those two categories are exactly the population this playbook is about — queries where traffic arrives and the sale does not. The Why column already tells you which of the two it is: "You're shown but shoppers click competitors" — a click problem. Your listing is losing the search result. "Shoppers click you but buy elsewhere" — a conversion problem. Your page is losing the visit. Take the top five by money. Do not try to fix twenty. Step 2 — Open the funnel and find the break Back on SQP Analysis, click the query. The four funnel stages, brand against market Four stages, each with your rate against the market's rate at that step. Walk them in order and find the first stage where you fall behind: | Your rate vs market | Broken at | The fix is in | |---|---|---| | impression share low | visibility | rank and bids | | CTR below market | the search result | main image, title, price, rating | | cart rate below market | the detail page | content, A+, reviews, price | | CVR below market at cart | checkout | price, delivery, stock, Buy Box | The comparison is what makes this usable. A 2% CTR is neither good nor bad until you know the market is at 1.5% or at 4%. Stop at the first break. Later stages inherit the problem from earlier ones, so fixing stage three while stage two is broken achieves nothing. Step 3 — Check it is the query, not the product The ASIN split for the same query The ASIN breakdown shows which of your products the query actually reached, with click, cart and purchase share each. This is the step that saves you from fixing the wrong thing: One ASIN dragging the rest — the query is fine. Fix or exclude that product. The query landing on a product it should not — a targeting problem, not a listing one. Negative-match it in the campaign sending it there. All ASINs performing similarly — now it genuinely is the query, and step 2's diagnosis stands. Step 4 — Act on what you found Losing the click. The main image does most of the work in a search result, then price, then rating. Look at the sellers winning the query and be honest about how yours compares at thumbnail size. Losing the cart. The detail page. Bullets, A+ content, images beyond the first, and the review count relative to the competition. Losing the purchase after the cart. The narrowest and most fixable set: price, delivery promise, availability, Buy Box. Check stock first — see Catch a stockout before it costs you rank. Losing on volume alone, winning on rates. The happy case. You convert better than the market and are not shown enough. This one is a bid and rank problem, and it is the only one where spending more is the right answer. Step 5 — Decide what not to fix Some queries should be abandoned rather than repaired, and saying so is part of the job: Competitor-brand queries. A better page cannot win a shopper who wants somebody else's brand. TrackIQ deliberately excludes these from Fix listing and Fix conversion recommendations for exactly that reason — provided the rival is in your competitor list. See Generic, Branded and Competitor Terms. Queries for a product you do not really sell. Adjacent terms that your broad match caught. Negative-match them. Queries where the market rate is also poor. If everybody converts badly on a term, it is a browsing query rather than a buying one. Step 6 — Re-check after a fortnight Listing changes take time to show. SQP reports weekly cohorts, so compare the two weeks after your change against the two before — not an arbitrary range. See SQP Is Weekly: Reading Week Cohorts Correctly. The measure of success is the rate against market at the stage you fixed, not the query's total sales. Sales move for many reasons; the rate tells you whether the change worked. What not to do Do not raise bids on a click or conversion problem. More impressions on a listing people do not click costs more money to be ignored. Do not change the listing and the bid in the same fortnight. You will not know which one moved the rate. Do not fix twenty queries at once. Five is enough to learn from and few enough to attribute. Do not read a single week on a low-volume query. The five-week window exists for a reason. Related - Search Query Cart Analysis: Why a Query Doesn't Convert — the screen, in full. - Recommended Actions: The Queries That Need a Decision — step 1. - Generic, Branded and Competitor Terms — why some queries are excluded. - SQP Is Weekly: Reading Week Cohorts Correctly — step 6.

Build a Monthly Board Pack

A board pack is not a bigger weekly report. The audience is different, the time horizon is different, and the questions are different — a board wants to know whether the business is working, not whether last Tuesday's ACoS was acceptable. Built once, it should take under an hour a month. Time: two hours to build. Under an hour monthly after that. Watch the walkthrough Prefer to watch? Eighty seconds on leading with total sales, and the two paragraphs nobody else can write. Step 1 — Decide what the board is deciding Every slide should serve one of three decisions. If it serves none, cut it: Should we put more money in? — growth, acquisition, category momentum. Is anything going wrong? — erosion, concentration risk, availability. Are we spending well? — efficiency, but as context rather than the headline. A board pack that opens with ACoS has misjudged the room. ACoS is an operating metric; it belongs in the appendix. Step 2 — Lead with the whole business, not the advertised part Analytics → Category Insights. Total Sales beside Ad Sales, per category This is the right opening because it is the only screen with Total Sales beside Ad Sales, which lets you show TACoS — how much of the business is bought. Per category, month over month and against the same month last year: Total sales, then TACoS, then total units. Three numbers, one table. The reading that earns its slot is the divergence: ad sales rising while total sales stays flat means advertising is replacing organic rather than adding to it. That is a strategic finding, and it is invisible on every advertising screen. See Benchmark Yourself Against Your Own Category. Step 3 — Show acquisition, not just sales AMC Cloud → NTB Customers. New customers by format, rate and volume NTB Users is the closest thing to a growth metric the account produces, and most board packs never include it. Show the count month over month, not just the share — a percentage can improve while the absolute number of new customers falls, which is the worst outcome wearing the best-looking number. If you want one derived figure, blended acquisition cost — total ad spend divided by NTB users — is rough but far more board-relevant than ACoS. Say that it is rough. Step 4 — Name the risks Boards remember the pack that told them about a problem before it arrived. Two are worth a standing slide: Concentration. If a small number of ASINs bring most of your new customers, say so and show the number. That is a real dependency, and it belongs in a risk register rather than in an ads report. See NTB ASINs Analysis: Finding Your Gateway Products. Availability. Days of cover on the products that matter, and any stock-out that happened during the month with what it cost in rank. See Catch a Stockout Before It Costs You Rank. Step 5 — Build it once, in a template Business Reports → AI Report Generator, then save the result. Describe the report properly, then set the comparison range Start from Executive Review, choose Slides, and describe it with the audience and the decision named: Monthly board pack. Total sales and TACoS by category against last month and the same month last year, new-to-brand customers by month, the ASINs driving acquisition, and any stock cover below four weeks. The decision is next quarter's investment. Turn on the comparison range. For a board pack this is not optional — a monthly number with nothing beside it is not information. Then ⋮ → Save as Template, and next month starts from a known-good structure. See Business Report Templates and the AI Report Generator. Step 6 — Write the two paragraphs nobody else can The pack's numbers will be right and its interpretation will be generic. That is the half you supply: What changed, and why. One paragraph. Not what the numbers did — what caused them. What we are doing about it. One paragraph. The decisions already taken, so the meeting starts from action rather than from explanation. Aria can draft the first — "what changed most vs last month and what caused it?" — but read it before it goes in. See Checking Aria's Work When an Answer Looks Wrong. A structure that holds up | Slide | Content | |---|---| | 1 | Headline: total sales, growth, one sentence | | 2 | Category performance — total sales and TACoS | | 3 | Acquisition — NTB users by month | | 4 | Risks — concentration and availability | | 5 | What we are doing about it | | Appendix | Efficiency: ACoS, spend by format, campaign detail | Five slides and an appendix. The appendix is where you put everything you would be asked about but should not lead with. What not to do Do not lead with ACoS. It answers a question the board did not ask. Do not change the structure monthly. A board reading the same shape each month spots change instantly. A board reading a new layout spends the meeting orienting. Do not present a number without a comparison. Do not hide a bad month. The pack that surfaces a problem early buys you the credibility to be believed about the next one. Do not send it unread. A generated pack read first by its audience is how a wrong number becomes a board decision. A note on delivery There is no email scheduling for reports today, so producing from the template and sending it yourself is the current path. It takes minutes once the template exists. Related - Business Report Templates and the AI Report Generator — step 5. - Benchmark Yourself Against Your Own Category — step 2. - NTB Customers: New vs Repeat, Counted as People — step 3. - Build a Weekly Executive Summary You Never Have to Write — the weekly sibling.

Audit Your Placement Multipliers

Placement multipliers are the highest-leverage setting most accounts never revisit. They were set once, at launch, from a guess — and they are still quietly multiplying every bid in the campaign. An hour, once a quarter, on evidence you already have. Watch the walkthrough Prefer to watch? A hundred and ten seconds on the four decisions, and the two mistakes that make the result unreadable. What a multiplier actually does In Amazon Ads you can raise a campaign's bid for a specific placement by a percentage. A 50% Top-of-Search multiplier means a £1.00 base bid becomes £1.50 when your ad is eligible for the first row of results. Two things follow, and both catch people out: It multiplies, it does not replace. Raise the base bid and every multiplier's effect rises with it. The multiplier is set in Amazon Ads; the evidence is in TrackIQ. This playbook is about producing the number, not about the control that consumes it. Before you start Set a window with enough clicks. Thirty days, or long enough that every placement in a campaign has a few hundred clicks. Placement data is thin by construction, and a quarter's worth of decisions should not rest on forty clicks. End the window a fortnight ago, not today. Attribution is still landing on recent days and will make every placement look worse than it is. See Time to Conversion: Checking Your Attribution Window. Know your target ACoS, per campaign or portfolio. Without it there is nothing to compare against. See Set a Target ACoS Across a Whole Portfolio. 1. Pull the evidence Amazon Ads → Placements, with the window set. Placements, campaign by placement The table gives one row per campaign per placement. Export it, or work on screen with the campaign filtered. See Placements: Top of Search vs Rest of Search. Work campaign by campaign. Comparing Top of Search in one campaign against Product Pages in another tells you nothing, because everything else differs too. The comparison is only valid inside a single campaign. 2. Score each placement For every campaign, write down four numbers per placement: | | Why | |---|---| | Clicks | is this even decidable? | | ACoS | against target | | Conversion rate | the mechanism | | RPC (revenue per click) | the cleanest single comparison | RPC is the number to sort on. ACoS blends price and conversion; RPC collapses them into one figure in your own currency, which is exactly what a bid decision needs. 3. Decide, one placement at a time | What you see | Read | Do | |---|---|---| | ACoS well under target, high RPC | under-buying a good position | raise the multiplier | | ACoS near target | correctly priced | leave it | | ACoS over target, low conversion | paying premium for weak traffic | lower it | | ACoS over target, good conversion | conversion is fine, the price is not | lower modestly, watch | | Under 100 clicks | not decidable | leave it and revisit | The fourth row is the one people get wrong. Good conversion with bad ACoS is a price problem, not a placement problem. Cutting hard there throws away traffic that was converting. 4. Move in steps you can attribute Change one campaign at a time, or a small set with the same shape. Move in increments of 25 percentage points, not from 0% to 200%. A large jump changes what you win as well as what you pay, and the result is uninterpretable. Do not change the base bid in the same week. If you change both, you cannot tell which did what — and the multiplier compounds on the new base, so the effect is larger than you intended. Write down what you changed and when. This is the whole reason next quarter's audit will be quick. 5. Wait, then read it back Give it two full weeks, plus the attribution window. Three days tells you nothing. Re-pull the same view for the new period and compare like with like: Did the placement's share of spend move as expected? Did ACoS move toward target? Did total campaign performance improve — not just that placement's? Shifting spend between placements can leave the campaign exactly where it was, which is a useful thing to learn. A quarterly rhythm Once a quarter is right. More often and you are reading noise; less often and you are leaving the setting to drift through a whole season. Also audit after any big change — a price change, a new hero product, a category shift, a Prime Day. Placement economics move with them. See Run a Post-Event Teardown. What not to do Do not set the same multiplier across every campaign. That is the guess you are trying to replace. Do not chase Top of Search unconditionally. It is the most expensive position on Amazon and it is not always worth it. The data says whether it is for you. Do not audit placements when the campaign has a different problem. If the whole campaign misses target on every placement, the issue is the base bid, the keywords or the listing — not the split. See Diagnose an ACoS Spike in Ten Minutes. Do not read a placement with forty clicks. Say "not decidable" and move on. If the Placements screen is empty It populates once your campaigns have placement-level reporting for the window. An empty screen means either the range is wrong or the campaign types you are running do not report placement — not that something is broken. See Placements: Top of Search vs Rest of Search. Related - Placements: Top of Search vs Rest of Search — the screen, in full. - Campaigns: Budgets, Placement and Audience Adjustments — where multipliers are set. - Set a Target ACoS Across a Whole Portfolio — the number you audit against. - Time to Conversion: Checking Your Attribution Window — why the window ends a fortnight ago. - Create a Sponsored Products Campaign — where the multipliers are first set. - Ad Type and Placement Mix — the benchmarks the multipliers should be serving. - Placement Bid Optimisation — the weekly routine behind the audit.