Pick the SKU with the best conversion rate and enough inventory to survive ninety days plus one lead time. Not the highest revenue, not your favorite. A hero SKU has to carry ad budget, ranking pressure, and a reorder, so choose the one whose margin survives launch level acquisition cost.
The short version
- Conversion rate is the selection criterion. Traffic is purchasable, conversion is not, so start where the page already works.
- Inventory cover is a hard filter, not a preference. A sprint that stocks out in week nine wasted the budget that created the demand.
- Margin must survive the expensive phase. You will pay above your comfortable acquisition cost while you buy rank.
- One SKU, genuinely. A sprint split across three products is three underfunded campaigns wearing one name.
- Write the stop line first. Decide now what result at day forty five would make you redirect the budget.
The instruction, and why it holds
Pick the product that already converts, then spend money on it. That sequence sounds obvious and almost nobody follows it, because the instinct is to pick the SKU that needs help, or the one with the most stock sitting in a warehouse, or the one the founder is proudest of.
The reason conversion comes first is arithmetic. Advertising buys visits. What a visit is worth is set entirely by the percentage of visits that become orders. Pointing budget at a SKU that converts poorly means paying full price for traffic and keeping a fraction of it, and doing that for ninety days at scale is how a launch budget disappears with a rank position to show for it and no profit.
The selection checklist
Run every candidate through these nine. A SKU that fails any of the first four is disqualified regardless of how well it does on the rest.
- Conversion rate above your catalog average. Done properly means comparing sessions to units over a period long enough to be stable, not a good week. If nothing in your catalog converts well, the sprint is a conversion project, not an advertising one.
- Inventory cover for ninety days of accelerated sales plus one full lead time. Done properly means modeling the sprint's expected sales rate rather than the current rate, since the whole point is to increase it.
- Margin that survives launch level acquisition cost. Done properly means running the numbers at a deliberately inefficient advertising cost, not at the efficient one you hope to reach in month three.
- A rating that is stable or improving. Done properly means checking the trend, not the average. A four point five sliding downward will keep sliding once volume increases, and volume amplifies whatever is causing it.
- A keyword set you can realistically win. Done properly means primary terms where the incumbents are beatable within the sprint, plus long tail terms that convert while the head terms are still being fought for.
- A supplier who can actually reorder. Done properly means a confirmed lead time and a quantity commitment, agreed before the sprint starts. Winning rank and then running out is the most expensive outcome in this entire page.
- A clean listing foundation. Done properly means images, copy, and A+ content already finished. Advertising a half built page pays retail price to display a page that is not ready.
- Room to grow order value later. Done properly means the SKU has a natural bundle, multipack, or variation path, so the wins from the sprint keep compounding after it ends.
- A written stop condition. Done properly means a defined checkpoint, a defined metric, and an agreed action if the metric misses.
Scoring your candidates
| Criterion | Pass condition | Weight |
|---|---|---|
| Conversion rate | Above catalog average and stable | High |
| Inventory cover | Sprint plus one lead time, confirmed | High |
| Margin at launch acquisition cost | Positive contribution at the inefficient number | High |
| Rating trend | Flat or rising | High |
| Keyword winnability | A realistic path on head or long tail | Medium |
| Reorder capability | Written lead time and quantity | Medium |
| Listing readiness | Images, copy, A+ complete before day one | Medium |
| Order value upside | Bundle or variation path exists | Low |
| Stop condition | Written and agreed before spend begins | Required |
If two candidates pass, take the one with better inventory cover. Stock is the constraint that cannot be fixed mid-sprint.
What most agencies will not tell you
The hero SKU decision is frequently made backwards, and the incentive explains it. The SKU with the most existing revenue produces the best looking report, so it gets chosen even when a smaller product with a better conversion rate would deliver more incremental growth per dollar. Ask whoever proposes a hero SKU to show you the conversion rate comparison across your catalog, and ask why the chosen one is the best use of budget rather than the safest thing to report on.
The second thing: a sprint on a product whose supply chain cannot support a reorder is worse than no sprint. You spend to build rank, the rank creates demand, the demand exhausts inventory, and the listing loses the position you just paid for. Sourcing capability is part of the SKU decision, not a separate department. We built our sourcing frameworks across more than 500 brands specifically because the growth plan and the supply plan are the same plan, and asking who handles reorders is a fair question to put to anyone bidding for your account.
Related answers
- How many SKUs to start with in a 90-day plan
- Top mistakes in 90-day Amazon plans
- What products to bundle for higher AOV
- Which product analytics to track weekly
- Amazon seller roadmaps and capital: the complete guide
If you want the hero SKU chosen from your own data rather than your instinct, that is where a free audit from Flapen starts.
