Business Reports is the part of Seller Central most sellers open first in the morning and understand least. It holds the sales and traffic numbers behind every listing decision: how many shoppers came, how many bought, and how often your offer was the one behind the Add to Cart button. It also holds a few definitions that trip up even experienced teams.
This guide covers what each Amazon business report shows, what the main columns mean in Amazon's own words, where the reports stop, and how to ask an AI client for the same data instead of exporting it. Amazon definitions were read in its Selling Partner API documentation and report schema on 10 November 2026.
In Seller Central, open Reports, then Business Reports. The page opens on the Sales Dashboard, a snapshot of today's ordered sales and units against earlier periods. In the menu beside it sit the reports themselves, in two groups: reports by date and reports by ASIN.
The data is the same underneath. Through the Selling Partner API, Amazon publishes it as one Sales and Traffic report that can be requested by day, week or month and by parent ASIN, child ASIN or SKU. Amazon lists it under the Brand Analytics role, so an application needs that role to read it. Every Amazon seller reporting tool that shows sessions and conversion gets them from this report, DataDoe included.
One row per day, week or month for the whole account. This is the Amazon sales report most people mean when they ask for "the numbers": ordered product sales, units ordered, total order items, sessions, page views, featured offer percentage and the two conversion percentages. It also carries units refunded, refund rate, units and orders shipped, and customer feedback counts.
Use it for trend questions. Did traffic fall or did conversion fall? Is this week behind the same week last year? It cannot tell you which product caused a change.
One row per product for the date range you choose. It carries the same sales and traffic measures per item, plus two share columns: the item's share of all your sessions and its share of all your page views.
It comes at three levels, and choosing the wrong one is the most common mistake:
The dashboard is the quick look: today so far, with comparison lines for earlier periods and filters for product and fulfillment channel. As an Amazon sales dashboard it answers "how is today going?" and little else. It has no fees, no ad spend and no profit, and you cannot join it to anything.
Amazon's report schema defines the columns precisely. The definitions are worth reading once, because several do not mean what their names suggest.
Because the top of the fraction is units, a shopper who buys four units in one session counts four times. On multipack-style purchases and consumables the unit session percentage can sit well above the share of visitors who bought anything. If you want "what share of visits ended in an order", read order item session percentage instead. When two people in a meeting quote different conversion rates for the same ASIN, this is usually why.
When another seller holds the featured offer, the sale goes to them, whatever your listing looks like. A falling unit session percentage with a falling featured offer percentage is a pricing or stock problem, not a listing problem. Read the two together. Our post on tracking Buy Box loss shows how to find the cause.
Sessions are counted per level. A shopper who opens a listing and clicks through three colours in one visit is one session for the parent and can be a session for each of the three children. So child sessions can add up to more than the parent's sessions, and a conversion percentage calculated from summed child rows will come out lower than the parent's own figure.
Neither is wrong. They answer different questions. Use the parent row for "how well does this listing convert?" and child rows for "which variation carries it?". Do not rebuild one from the other.
The same holds between the by-date and by-ASIN reports. The daily account row is Amazon's own total. It is not the sum of the product rows, because one visit that touches four of your products is one session for the account.
These limits are why "Amazon business analytics" tends to mean a weekly spreadsheet: a by-ASIN export, an ads export and a settlement file joined with lookups, redone every Monday.
DataDoe is the Amazon data layer for AI agents. It loads Seller Central, Vendor Central and Amazon Ads data into 130+ tables that an AI client reads over MCP, a developer reads through the REST API and a data team reads in BigQuery. The Sales and Traffic report lands in two of them, already joined to the things Business Reports leaves out.
.svg)
amazon_profit_by_sku_and_date): one row per SKU and child ASIN per day, with sales and units from shipped orders, settlement fees, cost of goods, ad spend and profit, next to the Sales and Traffic measures for that product. The parent_asin column lets you group variations.amazon_profit_by_date): one row per day for the account. It carries Amazon's own daily account totals from Sales and Traffic, such as total_sessions, total_page_views and avg_buybox_percentage, with B2B and B2C variants, beside the full profit and loss for the day.The table descriptions repeat the rules above so the AI does not have to guess:
asin_ are the per-child values summed for the day, kept separate on purpose.The daily fetch starts at 2am in the seller's local time. Data teams who want the report exactly as Amazon sends it can read it through the Raw BigQuery integration.
With the account connected to Claude, ChatGPT or another AI client, the Monday export becomes a set of questions:
The fourth and sixth questions are the ones Business Reports cannot answer alone, because they need traffic and money on one row.
DataDoe can also act on what the answers show. Listing and price changes are available as Actions. Actions stay off until you enable them, most AI clients ask before a change, and every change is logged.
Under Reports, then Business Reports. The first screen is the Sales Dashboard, and the by-date and by-ASIN reports are in the menu beside it.
A visit to your Amazon pages by one user. All activity within a 24-hour period is one session, however many pages the shopper views.
Units ordered divided by sessions, shown as a percentage. It is Amazon's standard conversion measure and counts units, not buyers.
The share of page views on which your offer was the featured offer, the one behind the Add to Cart button.
Through Amazon's API the Sales and Traffic report can be requested for up to two years back. Store it yourself if you need longer history.
DataDoe costs $97 a month with a 14-day free trial. Connect Seller Central, wait for the first load, and ask your AI client the question you would normally answer with a by-ASIN export and a pivot table. The page for Amazon sellers shows what else sits in the same schema.