Prime Day 2026 finished on 26 June. If your account is like most, the fortnight since has been spent catching breath, watching the returns wave come in, and quietly going back to normal. That's the standard move. It's also the expensive one.
At rankhouse we treat the two weeks after any major event as the highest-leverage fortnight of the quarter. The event itself is a spike. What you do with the data and the customers it generated is what compounds. Here are the ten jobs, in the order we run them across client accounts.
1. Cost the event honestly, per product
Before any celebration: what did each product actually earn during the event window? Deal price minus referral fee, minus FBA fee, minus deal funding, minus attributed ad spend, minus your landed cost. Per unit, per ASIN. Not an account-level blend, because blends are where losing products hide.
Every event we audit contains at least one "hero" deal that lost money on every unit while topping the revenue chart. You cannot make that call from a dashboard that averages everything together. This is the same discipline as everyday trading, which we've laid out in the case for per-ASIN TACoS, just applied to a four-day window.
| Week | Priority | Why it matters |
|---|---|---|
| Week 1 | Hold rank: keep ads on winners, do not slash budgets | The velocity you bought is still compounding |
| Week 2 | Harvest: review requests, search term mining, new exact campaigns | Event traffic reveals terms you could not see before |
| Week 3 | Debrief the deals: unit economics per promotion, not blended | Some deals bought profit, some bought expensive turnover |
| Week 4 | Re-forecast stock and cash into the next event | December is decided months out |
2. Mine the search term report while it's fresh
The event pushed a quarter's worth of traffic through your campaigns in four days. Inside the search term report is a list of queries that converted for you at scale, some of which you weren't deliberately targeting. Harvest them into exact match now, while the data is dense. This is the fastest keyword research you will do all year, and it's free.
3. Run the kill list
The same report shows the opposite: terms that took meaningful spend during the highest-intent shopping event of the summer and still produced nothing. If a term couldn't convert on Prime Day, it is not going to convert on a wet Tuesday in August. Negate it and redeploy the budget. We've written a full method for this in the kill list.
4. Chase the new-to-brand cohort
Check your new-to-brand percentage for the event window. Those customers bought you at a discount, which means the relationship starts underwater and only pays if they come back at full price. Three levers, in order of power: Subscribe & Save capture, a follow-up flow through Brand Tailored Promotions where eligible, and remarketing through Sponsored Display audiences. The economics of the first lever are strong enough that we've given it its own piece.
5. Let the returns wave settle before you judge anything
Event-driven purchases return at a higher rate than everyday purchases. The wave takes two to three weeks to fully land. Any verdict on event profitability written before it settles is provisional, and any agency that sends you a triumphant wrap-up on day two is reporting revenue, not money.
Want this level of thinking inside your account?
rankhouse is a boutique specialist agency: experienced Amazon professionals, no juniors, led personally by founder Dan Whalley. Fixed fee from £1,500 + VAT a month. Every engagement starts with a free profit audit built from your own data.
Get the free audit at rankhouse.co.uk →6. Re-forecast stock from actual run rates
The event either cleared more than you planned or less. Both change your Q4 order maths, and July is when those orders go in. Recalculate weeks of cover per product from post-event run rates, not pre-event forecasts. December stockouts are almost never December problems. They're July problems that stayed quiet, which is why we've covered the stock maths separately.
7. Read the competitor movements
Prime Day scrambles the search grid. Some competitors bought rank with deep deals and will fade. Some held rank without discounting, which tells you their organic base is strong. Screenshot your money keywords now and again in two weeks. The gap between the two pictures is a map of who bought their position and who owns it.
8. Bank the review opportunity
Every unit sold during the event is a review request opportunity on a timer. Requests convert best once the customer has actually used the product, which for most consumables means the window is open right now. A four-day sales spike should produce a visible review lift three to five weeks later. If yours doesn't, the machine is broken, and here's how to build a compliant one.
9. Test into the quiet
July and August are the flattest trading months of the year, which makes them the best testing months. Main image tests, title experiments, A+ variants. The traffic is calm enough that results are clean, and the winners are locked in before Prime Big Deal Days in October. A test started this week reports by late August. A test started in September collides with Q4 and reports too late to matter.
10. Write the October plan while the June lessons are warm
Prime Big Deal Days is expected around 7 to 8 October. Everything the June event taught you about your deal structure, your bid curve, your stock cover and your creative applies directly, and it is never better documented than it is right now. Write the plan this month. Future you, in the last week of September, will be very grateful.
Notice what none of these jobs are: none of them are about the next promotion. The fortnight after an event is analysis, harvesting and positioning. The brands that treat it that way turn one good event into a better quarter. The brands that go straight back to business as usual pay for the same lessons again in October.
If you'd like a second pair of eyes on your event data, the free audit we run at rankhouse covers exactly this: per-product event economics, the search term harvest, and the kill list, built from your own account data. It's the same diagnostic we run before taking on any client, and it exists because we'd rather show you the method than describe it.
Questions we get asked about this
How long after Prime Day should I wait before judging profitability?
Two to three weeks, because event purchases return at a higher rate than everyday purchases and the wave takes that long to land fully. Any profitability verdict written in the first week is a revenue report wearing a profit costume. The practical approach is two passes: a provisional read in week one, good enough to catch anything alarming and to start the search term work, then a final costed verdict once returns settle, which becomes the record you plan the October event against. Date both versions, because the gap between them is itself useful information about your return rates under discount conditions.
What exactly should I harvest from the search term report?
Two lists, pulled while the data is dense. First, converting queries you weren't deliberately targeting: these go into exact match with their own bids, because the event just proved demand and relevance at scale, and that evidence is worth more than any keyword tool's estimate. Second, high-spend queries with zero sales during the highest-intent shopping period of the summer: these go onto the kill list, because a term that couldn't convert on Prime Day has had the fairest trial it will ever get. Run both motions together and the account comes out of the event with permanently sharper targeting, which is a benefit that outlasts the sales spike.
How do I keep the new customers Prime Day brought in?
Accept that they bought you at a discount, which means the relationship starts underwater and only pays on the repeat. The three levers, in order of power: Subscribe and Save capture, because converting a deal buyer into a subscriber turns one discounted order into a scheduled annuity; Brand Tailored Promotions where your account is eligible, aimed specifically at recent first-time buyers; and Sponsored Display remarketing audiences to stay visible during the natural repurchase window. The economics only work if you know your repeat rates, so check Repeat Purchase Behaviour in Brand Analytics before deciding how much a new customer is worth chasing.
Is it worth running listing experiments in July and August?
It's the best window of the year. Traffic is calm enough that test results are clean, volumes are still sufficient for significance on most established ASINs, and, crucially, the timeline works: a main image or title experiment started in late July reports by early September, in time to lock the winner in before Prime Big Deal Days traffic arrives in October. Started in September, the same test collides with event noise and reports too late to matter. Brands treat summer as dead time; it's actually the laboratory the rest of the year's results come from.
When should I start preparing for the October event?
Now, and not rhetorically: the twelve-week countdown to an expected 7 to 8 October event starts in mid-July. The long-lead items are stock orders, which ride the same Q4 supply chain crunch as your Black Friday cover, and experiments, which need weeks to conclude. Deal submissions, creative and campaign structure follow through August and September. June's event data is the best preparation asset you own, and it's never better documented than it is right now, which is why writing the October plan is the last job in the post-event fortnight rather than the first job of autumn.
The next step is twenty minutes.
If any of this reads like your account, the fastest way to find out is the free audit: per-product profitability, where the ad spend is leaking, and what we would fix first. No pitch deck, no obligation.
Book a chat with Dan →
About the author
Daniel Whalley is the founder of rankhouse, a boutique specialist agency for Amazon-focused growth in FMCG, health, wellness and beauty brands. He has spent 10 years inside Amazon accounts, generating £100M+ for the brands he works with, and manages £500k+ a month in ad spend across the UK, Europe and the US. He writes from inside the accounts he runs, not from the sidelines. Connect on LinkedIn → · amazon@rankhouse.co.uk