Amazon business reports: the weekly numbers habit
Everything you need to run an Amazon account already sits inside Seller Central, unread. This is a guide to the five reports that matter every week, how to read unit session percentage as your real conversion rate, and how to compress it all onto one page.
- Five reports cover ninety percent of weekly decisions: sales and traffic by ASIN, inventory, returns, advertising, and payments. Ignore the rest until these are a habit.
- Unit session percentage is your conversion rate. Sessions falling means a traffic problem. Sessions steady with conversion falling means a listing or price problem. The split tells you what to fix.
- Build a one page weekly dashboard with the same columns every week. Trends across weeks are the signal. A single week's numbers are mostly noise.
Most sellers run their Amazon business on feelings and their bank balance. The irony is that Amazon hands over more operating data than almost any retail channel in India. It just puts that data in reports nobody opens. This guide covers where the reports live, the five that matter weekly, and how to compress them into one page you can read in ten minutes every Monday.
Where the reports live
Everything sits under the Reports menu in Seller Central. Business Reports holds sales and traffic. Fulfilment reports cover FBA inventory and returns. The advertising console keeps its own reporting section. Payments has a dedicated dashboard. Five doors, one habit: open them on the same day every week, pull the same numbers, and write them down in the same place.
Report one: sales and traffic by ASIN
This is the core report. Set it to child ASIN view and you get, per SKU, the numbers that describe your entire funnel:
- Sessions. Unique visits to the listing. This is your traffic.
- Page views. Total loads of the page. One shopper visiting three times counts once in sessions, three times here.
- Units ordered. What actually sold.
- Unit session percentage. Units ordered divided by sessions. This is your conversion rate.
- Buy Box percentage. How often your offer held the Buy Box when the page was viewed. If this slips, every other number on the row is understated.
Unit session percentage is your CVR
Amazon does not label anything CVR, but unit session percentage is exactly that. It answers the only question that matters after traffic arrives: of the people who saw the listing, how many bought?
Read it as one half of a two by two. Traffic and conversion can each be rising or falling, and each combination points somewhere different:
| Sessions | Unit session % | Likely story |
|---|---|---|
| Falling | Steady | Visibility problem. Ranking loss, ad spend cut, or a competitor took your shelf. |
| Steady | Falling | Listing or offer problem. Price undercut, new negative reviews, images gone stale, or a Buy Box slip. |
| Falling | Falling | Compounding decay. Conversion drops fed back into ranking. Treat as urgent. |
| Rising | Falling | Wrong traffic. Broad ad targeting is pulling in shoppers who bounce. |
This split is how you spot listing decay early. A listing rarely dies in a week. Sessions erode a few percent at a time, or conversion drifts down after a competitor refreshes their images. Week on week comparison of these two numbers per ASIN catches the drift while it is still cheap to fix.
Report two: inventory
Pull FBA inventory levels and compute days of cover for each important SKU: units available divided by average daily sales. Anything under your replenishment lead time is an emergency, because a stockout does not just pause revenue. It hands your ranking to whoever stays in stock. Check stranded inventory in the same pass. Stranded units are paid for, stored, and invisible to shoppers.
Report three: returns
The FBA returns report lists what came back and why. Read the return reasons, not just the rate. A cluster of defective or not as described reasons is a product or listing accuracy problem that no amount of marketing fixes. Size and fit clusters in apparel point at the size chart. Returns data is the closest thing Amazon gives you to customer interviews at scale.
Report four: advertising
The ad console will drown you in metrics. Weekly, you need four per campaign: spend, sales, ACOS, and the search terms that consumed the most money. The search term report is where waste hides. Terms with high spend and zero orders should be negated in the same sitting. Everything deeper than this is optimisation work for a different day; the weekly job is catching leaks.
Report five: payments
The payments dashboard reconciles what Amazon actually paid you against what you sold. Weekly, scan for two things: fee lines that changed, and reimbursements you are owed for lost or damaged inventory. Sellers who never read this page fund Amazon’s errors silently.
The one page weekly dashboard
Build a single sheet with one row per week and columns for: total sessions, total units, blended unit session percentage, ad spend, ad sales, ACOS, returns rate, and days of cover for your top SKUs. Below it, a small table of your five most important ASINs with their individual sessions and conversion.
The discipline is sameness. Same columns, same day, same source reports. After six weeks the sheet starts talking. You will see sale events pull demand forward and the dip that follows. You will see a listing decay two weeks before revenue shows it. This weekly rhythm is the backbone of any serious Amazon India Account Management operation, whether you run it in house or an agency runs it for you.
Common misreads
- Comparing a sale week to a normal week. Event weeks distort everything. Compare event to event and normal to normal.
- Trusting conversion on tiny sessions. Ten sessions and two orders is not a twenty percent CVR. It is noise. Judge conversion only where session counts are meaningful.
- Reading parent ASIN data as the whole truth. Parents aggregate children. A strong variation can mask a dead one.
- Ignoring Buy Box percentage. If you held the Buy Box eighty percent of the time, your true conversion on owned traffic is better than the row suggests, and the fix is a Buy Box fix.
- Treating one bad week as a trend. Act on two to three week patterns, not single data points.
Make Monday the reporting day
Block thirty minutes every Monday morning. Pull the five reports, fill the one page sheet, and write two sentences: what changed, and what you will do about it. That is the whole system. The sellers who compound on Amazon are rarely the ones with the cleverest hacks. They are the ones who have read their own numbers every week for two years and can smell a problem a fortnight before it reaches the bank balance.