Five numbers matter in launch week: sessions, unit session percentage, primary image click-through rate, advertising cost of sales, and review velocity. Revenue is a lagging summary of those five. Track them daily, compare each against the assumption you wrote before launch, and act on the gap rather than on the total.
The short version
- Revenue in week one tells you almost nothing. It is the output of five inputs, and only the inputs are actionable.
- Write your expected number before launch. A metric without a prior expectation cannot be judged, only admired.
- Conversion is the gate. If unit session percentage is weak, more advertising spend buys more of the same disappointment.
- Daily cadence for seven days, then weekly. Launch week is the only period where daily reading is worth the noise.
- Two numbers are near useless in week one. Ranking position and total sales rank move for reasons you did not cause.
The buyer-side lesson
Before Flapen I ran data and technology at BRANDED and at Moonshot Brands, two large Amazon aggregators, and part of that job was receiving agency reports on brands we owned. The reports that arrived full of revenue charts were the ones I trusted least, because revenue told me what had happened without telling me what to do next. The useful reports were short, showed the same five inputs every week, and marked each one against the number we had agreed to expect.
That is the standard I would hold any partner to, and the standard worth holding yourself to in your own launch week spreadsheet.
The five that matter, compared
| Metric | What it actually tells you | A bad reading means | Wrong response |
|---|---|---|---|
| Sessions | Whether anyone is finding the listing | Indexing, keyword coverage or bid problem | Rewriting copy nobody has read yet |
| Unit session percentage | Whether the page persuades the people who arrive | Price, images, reviews or a targeting mismatch | Increasing budget to send more traffic |
| Primary image click-through rate | Whether the thumbnail wins the search results grid | The main image loses against the competing tiles | Editing the bullet points |
| Advertising cost of sales | What a customer currently costs you to buy | Bids, match types or a conversion problem upstream | Pausing all advertising in week one |
| Review velocity | Whether social proof is accumulating fast enough | Follow-up, insert compliance or Vine gaps | Anything that breaches Amazon's terms |
Sessions before anything else
If sessions are near zero, no other metric is readable. Check indexing on your main keywords, check that campaigns are actually delivering, and check that the offer is not suppressed. This is a mechanical problem with a mechanical fix, and it is the first thing to look at on day two.
Conversion decides the week
Unit session percentage is the number I would keep on screen. When traffic is arriving and the page is not converting, spending more is the natural instinct and the wrong one, because the extra spend buys extra evidence of a problem you already have. Fix the constraint, then buy traffic. Price, main image and the top three bullets are where most of the movement lives.
Two metrics that mislead in week one
Keyword rank moves on very small volumes early, so a jump to page one on a low volume term is not a signal. Total profitability is similarly unhelpful, because your launch advertising cost of sales is supposed to be uncomfortable while you buy rank, reviews and data. Judge profitability at maturity, not on day five.
What to compare each number against
A metric on its own is a fact. A metric next to an expectation is a decision. Before launch, write down the sessions you expect from your opening budget, the conversion rate typical of your category, the review pace your follow-up should produce, and the advertising cost of sales you consider acceptable for a launch phase. Then read the daily numbers as gaps.
For our own clients this is a written Slack update once a week, a live review every two weeks, and access in between, which is deliberately boring. The value is not the frequency, it is that the same five lines appear every time so a trend is visible without anyone reconstructing history.
What most agencies will not tell you
Launch week reporting is where dashboards get used as theater. Twenty charts arrive, the good ones are large, the bad ones are in a table at the bottom, and nobody states which number decides the next action.
The second thing: some of what you should be tracking is uncomfortable for the person reporting it. Return rate, negative review themes, and the gap between promised and actual timelines all belong in launch week reporting and all reflect on execution. Ask for them explicitly in the first report, and notice whether they arrive in the second one without being asked again.
Related answers
- What to use for tracking Amazon launch milestones
- KPIs an Amazon agency should report weekly
- Ideal launch day schedule for Amazon new product
- What does a good Amazon account audit include
- Amazon launch services: the complete guide
If you want the five lines read for your own listing before launch, that is the free audit at Flapen.
