Key takeaways:

  • Amazon revenue breaks down to four numbers: impressions, click-through rate, conversion rate and price. Nothing else moves sales.
  • Only one of the four is usually the binding constraint at any given time. Fixing the other three costs money and changes nothing.
  • Sessions and unit session percentage live in Business Reports. Ad impressions and CTR live in Campaign Manager. Search impressions and search CTR live in Search Query Performance if you have Brand Registry.
  • A one-point gain in conversion rate is multiplied by every session you already pay for, which is why page work usually beats bid work.
  • Diagnose before you spend. The order is traffic, then click-through, then conversion, then price.

Amazon sales come from one equation: impressions × click-through rate × conversion rate × price, with units per order as a small multiplier on the end. Sessions are what you get after click-through. Orders are what you get after conversion. If sales are flat, one of those four numbers is the bottleneck, and the fastest route to growth is finding which one before you touch a bid.

That sounds obvious. In practice most accounts respond to flat sales by raising budgets, which only helps if impressions were the constraint. If the real problem is a main image that nobody clicks or a page that nobody trusts, more traffic just makes the loss bigger.

Where each number actually lives

Each factor in the equation has one authoritative report. Pull them in this order.

Factor Where to find it What a problem looks like Starting playbook
Impressions Campaign Manager for paid; Search Query Performance for organic search impressions Impressions fall while bids and budget are unchanged D2 · Not enough impressions (traffic)
Click-through rate Campaign Manager by campaign and placement; Search Query Performance by query CTR drops on queries where impression share held D3 · Low click-through rate (CTR)
Sessions Business Reports, Detail Page Sales and Traffic by Child ASIN Sessions down more than roughly 15% week over week D2 · Not enough impressions (traffic)
Conversion Unit session percentage in the same Business Report Units per session falls while sessions hold D4 · Low conversion rate (CVR)
Price and order value Business Reports ordered product sales divided by units; your own promo calendar Average selling price slides from coupons, variation mix or a price match D24 · Low average order value (AOV)

Two notes on the data. Unit session percentage is units ordered divided by sessions, not orders divided by sessions, so multipacks and quantity buys inflate it slightly. And Search Query Performance covers search impressions only, so a drop there does not account for browse, deal pages or external traffic. Reports and field names change; check the current definitions in Seller Central before you build a model on them.

Why the order of operations matters

Here is an illustrative set of numbers for a single ASIN over 30 days. These are an example, not a benchmark.

  • 400,000 impressions
  • 1.0% click-through, so 4,000 sessions
  • 9% conversion, so 360 units
  • $35 average selling price, so $12,600 in sales

Now change one factor at a time by the same relative amount. Raising impressions 10% costs real money, because the extra impressions come from higher bids or wider targeting, and those incremental clicks usually convert worse than your current ones. Raising conversion from 9% to 9.9% costs a photographer and a week of A+ work, and it applies to all 4,000 sessions, paid and organic. It also lowers your cost per acquisition on every campaign at once, which is the part people forget.

This is the compounding argument for working on listing and creative alongside ads rather than in sequence.

A8Why ads, listings and creative work better togetherbetter pages mean cheaper clicks; cheaper clicks mean more sales; more sales mean higher organic rank and lower dependence on ads. One loop takes about a quarter for most accounts.

How to find the bottleneck in twenty minutes

Run the checks in the same order every time, so you stop arguing about causes.

Read this flowchart as text
  1. Sales flat or falling
  2. Pull Business Reports by child ASIN last 8 weeks, week over week
  3. Sessions down more than 15 pct?
  4. Traffic constraint check indexing, budget pacing, organic rank, eligibility
  5. Unit session percentage falling?
  6. Conversion constraint check price position, rating, Buy Box, stock, page content
  7. Average selling price falling?
  8. Price or mix constraint check coupons, promos, variation and multipack mix
  9. Click-through constraint check CTR by query and placement in Search Query Performance
  10. One constraint, one owner, one number to watch

Two rules make this reliable. Compare against the same ASIN's own history, not a category average you read somewhere, because conversion varies enormously by price point and category. And compare like periods: week 47 against week 46 and against week 47 last year, never a four-day stretch against a full week.

What each bottleneck actually needs

Traffic constraint. Check in this order: is the listing live and in stock, is it indexed for the terms you expect, is budget running out before evening, are bids losing top-of-search placement. Indexing and Buy Box problems are cheap to fix and account for more sudden traffic drops than bidding does.

Click-through constraint. CTR is driven by the main image, the price shown in results, the review count and star rating, the first 60 characters of the title, and whether a badge or coupon flash is present. Sponsored Products CTR is typically well under 1%, so judge movement against your own baseline rather than an absolute number. Test one variable at a time.

Images, A+ & Brand StoreManage Your Experiments: What to A/B Test First on AmazonManage Your Experiments split-tests titles, main images, bullets, descriptions and A+ Content on live traffic. What to test first and how to read the result.

Conversion constraint. The usual causes, in rough order of frequency: price moved relative to the top competing offers, rating slipped, you lost the Buy Box, the gallery does not answer the obvious objection, or A+ content is missing or generic. Conversion problems also show up as a rising ACoS even when nothing in the ad account changed, which is why ACoS alone is a poor diagnostic.

Strategy & growthACoS, TACoS and Profit: The Amazon Metrics That Actually MatterACoS measures ad efficiency, TACoS measures how much your sales depend on ads, and profit per unit decides whether either matters. How to read and act on each.

Price and order value constraint. Sales can fall while units hold if a coupon is live, if the cheapest variation took over the mix, or if you matched a competitor and forgot to unwind it. Check ordered product sales divided by units ordered, by child ASIN, before you blame demand.

Three mistakes that hide the real constraint

Reading the account total. Portfolio-level sessions and conversion average away the one ASIN that broke. Always diagnose at child ASIN level, then roll up.

Confusing clicks with sessions. Ad clicks and sessions are different measures from different systems. A visitor who clicks two ads in one visit is two clicks and one session. Use ad clicks to judge campaigns and sessions to judge the listing.

Fixing the factor you are best at. Ad teams find ad problems. Designers find creative problems. The point of running the equation in a fixed order is to remove that bias from the decision.

How we run this

Every engagement starts by scoring the four factors rather than assuming the brief we were given is correct. The audit output is a single statement: this is the constraint, this is the evidence, this is what we do in week one.

A2The growth equation — which service moves which numbera diagnosis instead of a guess. When sales are flat we can show you whether it is a traffic, click-through, conversion or margin problem — and which service fixes it.

D1Sales are down — where is the problem?one owner, one playbook and one number to watch — instead of a debate about whose problem it is.

Strategy & growthAmazon Account Audit: The 30 Things We CheckEvery engagement starts with the same audit: 30 checks across ads, listings, creative, catalog, inventory and unit economics, turned into a scorecard and plan.

QUESTIONS

Frequently asked questions

What is the Amazon sales formula in its simplest form?

Sales equal sessions times unit session percentage times average selling price. Sessions themselves are impressions times click-through rate. Working backwards from the sales figure through those four numbers tells you which lever is actually holding you back.

How much data do I need before the diagnosis is reliable?

For a mature ASIN, eight weeks of weekly data is usually enough to separate a trend from noise. For low-volume ASINs, use four-week blocks instead of weeks, because a handful of orders can swing unit session percentage by several points.

Can advertising fix a conversion problem?

Not directly. Better targeting sends more relevant traffic, which lifts measured conversion a little, but it cannot fix a price gap, a 3.6 star rating or a gallery that fails to answer the main objection. Ads amplify whatever the page already does, in both directions.

Which factor should I work on first if two look weak?

Work on the one with the largest multiplier effect on existing traffic, which is almost always conversion, unless traffic has collapsed for a technical reason such as a suppressed listing, lost Buy Box or a term that stopped indexing. Those are urgent and usually cheap to fix.