Operations & Reporting

The Weekly Meeting

Most Amazon problems are not sudden. A product's conversion rate slips, its TACoS drifts up, its stock cover shortens — and none of it is alarming on the day it happens. Three months later it is a serious problem with no obvious cause. The weekly meeting is the routine that catches those while they are still small. One hour, every Monday, across every product. Watch the walkthrough Prefer to watch? One hundred and twenty-one seconds on the five checks and why their order matters, how to scale your response to the market, and the ending that makes the whole thing work. What it is for Find problems and trends early, while they are cheap to fix. Decide specific actions, and put time in the week to do them. Build a feel for your products and markets, which is the benefit nobody expects and everybody values most. After six months of this you know what normal looks like for each listing, and anomalies announce themselves. It is not a reporting exercise. The output is a task list with owners and dates, not a document. The five checks, in order The order matters, because each one changes how you read the next. | # | Check | Question | |---|---|---| | 1 | Sales | which way is this product trending? | | 2 | Stock against sales | will it run out? | | 3 | TACoS | is total spend inside its range? | | 4 | ACoS | is the advertising itself efficient? | | 5 | CVR, sessions, rating | is anything drifting underneath? | Work through products one at a time, best-selling first. Give each one attention in proportion to what it contributes. Your top product deserves ten minutes; your twentieth deserves thirty seconds. 1. Sales Look at daily units and revenue, this week against previous weeks. You are establishing direction, not judging a single number. Then, for leading products, look at the market. Check what competitors are doing. This comparison decides how hard you act: Your sales down, market down — not a problem. The category is soft; aggressive intervention would be spending into a downturn. Your sales down, market flat or up — this is the one to act on decisively. You are losing share. Your sales up — make lighter changes. Something is working, and the risk is breaking it. Sales performance, week by week 2. Sales against stock Which products are heading for a stockout, and how soon? The answer sets your advertising decisions for the week, because a product about to run out should be slowing down, not scaling. Scale your response to how close it is: 45 days out — ease off slightly. 14 days out — cut hard. Take TACoS down to near zero. The levers, in order: stop deals, cancel or reduce outside traffic and promo codes, lower advertising spend, raise the price. Act fast here. Every day at full velocity on a constrained product is a day of stock you needed later. See When You Are Going Out of Stock and Monitoring Inventory Against Sales. 3. TACoS against its range Inside the range: nothing to do. Above the range: work out which channel is responsible — advertising, outside traffic, deals, or promo codes — and reduce the worst performer rather than cutting everything. Deals and promo codes are usually the quickest to pause. Outside traffic can be reduced or trimmed to its best campaigns. Advertising is the slowest to move and the most valuable, so cut it last. Below the range: you are under-investing. Find ways to spend more — that is unused, budgeted growth. And adjust the range itself when circumstances change. Low stock justifies a temporarily lower range; a holiday period may justify a higher one. See Setting Target TACoS per Product. 4. ACoS against target Within about five points of target: fine. Above target, and TACoS is also high — the advertising is both inefficient and too large. Cut waste aggressively; it fixes both at once. Above target, but TACoS is normal — the spend level is right and the efficiency is not. This is an optimisation job rather than a budget one: review the portfolio, cut bids on poor targets, raise them on good ones, and test new targets. Block one to three hours per portfolio rather than trying to fix it inside the meeting. The meeting identifies; the block resolves. See Bid Optimisation: Six Rules and The Bleeder Reports. Expect a lag. Fixing ACoS often raises TACoS for a week or two as you test. Watch it in the following meetings rather than reacting again immediately. 5. Conversion, sessions and rating The slow-moving metrics, and the ones that reveal problems the others cannot. Conversion rate — falling means the listing, price or competition changed. Sessions — falling means rank or advertising changed. Rating — a drop here explains a fall in both. See Why a Rating Drop Costs More Than It Looks. You are comparing against the last few weeks, not against a target. A conversion rate of 9% is neither good nor bad; 9% when it was 13% a month ago is a problem. Do not diagnose in the meeting. Write "investigate conversion drop on X", assign it, and book the time. Ending the meeting properly This is the part that makes it work, and the part most often skipped. Every issue becomes a task with a named owner and a date. An observation with nobody's name on it does not get fixed. Filter ruthlessly. A weekly review generates more ideas than you can act on. Keep only what serves this week's actual priority — the TACoS problem, the ACoS problem, the stock problem. Everything else is a distraction wearing the costume of productivity. Block the time. "Optimise the portfolio" without a calendar entry does not happen. The part that used to take longest The traditional version of this meeting is mostly assembly: exporting profit data, exporting advertising data, pulling ACoS per portfolio by hand, fetching deal fees, adding outside traffic costs, and reconciling it all per parent ASIN in a spreadsheet before anyone can look at anything. That is an hour of work to produce a view, before a single decision gets made — which is why the meeting is the first thing dropped in a busy week. In TrackIQ the view already exists. Sales, advertising spend, ACoS, TACoS, conversion, sessions and stock sit together per product over a date range you set. See Understanding Business Reports, Understanding Your TrackIQ Dashboard and Weekly Business Reports and AI-Generated Insights. The hour is then spent deciding rather than assembling, which is the only part that ever produced any value. If you do not use this exact process Whatever you run instead should cover three things: Stock against sales, weekly, so no product runs out unnoticed. TACoS, so you learn about overspending now rather than in a monthly profit report. ACoS per product or portfolio, so advertising problems are caught before they damage the flywheel. Related - The Nine-Report PPC Routine — the advertising work this meeting schedules. - Setting Target TACoS per Product — where the ranges come from. - Monitoring Inventory Against Sales — the stock check, in detail. - The Four Weekly Checks — the checks that make up the meeting, in order.

Escalating an Amazon Case

Seller Support resolves most things. When it does not, the difference between a problem fixed in two days and one that drags for three weeks is knowing the escalation ladder and working it in order. Three rungs, each with its own trigger and its own timeframe — plus one case where you skip all of them. Watch the walkthrough Prefer to watch? One hundred and forty-three seconds on building a case that can be actioned, escalating by phone with a script, and knowing when severity justifies skipping the ladder. Rules that apply throughout Start at step 1 unless it is genuinely urgent. Account suspension or the removal of a revenue-critical ASIN justifies using every route at once. Almost nothing else does. Document every action. Case IDs, dates, who you spoke to, what was promised. Move up when a step fails inside its timeframe — not before, and not much after. Stay factual, concise and respectful. Every rung is staffed by someone who did not cause your problem, and the tone materially affects whether they escalate for you. Step 1: open a case Always the starting point. Right for listing errors, wrong category assignments, variation and linking problems, catalogue issues and policy questions. Seller Central → Help → Get Help and Resources. Pick the closest issue, or "My issue is not listed" to describe it yourself. Choose the correct category — product detail page, account health, fulfilment — because it decides which team receives it. A misrouted case fails slowly. Include everything, first time: ASINs. Any existing case IDs. A timeline of events. Screenshots. Feed files and error messages. The exact resolution you want. That last one matters more than people expect. "This is wrong" invites a generic reply. "Please change the browse node from X to Y" gives someone something to action. Expect a first response within 24 hours. Two parallel routes worth knowing Advertising and campaign manager issues go through the ads console support, not Seller Central. Raising them in the wrong place wastes a day. Catalogue issues should also be raised through Brand Registry support, in parallel with the Seller Central case. Two teams, two chances. If you have a strategic account manager, email them as well as raising the case. Not instead — the case number is what everything else references. If it stalls Reply once in the same thread with clarification or new evidence. If nothing moves after 24 hours or two exchanges, go to step 2. Step 2: phone When the case has stalled, been closed prematurely, or the associate has misunderstood the request. Also when time matters — a suppressed ASIN or rank falling because of a catalogue error. Seller Central → Help → Manage Support Cases, select the case, reply. Some cases only offer email; when that happens, open a new case and reference the previous case ID. On the call, do three things: Summarise briefly and give the case ID. Ask explicitly for escalation to a higher-level specialist or manager. Ask them to document the notes in the case and keep it open until resolved. A case closed as resolved when it is not resets your progress. A script that works Opening: "I'm calling to escalate case 1234 regarding ASIN B09Z4X4383." Issue: "This ASIN is assigned to the wrong browse node — Boat Tie Downs. It's a Boat Trailer Guide-On and belongs under Sports & Outdoors → Boating & Sailing → Boat Trailer Accessories → Boat Trailer Guides & Rollers." Request: "I've already submitted a flat file and a category change request and the node hasn't updated. Could you escalate to a higher-level catalogue specialist and note that we're requesting a browse node correction from X to Y?" Close: "I'd like the case kept open until the update is confirmed. Could you document these notes in the case?" The pattern: identify, state the correct answer, ask for a specific escalation, ask for documentation. Expect a follow-up by email within 24 hours of the escalation. If it still stalls Go to step 3. And on a critical issue, keep contacting support daily — you reach a different person each time, and outcomes genuinely vary with who picks it up. Step 3: public escalation When the case is stuck and phone support has produced nothing in 24 hours. Amazon's seller help account on X can be reached by direct message or a public post. Both create a timestamp and often nudge internal routing. Include: the case ID, the ASINs, one or two sentences of summary, and the specific action requested. Use your brand account or a dedicated support handle, and keep the tone factual. This is a routing mechanism, not a complaint channel — a measured post gets picked up, an angry one gets ignored. When to escalate immediately Skip the ladder and use every route at once for: Account suspension. Removal or suppression of a revenue-critical ASIN. Anything actively costing significant daily revenue. In those cases the cost of waiting exceeds the cost of over-escalating. Everywhere else, working the ladder in order is faster, because each rung expects the previous one to have been tried. Which means the judgement call is really about severity, and severity is a question about revenue rather than about how annoying the problem is. Issues ranked by severity, not by when you noticed them Ranking your open problems by what they are actually costing per day is what tells you which one earns a phone call today — see Alert Categories, Priorities and What Actually Triggers Them. Working from an inbox of Amazon notifications instead sorts your problems by arrival time, which has nothing to do with their cost. Common mistakes Escalating before step 1 has had its 24 hours. Vague requests that give nobody an action to take. Miscategorising the case, so it sits with the wrong team. Raising advertising issues in Seller Central, or catalogue issues without also using Brand Registry. Letting a case be closed before the problem is actually fixed. Losing the case ID, which every later step needs. Related - Common TrackIQ Questions — problems that are not Amazon's to fix. - Triage a Rank Drop — diagnosing before escalating. - Find the Variation Dragging Your Rating Down — catalogue work that often needs a case.

Better Payment Terms From Suppliers and Freight

Amazon is a capital game. To do $10m a year you have to be holding around $2m in inventory — without it, the revenue is arithmetically impossible no matter how good the listings are. Which means growth is usually limited by cash, not by demand. But you do not necessarily need more money. You need more money that is free. Watch the walkthrough Prefer to watch? One hundred and fifty-five seconds on why growth is limited by cash rather than demand, the four arguments for supplier credit, and how to get Net 60 from a freight forwarder. Why the cycle strangles you Look at what happens to a dollar of inventory. You pay the supplier before production. Sea freight from Asia takes 45–75 days. Then Amazon holds your sales revenue roughly 30 more days after the sale. So you are funding that unit for something like four months before the money comes back. Every unit you want to sell more of extends that gap. The tactics below close it with supplier and freight credit rather than borrowing — which is the point. A loan at 10% a year is a real cost. Payment terms are free. Ask your existing suppliers for terms This is where the serious credit is, and it depends on relationship, volume, and how much your business matters to them. If you have worked with a supplier for years, the answer is usually yes. Negotiate on the messaging app they actually use, not email. Terms are a relationship concession, and relationships are not built in an inbox. The four arguments, in order of weight 1. Mutual growth. Credit is not charity — it lets you order a larger quantity, which lets you spend more on marketing, which sells more units, which makes the next order bigger. Say that explicitly: you are the constraint on my order size, and removing it grows your business too. 2. The payment cycle. Explain the four-month gap above. Most suppliers do not know Amazon holds seller funds for a month, and it makes the request look structural rather than opportunistic. 3. A yearly plan, not a purchase order. Tell them how many units you intend to order across the coming 12 months. The same request framed against an annual volume lands very differently from one framed against a single PO. 4. Social proof. Mention that your other suppliers already work on these terms. Real, but the weakest of the four — lead with growth, not with comparison. What to ask for Open with 25% deposit, 45% on ready date, 30% at 60–90 days after shipment. Then expand gradually as trust builds. Terms are a ladder, not a switch. Long term, push toward 100% at delivery. Some suppliers will agree once your reliability and volume are strong. And once you have real leverage, ask them to hold finished stock — or at least raw materials — and ship on demand without upfront payment. That collapses your lead time and your cash gap in one move. Then do the same with freight Freight forwarding is a commodity. Several providers offer effectively the same service at different prices, which is exactly the condition under which asking works. Get multiple quotes. The first purpose is simply not to overpay. Then choose on terms, not only on price. When two quotes are close, take the one with better payment terms and better communication. A slightly higher rate on Net 60 beats a slightly lower rate paid up front. After 3–4 months of consistent shipments, ask for Net 60 — paying 60 days after the shipment lands. Forwarders grant this far more readily than suppliers do, because the service is commoditised and they do not want to lose the volume. Q4 is the easiest ask of the year Ask for better terms on specific POs and specific shipments, not as a permanent change. And be direct about why: without the terms, you cannot place the order at that size. Both suppliers and forwarders expect this conversation in Q4, because every one of their customers is trying to fund a peak. A request scoped to one season and one order usually goes through smoothly. What this is worth Work out your own number before you negotiate. Take the order value, multiply by the extra days you are asking for, and you have the amount of free working capital the conversation is worth. Then decide what you would do with it. Free cash flow only helps if it goes somewhere — more inventory on your best-performing products, a bigger advertising budget behind a launch, or simply the ability to take money out of the business. Sizing your inventory need is the input to all of this — see Inventory Forecasting and Purchase Orders. Sales and inventory, per product Common mistakes Not asking. The overwhelmingly most common failure. These terms are standard for sellers who requested them and absent for sellers who did not. Asking by email, where a relationship request reads as a procurement demand. Framing the request around one purchase order instead of an annual volume. Choosing a freight forwarder on price alone, and giving up 60 days of free capital to save a small amount per shipment. Treating financing as the first move. Supplier and freight terms cost nothing; borrowing costs interest. Exhaust the free credit first. Freeing up cash and then not deploying it. Idle working capital earns you nothing. Related - Inventory Forecasting and Purchase Orders — how much stock the terms need to fund. - Where Your Margin Actually Goes — the cost side of the same problem. - PPC Payments and Cash Flow — the other free credit line.

PPC Payments and Cash Flow

Most sellers let Amazon charge advertising to their account balance every $500 and never think about it again. That default is costing you two things at once: the free credit you would get by paying later, and the cashback you would get by paying with a card. This is a one-time project. A week or two of admin, and it earns quietly in the background forever after. Watch the walkthrough Prefer to watch? One hundred and forty-five seconds on raising your billing increment, switching ad spend to a credit card, and sizing the cashback against your net margin. The two levers 1. Raise the billing threshold. Amazon bills advertising in increments — $500 by default, and it will raise that to $5,000 or $10,000 on request. Fewer, larger charges means each one lands later. 2. Pay by credit card instead of account balance. That adds your card's own statement period on top. Stacked, the two put roughly 60 days between spending on ads and paying for them — free credit, from money you were spending anyway. Why the threshold matters more than it looks At a $500 increment and $30,000 a month of ad spend, you are charged sixty times a month — effectively paying same-day. At a $10,000 increment, three charges a month. The average dollar of ad spend now sits unpaid for days rather than minutes, and that difference is working capital. Check your actual monthly spend first, because it decides which increment is realistic to ask for. Campaign spend, the number that sizes the ask How to request the increase 1. Go to the advertising campaign manager. 2. Open the help icon and choose Support Center. 3. Pick Billing and Payments, then Contact us. 4. Choose Payments and Invoicing → Payment frequency and credit ladder. 5. Start a live chat and make the request directly. A message that works: "Please adjust my payment ladder to $10,000. My credit card is [last four], limit [amount]. My Ads ID is [entity ID]." Your Ads ID is in the console URL — the value after entityId=. Your card limit must exceed the increment you are asking for. Requesting a $10,000 ladder on a $5,000 limit gets declined on the spot. And expect a ladder, not a leap. Amazon raises these gradually. If you are refused for low spend or a non-US account, keep spending consistently and ask again in a month or two. Paying by card Change the payment method in the campaign manager from account balance to a card. Then work out what the cashback is actually worth, because at real ad spend it is not a rounding error: | Monthly ad spend | Cashback rate | Extra profit per year | |---|---|---| | $10,000 | 2% | $2,400 | | $30,000 | 2% | $7,200 | | $30,000 | 3% | $10,800 | And read that against your margin. At a 10% net margin, $10,000 of cashback is worth the same as $100,000 of additional sales — about $8,000 a month more revenue you did not have to sell. That comparison is why this is worth a fortnight of admin. Very little else in the business produces that much profit for that little work. Choosing the card Compare candidates on the same six questions, so the answers line up: What is the spending limit, monthly and annually? What cashback rate, and in what form — cash, statement credit, or points? Can you add more cards if you outgrow the limit, and how many? Can it be linked to an online business account, or only a traditional bank? What are the foreign-currency fees, if you pay suppliers in other currencies? What is the statement and due-date cycle — this is the part that determines your actual float. Then negotiate. Tell the provider you have compared others and had better terms on a specific dimension, and ask them to match. Card terms are more negotiable than most sellers assume. Plan the limit before you need it Calculate how much ad spend the card can absorb at your current rate, and at the rate you expect during Q4. A card that hits its limit mid-month stops the whole arrangement — Amazon fails the charge, and you are back to account balance at the worst possible moment. So work out when you will need a second card, and apply before peak season, not during it. The wider version of the same idea Cashback on ads is one of four places money leaks on its way through the business: Cashback on card spend — covered above. Receiving payouts from Amazon — aim for zero fees. Paying suppliers, freight and team — keep transfer fees minimal. Currency conversion — can be as low as around 0.15%, and is frequently many times that by default. Audit all four, then fix the worst one. If only cashback is weak, a card solves it. If several are weak, the account you bank through is the problem. Common mistakes Leaving the $500 default in place, which is the whole cost of this article. Asking for a threshold above your card limit, and getting an automatic no. Treating a refusal as final rather than as a ladder you climb with consistent spend. Choosing a card on headline rate alone, ignoring the limit and the statement cycle. Letting the card hit its limit in Q4, exactly when ad spend peaks. Fixing the cashback and ignoring the conversion rate, which on an importing business is often the larger leak. Related - Better Payment Terms From Suppliers and Freight — the larger free credit line. - Where Your Margin Actually Goes — the rest of the cost stack. - Campaigns: Budgets, Placement and Audience Adjustments — where the spend numbers come from.

The Brand Manager's Calendar

Once every process in this training is running, the question stops being what should I do and becomes when do I do it. This is the calendar that holds all of it together. The shape is simple: a daily check that takes minutes, a weekly cycle built around one meeting, a fortnightly harvest, and a monthly planning block. Adjust it to suit you — but keep the cadences. Most of these processes fail not because they are wrong but because they get run irregularly. Watch the walkthrough Prefer to watch? One hundred and sixty-eight seconds on what belongs in the daily check, how the weekly cycle is ordered, and why monthly work needs blocked time. Daily — minutes, not hours Check each listing is live and correct: Buy Box held, main image right, price right, coupons as intended. Check account health. Check performance notifications. That is the whole daily job. It exists to catch the things that cost you money by the hour — a suppressed Buy Box, a listing gone inactive, a policy warning — see Escalating an Amazon Case. Everything else waits for the weekly cycle, because reacting daily to normal noise is how people spend their whole week optimising. The daily check, in one view A daily pass over Buy Box, price, images and account health across a catalogue is a five-minute job when it is one screen and an hour when it is a tab per product — see Alert Categories, Priorities and What Actually Triggers Them. Weekly — the meeting, then the work 1. The weekly meeting Review three things: last week's performance table, inventory levels, and keyword performance. Then decide what needs attention this week and which products you are pushing versus fixing. See The Weekly Meeting. 2. Urgent tasks and quick wins — immediately after Cut budgets on products selling too fast and heading for stockout. Raise budgets on products back in stock. Cancel deals where they no longer make sense. Adjust target TACOS up or down per product. Add or remove promotions and coupons. Do these in the hour after the meeting, while the decisions are fresh. They are small, they are quick, and they are the ones that get forgotten. 3. The weekly PPC work Bleeder reports 1.0 and 2.0 — see The Bleeder Reports. Bidding optimisation — see Bid Optimisation: Six Rules. Placement bid optimisation — see Placement Bid Optimisation. Deals for the coming period — see Running a Deal Calendar. MSRP renewals — see Adding a List Price Strike-Through. Fee review, including referral fee discounts still applying. Inventory level check, and ship more where needed — see Monitoring Inventory Against Sales. 4. Fix product issues Whatever time is left goes to the problems the meeting surfaced. This is the proactive part of the week — listing work, relaunches, ranking pushes, offer improvements. It is also the first thing to get squeezed, which is why it sits explicitly on the calendar rather than in the gaps. Fortnightly The harvesting report, and launching the campaigns it produces — see Search Term Harvesting: Criteria and Bids. Fortnightly rather than weekly, because search terms need time to accumulate enough clicks to be judged. Monthly — set aside one or two days Review your P&L and set target TACOS per product for the coming month — see Setting Target TACOS per Product. Inventory planning and sales forecasting — see Inventory Forecasting and Purchase Orders. Lifetime bleeder targets — see Lifetime Bleeder Targets. Ad type performance and ad placement performance reports, for the high-level view — see Ad Type and Placement Mix. Portfolio and account-level share of spend. Best Seller badge opportunities. Subscribe & Save trend. Block the time. Monthly work has no natural trigger, so it only happens if it is scheduled. Why the cadences are what they are Daily is for damage. Things that cost money by the hour. Weekly is for optimisation. Enough data to be readable, frequent enough to compound — and slow enough that you are not reacting to noise. See Optimisation as a Calculated Bet. Fortnightly is for anything needing accumulated clicks. Monthly is for decisions rather than adjustments — targets, forecasts, structural reviews. Put every one of these in a calendar and a task tracker. A process that lives only in someone's memory is a process that runs when they are not busy, which is never. Common mistakes Doing PPC work before the weekly meeting, and optimising products you were about to deprioritise. Skipping the quick wins, which are the cheapest actions of the week. Running harvesting weekly, on data too thin to judge. Letting monthly work slide, because nothing forces it. Treating the daily check as optional. It is the only defence against fast, expensive failures. Filling the whole week with optimisation, and never reaching the product work that actually grows the business. Related - The Weekly Meeting — the meeting the week is built around. - The Nine-Report PPC Routine — the reports inside the weekly block. - Setting Target TACOS per Product — the monthly decision everything else runs against.

Focus on Two or Three Products

Pushing a product forward costs money and attention, and both are finite. Spreading them across eight products means none of them gets enough of either. So if you are doing under about $20–30k a month, work on two or three parent ASINs at a time — even if you have twelve. And then do something explicit about the rest, which is the half people skip. Watch the walkthrough Prefer to watch? One hundred and fifty-five seconds on why the number is two or three, diagnosing which parents are bleeding and why, and the four options for everything you are not working on. Why the number is two or three More investment per product means a higher chance that product works. That is the whole argument. For scale: eight-figure sellers launching new products work on five or more at once — with a full team behind them, deep pockets, and a process they already know. Assuming a smaller operation can match that is not ambition, it is arithmetic that does not work. Once you have run three products through the full process and know it properly, duplicating it across the rest is much easier. The constraint is not the products; it is you. This only applies if you have more than three parent ASINs. With three or fewer there is nothing to choose. Step 1: find the best opportunities Put every parent ASIN through the comparison table — see The Product Comparison Table. That gives you the two or three strongest opportunities on offer strength, margin, review position and headroom. Do not commit yet. The next step frequently changes the answer. Step 2: find what is bleeding Open your P&L, make sure landed costs and advertising costs are complete, and look at last month — or month-to-date if you are past the 15th. Group by parent ASIN, sort by profit, and list every parent that lost money. Then work out why, for each one. There are only two answers: TACOS too high — you are overspending on advertising relative to what the product returns. Gross margin too low — and then the follow-up: is the price too low, or is COGS too high? Profit by parent ASIN, with the reason visible Grouping by parent, sorting by profit and reading TACOS and gross margin beside it is the whole diagnostic — see Understanding Business Reports and Where Your Margin Actually Goes. Reconstructing a per-parent P&L from settlement reports and ad exports every month is exactly the task that stops happening in a busy quarter. Step 3: choose You now have both halves: where the opportunity is, and where the money is going. Pick on two criteria: Highest likelihood of success. Highest upside if it works. The comparison table's winners usually get picked — unless there is an inventory situation that makes pushing one impossible. There is no point launching a product you cannot keep in stock; see Deciding Your PPC Daily Budget. Step 4: deal with everything else This is the step that actually frees the resources, and it is the one that gets skipped. For every product you are not focusing on, do one of four things: Cut most of its advertising spend. Pause its advertising entirely, so it cannot lose money. Raise its price until it is at least breaking even — see Raising Your Price Safely. Discontinue it, if the opportunity is weak and cash is tight. The rule underneath all four The only products allowed to lose money are the ones you are actively working on. A loss on a focus product is an investment with a plan attached. A loss on an ignored product is just a loss. The exception is a product you have a clear plan for — a relaunch scheduled, a new variation coming, a supplier change landing. Then a deliberate loss is fine. What is not fine is a product quietly losing money because nobody looked. Common mistakes Choosing focus products from the comparison table alone, without checking which products are bleeding. Picking four or five because they all look promising. Doing steps 1–3 and skipping step 4, so the resources you were freeing up are still being spent. Ignoring the inventory position on a product you intend to push. Discontinuing too fast, before checking whether price or COGS explains the loss. Never revisiting. When a focus product stabilises, it leaves the list and something else comes on. Related - The Product Comparison Table — the opportunity half of the decision. - Where Your Margin Actually Goes — diagnosing why a product loses money. - Setting Target TACOS per Product — how the non-focus products get held at breakeven.

Onboarding a VA

Most of the routine work in this training — the reports, the exports, the filtering, the pausing — is delegable. The reason it usually is not delegated well is onboarding, not capability. A VA who understands the tasks but not the business makes locally correct decisions that are globally wrong. Pausing a keyword you were holding for rank. Cutting a budget on a product mid-launch. So the onboarding has to cover the business, not just the process. Watch the walkthrough Prefer to watch? One hundred and eighty seconds on why business context comes before task training, the access to grant before day one, and keeping judgement while delegating assembly. Before day one Decide the hours — per week or per month — and be specific. Decide how you will track them. A simple shared timesheet works: daily start and end times, tasks completed, hours totalled. Whatever you use, you want to be able to cross-check hours against deliverables weekly or monthly, otherwise you have no idea what you are buying. Decide payment terms and method before the first invoice is due, not after. And decide whether you want a contract. That is your call; there is no universally right answer. Step 1: the introductory email Keep it short and unambiguous. Three things: Exactly how many hours they are engaged for. Payment method options, asking them to confirm one. The address to send invoices to. Getting these agreed in writing before the first call removes the entire category of misunderstanding that damages these relationships early. Step 2: the introductory call This is the step that decides whether the arrangement works, and it is much more than introductions. 1. The business Who you are and the story behind the brand. Your catalogue — main products, best sellers, what is launching. Your competition — what differentiates you, and the real challenges in your market. This is the context that turns a task-follower into someone who can flag a problem. Skipping it is the single most common onboarding failure. 2. The role What they are responsible for, and what success looks like in each area. Walk through the processes relevant to their role, rather than sending links. Invite questions. A VA who is uncertain and does not ask will guess. 3. Hours and availability Total hours, time zones, working hours, and expected response times. And a regular check-in schedule — daily or weekly updates, and a recurring meeting. 4. Communication and tools Your communication channel. Your task tracker. Your time tracker. What reporting you expect, and how often. 5. Next steps Recap hours, payment and reporting. Confirm what access they need. Name their first tasks, so they finish the call with something to start on. Step 3: Seller Central access Invite them with the permissions the role actually needs: Advertising Console — Campaign Manager. Business Reports. Manage All Inventory — not just FBA. This trips people up constantly. Add Products via Upload. Brand Analytics. Orders, including Create MCF Order, if fulfilment work is in scope. Grant these before their first working day. A VA waiting on permissions is a VA you are paying to wait. What to delegate first Start with the reporting work, because it is well-defined, repetitive, and its output is reviewed by you before anything is actioned. Good early candidates: Bleeder reports — pulling and filtering — see The Bleeder Reports. Lifetime bleeder targets, monthly — see Lifetime Bleeder Targets. Search term harvesting exports — see Search Term Harvesting: Criteria and Bids. Inventory tracking and forecasting inputs. Reimbursement claims. Buyer messages. A+ content updates and storefront work. The pattern: the VA produces the report and marks recommendations; you approve; the VA executes. Keep the judgement with whoever understands the strategy until they have earned it — see Optimisation as a Calculated Bet. What the VA produces, and what you approve Most of what you are delegating is assembly — export, filter, sort, mark — and the smaller that assembly step is, the more of the hours go to judgement instead — see Rule-Based Optimization: Automating Bids and Budgets. It also changes what you can hire for. Set the reporting schedule explicitly Build a schedule that maps each report to its cadence — daily, weekly, fortnightly, monthly — and put it where both of you can see it. It should match your own calendar — see The Brand Manager's Calendar. How long each process takes depends on your catalogue size, data volume and product count, so measure the first few runs rather than estimating. That is also how you find out whether the hours you have agreed are realistic. Common mistakes Onboarding the tasks without onboarding the business. Sending process documents instead of walking through them. Granting FBA-only inventory access, then wondering why uploads fail. Leaving payment terms until the first invoice. Delegating judgement before context — pausing and bidding decisions on day one. No time tracking, so you cannot tell whether the hours match the output. No fixed check-in, so problems surface a month late. Related - The Brand Manager's Calendar — the cadence the reporting schedule mirrors. - The Nine-Report PPC Routine — the reports most worth delegating. - In-House PPC or an Agency — the wider version of the same decision.

TACOS and ACOS Range Monitoring

Two numbers get checked every week, per product, and in this order: TACOS first, then ACOS. TACOS is the one that guarantees your profit. It is the share of total sales you are spending on advertising, and holding it inside a range you chose is how a monthly profit target actually gets hit rather than hoped for. ACOS tells you whether the advertising itself is efficient. Which matters, because 75–100% of most Amazon businesses' marketing spend goes through it. Watch the walkthrough Prefer to watch? One hundred and eighty-two seconds on why TACOS is a range and gets checked first, the arithmetic for correcting a week, and the bleeder sweep when both numbers fail. TACOS: hold it in a range, not at a number Set a monthly target TACOS per product from your profitability review, based on the product's profit and the stage it is at — aggressive for growth, tight for profit. See Setting Target TACOS per Product. Express it as a range of two or three points, not a single figure. 7–9%. 10–12%. 12–15% if you are pushing. Then each week, check last week's TACOS against that range. In range → do nothing. Move to the next product. Above range → reduce spend. Below range → increase spend. That last case is the one people ignore, and it is a real miss. Target 10–12% and spending 7.8% means you have profit budget you are not using — see Optimisation as a Calculated Bet. Correcting weekly is what makes the month land where you planned. Check monthly and you find out afterwards. The arithmetic, worked Target 12–14%. Last week: 16%. Sales $7,000. Spend $1,100. Aim for 13.5%, just inside the range — 13.5% of $7,000 = $945. $1,100 − $945 = $155 to remove from the week. Divide by seven: about $22 a day. Now decide where it comes from — the campaigns with the worst ACOS, or the outside traffic, promotion or deal spend that is performing worst. Whichever is returning least. Going the other way, if you are under range, you have options beyond bids: raise budgets on campaigns hitting their cap, raise bids or placement adjustments on your best targets, test new targets, campaigns or ad types, or add a deal, promotion or outside traffic. Spend against sales, per product, week by week Doing this per product every week means holding target range, actual TACOS, spend and sales side by side for the whole catalogue — see Understanding Business Reports. Rebuilding that from an ads export and a sales report each Monday is most of why the check gets skipped. ACOS: then check efficiency Once TACOS is handled, look at PPC ACOS per product against its target. A common general target is 30–35%, but higher-priced products legitimately run higher. Use your own number. Within ±5% of target is fine. It will never sit still, and chasing normal variation wastes the week. More than 5% above target is worth addressing. Target 30% and running 37–40% means recovering five points would be a substantial win. Mark the one, two or three products where that is true, and work on those. When both are wrong at once High TACOS and high ACOS together is the situation that justifies aggressive action. Because both of the main advertising metrics are failing, and the cause is usually the same: spend landing on targets that do not convert. Run the full bleeder sweep — high-click zero-sale targets, targets converting above your threshold, and the lifetime pass. See The Bleeder Reports and Lifetime Bleeder Targets. That attacks both numbers from the same side: cutting the waste lowers spend, which lowers TACOS, and removing poor targets lowers ACOS. Why TACOS comes first ACOS measures the advertising. TACOS measures the business. A campaign at a beautiful ACOS on a product whose TACOS is 20% is not saving you — you are still spending a fifth of revenue on advertising. And the reverse: a product comfortably inside its TACOS range does not need its campaigns interfered with, however imperfect they look. So TACOS decides whether to act. ACOS decides what to do. Common mistakes Setting a single target rather than a range, and correcting constantly against noise. Only acting when TACOS is too high, and leaving profit budget unspent when it is low. Checking monthly, so the month is over before you can influence it. Chasing ACOS variation inside the ±5% band. Using one ACOS target across a catalogue with very different price points. Cutting spend without deciding where from, and taking it off the campaigns that were working. Related - The Weekly Meeting — the meeting these two checks sit inside. - Setting Target TACOS per Product — where the range comes from. - The Bleeder Reports — what to run when both numbers are wrong.

The Four Weekly Checks

The weekly meeting works because it puts the same numbers in front of you every week — and the value is entirely in the comparison, not in any single week's figure. Five weeks in, you know what normal looks like for each product. A drop that would be invisible on its own becomes obvious against the run of weeks beside it. These are the four checks, in the order to run them. Watch the walkthrough Prefer to watch? One hundred and seventy-five seconds on reading each metric against its own recent history, checking the market before judging a sales trend, and only tracking what you will act on. Start with the 80/20 Work your best products first, while you are fresh. They deserve the most attention and they get it least, because a meeting that runs top to bottom alphabetically spends its energy on products that do not matter. Check 1: sales trend Weekly sales, daily sales, revenue. Read the trend against the previous weeks, not the absolute number. Then check the market, for your leading products Search your main keyword, look at the top and direct competitors, and judge whether the market overall is normal, up, or down. This is what makes the sales trend interpretable. Sales up 10% while the market is up 100% is a bad week wearing a good number. After ten or fifteen weeks of doing this you will know your competitors' normal ranges by eye, which is when the check gets fast. Then act on what you find Trend positive → change less. If it is compounding, let it compound. Confirm ACOS and TACOS are in range, confirm inventory is adequate, and keep whatever is working running — deals, outside traffic, promotions. Trend negative while the market is normal → act now. That is your product losing ground rather than the category softening. Your options, in rough order of speed: raise TACOS and push it through advertising, run a deal, drive outside traffic, start a promotion. Then, deeper: check whether ACOS has deteriorated, test new campaigns, or work on the listing itself. Sales trend per product, against previous weeks The whole method depends on last week sitting next to the previous ten, per product — see Understanding Business Reports. A single week's export tells you a number; the run of weeks tells you whether to act. Check 2: sales against inventory Compare the sales pace to the cover you have left — see Monitoring Inventory Against Sales. Selling faster than planned is only good news if you can supply it. This is where budgets get cut on products heading for a stockout and raised on products back in stock — see Going Out of Stock. Check 3: sessions and conversion Both from Business Reports, both read against the previous weeks. Sessions. Five weeks running at 5,000–6,000 a week, then suddenly 3,000, is a dramatic change that has to be investigated this week. Block time for it. Look at the market, the listing, the back end — something specific caused it, and a fall that size will show up in sales immediately. Conversion. Five weeks at 65–75% in Business Reports, then 55%, is a ten-point drop and worth one to two hours of dedicated investigation. Business Reports conversion is not your true conversion rate — it is calculated its own way. That does not matter here, because you are comparing it against itself. Check 4: rating Watch the trend, not the value. 4.5 → 4.4 is alarming, even though 4.4 is a perfectly good rating. The direction is what matters, because a trend continued takes you to 4.0, and at that point the product is in serious trouble. Address negative rating trends the week you see them — see The Rating Recovery Action Plan. Only track what you will act on You can add as many columns as you like. Add only the ones that will change what you do. A metric you look at and never act on costs you time every single week, forever. Why the weekly cadence is the point Every one of these checks is a comparison against recent history, and that only exists if you look at the same numbers on the same schedule. The output is tasks. Each check that fails produces a specific investigation with time blocked against it — see The Brand Manager's Calendar. That is what makes the meeting powerful: you find negative trends within a week of them starting, rather than at the end of a quarter. Common mistakes Reading this week's number without the previous weeks beside it. Working the catalogue in order rather than starting with your best products. Judging sales without checking the market, and misreading a category-wide move as your own. Investigating a session drop next week. A 40% fall is this week's job. Reacting to a rating value rather than a rating trend. Adding columns you never act on. Leaving the meeting without tasks. The checks are only useful if they produce blocked time. Related - The Weekly Meeting — the meeting these checks make up. - TACOS and ACOS Range Monitoring — the two advertising checks that follow these. - Triage a Rank Drop Across Your Top Keywords — where a session drop investigation usually goes next.