Amazon Product Launch Playbook: The First 90 Days

Key takeaways: A successful Amazon product launch is won mostly before day one: a listing that indexes and converts, stock for at least 90 days, a keyword list in three tiers and a review plan that stays inside Amazon's rules. The first 30 days are about indexing, first reviews and learning which search terms convert. Days 31–60 are about pushing rank on a short list of attainable keywords. Days 61–90 are about pulling ACoS down toward its steady-state target without losing the rank you paid for. Launch ACoS is high on purpose; the mistake is not planning when it comes down.

What a launch is, and what it is not

A launch is the period between a product going live and the point where its organic sales can carry a meaningful share of its total. During that window you are buying three things with advertising: data on which search terms convert, sales velocity that Amazon's ranking uses as a signal, and the first reviews that make the next sale easier.

A launch is not a sales target. Judging a new product by the ACoS of a mature one in its first month guarantees you will either give up too early or cut spend at exactly the moment rank was starting to move. Set launch-specific targets and exit criteria for each stage, and report the launch separately from the rest of the account so it does not distort the picture.

This Amazon product launch playbook is the plan we run for a new product with a realistic budget, a defensible margin and a listing that is ready. If any of those three is missing, fix it first.

Before day one: the four things that must be ready

A listing that converts. A launch sends expensive traffic to a page with no reviews. The page has to do more work than a mature listing, not less. That means a main image that wins the click on a crowded results page, six more images that each remove a specific doubt, bullets written as answers to real questions, A+ Content if you are in Brand Registry, and every attribute filled. If the images are not ready, delay the launch rather than launch with placeholders. Amazon Listing Images: The 7-Slot Gallery That Sells covers what each of the seven slots is for.

Stock for the whole plan. Work out expected units per day at the launch's peak, multiply by 90, add a buffer for the restock lead time, and make sure it is in FBA or on its way. Running out of stock in week six of a launch throws away the rank you paid for; recovering it costs more than the stock would have. We never scale spend into a stockout, and neither should a launch.

A keyword list in three tiers. Tier one: the two to five high-intent terms where you realistically want page-one rank by day 90. Tier two: ten to twenty supporting terms with lower volume that you can win faster. Tier three: the long tail for discovery. Every term should be verified as indexed 48 hours after the listing goes live. If a tier-one term does not index, fix the listing before you spend on it.

A review plan inside Amazon's rules. Enroll the ASIN in Amazon Vine if you are Brand Registered; it is the one allowed way to get early reviews on a product with none. Turn on the Request a Review button or an approved automation for it. Use product inserts only to say thank you and offer support, never to ask for a positive review or to steer unhappy customers away from reviewing. Anything else, including incentivized reviews, review swaps and "review services", risks the listing and the account. Amazon Review Strategy That Follows the Rules goes into what is allowed.

The 90 days at a glance

Days 1–30: index, learn, get the first reviews

The first month has three jobs: confirm the listing indexes for everything on the list, find out which search terms actually convert, and get the first handful of reviews on the page.

Week one. Verify indexing on all three tiers. Launch an automatic campaign with all four targeting groups split into separate ad groups or campaigns so you can read them, and a broad-match campaign on the tier-one and tier-two terms. Set daily budgets high enough that neither campaign runs out before evening; a launch campaign that caps at noon is learning from half a day. Bid to win impressions, not to hit a target ACoS. Set up the review request process and confirm Vine enrollment.

Weeks two to four. Read the search term report weekly. Any search term with two or more orders is a candidate for its own exact-match campaign. Any term with meaningful spend and no orders after several clicks is a negative candidate, though be slower to add negatives in a launch than in a mature account; the sample is small and the listing has few reviews. Watch conversion rate by search term, not just overall.

What ACoS looks like. High, and expected. With no reviews and no rank, you are paying full price for every sale. In many launches ACoS in month one is well above break-even, sometimes double the steady-state target. That is the cost of the data and the velocity. The question is not "why is ACoS high" but "is conversion rate reasonable for a product with this few reviews, and are the converting terms emerging".

Exit criteria for day 30. Indexed on all tier-one terms. At least a handful of reviews, ideally five or more, with a rating at or above 4.3. A list of search terms that have converted at least twice. Conversion rate that is not far below the category norm once you account for the review count. If any of these is missing, do not move to the rank push; fix the gap.

Days 31–60: push rank on the terms that matter

Now you spend deliberately. The goal of this stage is to move up the organic results on the short tier-one list, and rank is earned mostly by sales through that search term.

Structure. Move each converting tier-one term into its own exact-match campaign so you control its bid, budget and placement independently. Keep the broad and auto campaigns running at lower budgets for continued discovery. Add the terms that are now in exact campaigns as negatives in the broad and auto campaigns so they do not compete with themselves. Amazon PPC for Product Launches: Structure, Budget and Timing covers the campaign structure and budget split in more detail.

Placement. Top of search is where rank is won in most categories, because it is where the highest-intent clicks happen. Use a top-of-search placement modifier on the tier-one exact campaigns, and accept that it raises cost per click. Check the placement report weekly: if top-of-search converts at two or three times the rate of the rest, the modifier is doing its job.

Price and offer. A launch price a little below where you intend to settle, or a coupon that shows a badge on the results page, lifts both click-through and conversion during the rank push. Plan how and when it comes off; a price rise later is easier when reviews have accumulated.

Watch stock. This is the stage where units move fastest. Check days of cover weekly against the restock lead time. If cover falls below the lead time plus a two-week buffer, slow the push rather than run out.

Exit criteria for day 60. Page one or two organic rank on most tier-one terms. Conversion rate at or near the category norm. Review count growing at a steady rate through Vine and review requests. Organic orders visible as a share of total orders in the Business Reports, even if small. If rank has not moved on a term after four weeks of top-of-search spend, that term may be too competitive for now; narrow the list rather than keep paying.

Days 61–90: bring ACoS down without losing the rank

The last month is where many launches fail, in one of two ways. Either spend stays at rank-push levels indefinitely and the product never becomes profitable, or spend is cut hard the moment rank appears and the rank goes with it. The plan is a taper.

Bids. Reduce bids on the tier-one exact campaigns in steps of 10–15% every week or so, watching organic rank after each step. If rank holds, take the next step. If it slips, hold or restore. The goal is to find the lowest bid that keeps the position, not to hit a target ACoS on a date.

Negatives and waste. By now you have 60 days of search term data. Go through it properly: add exact and phrase negatives for the terms that spent without converting, tighten broad match where it leaks, and cut the auto campaign budgets to a discovery level of 15–25% of the total.

Placement modifiers. Reduce the top-of-search modifier before you reduce base bids, so you keep the cheaper placements that are converting.

Organic share. Track ad-attributed orders as a share of total orders for the ASIN. As rank holds and reviews grow, this share should fall while total orders hold or rise. That falling share is the launch working. If total orders fall as fast as ad orders, you are cutting too quickly.

Exit criteria for day 90. Rank held on the tier-one terms through the taper. ACoS inside or approaching the steady-state band for the product's margin. Organic share of orders rising week over week. Review velocity steady. A decision made about the next stage: scale, hold, or fix.

What the numbers should look like

No two launches are the same, and anyone who promises a specific rank or ACoS by a specific day is guessing. But the shape is consistent. ACoS starts high and comes down through the taper. Conversion rate rises as reviews accumulate. Organic share of orders starts near zero and climbs. Total orders should rise through stages one and two and hold through stage three even as ad orders fall.

If you want to understand which of ACoS, TACoS and profit to watch at each stage, ACoS, TACoS and Profit: The Amazon Metrics That Actually Matter explains how the three fit together. For a launch, TACoS on the ASIN falling month over month from day 60 is the clearest single signal that the plan is working.

Common launch mistakes

Launching before the listing is ready. Placeholder images, three bullets and no A+ Content mean every expensive launch click converts at a fraction of what it could. Delay a week rather than waste a month.

Judging launch ACoS against mature targets. Month-one ACoS is supposed to be high. Cutting bids in week three because ACoS is above break-even stops the launch before it has produced the data or the velocity it exists to create.

Spreading spend across too many keywords. Rank is won term by term. Fifty keywords at low bids produce no rank on any of them. Two to five tier-one terms with real budget behind them produce rank you can build on.

Running out of stock. The most expensive mistake on this list. Rank earned over six weeks can be lost in a fortnight of being unavailable, and the launch spend has to be repeated.

Cutting spend the day rank appears. Rank is held by continued sales velocity. A taper of 10–15% a week with a check on rank at each step keeps it; a 50% cut loses it.

Anything outside the rules on reviews. Incentivized reviews, review swaps and steering unhappy customers away from feedback are the fastest way to turn a launch into a suspended listing. Vine and review requests are slower and they work.

How WeSpark runs a launch

Our 90-day launch program follows the three stages above, with written exit criteria for each and the launch reported separately from the rest of the account so its ACoS does not distort the mature products. Before day one we build the seven-slot image set, verify indexing on the tiered keyword list, enroll Vine and check that stock covers the plan. The launch program is priced from $3,500 a month, and the written scope sets out the targets for each stage before anything is spent.

C1Launching a new product — the 90-day plana launch with clear exit criteria for each stage — and an honest view of why launch ACoS is high on purpose.

E10How we grow reviews within Amazon's rulesmore reviews per 100 orders and a rating at or above 4.3 — with nothing that puts your account at risk.

B7The 7-slot image set — what each image is fora gallery where every image has a job: the main image wins the click; slots 2–7 remove doubt and win the sale. This is also why we ask for raw photos from several angles.