Amazon operations

A 15-minute Amazon performance review

Use this short review to identify what changed, find the ASINs responsible, and leave with a small set of evidence-backed actions.

Collins Paris3 min read

A useful performance review ends with a decision. Touring every dashboard rarely does.

Fifteen minutes is enough when the report is prepared and the questions stay narrow. The purpose is to find meaningful change, decide what deserves investigation, and assign the next step.

Set it up once

Save a view or worksheet with:

  • Ordered product sales
  • Units ordered
  • Average selling price
  • Sessions
  • Unit Session Percentage
  • Featured Offer percentage
  • Current period
  • Previous comparable period
  • Same period last year when available
  • Absolute contribution to the sales change by child ASIN

Use complete periods. For a weekly review, the last seven complete days compared with the seven before them is a practical starting point. Add year-over-year context for seasonal products.

Amazon's Business Reports provide sales and traffic data by date and ASIN, including parent, child, and SKU views. Amazon documents the available groupings here.

Minutes 0 to 2: verify the frame

Before reading a percentage, ask:

  • Is the current period complete?
  • Are the weekdays comparable?
  • Did a promotion, event, stockout, or price change distort one period?
  • Are we looking at the same marketplace and ASIN level?

If the frame is wrong, stop. A fast review of mismatched data only produces a fast mistake.

Minutes 2 to 5: read the topline

Look at revenue, units, and average selling price first.

Revenue shows the business result. Units show product demand more directly. Average selling price explains part of the gap between them.

Then add sessions and Unit Session Percentage:

  • Sessions tell you whether traffic changed
  • Unit Session Percentage tells you how many units the traffic produced
  • Featured Offer percentage tells you whether your offer was attached to the primary purchase action

Do not investigate yet. Write down the one or two movements large enough to explain the period.

Minutes 5 to 9: find the products responsible

Sort child ASINs by absolute contribution to the change.

If total sales are down $40,000, find the ASINs that make up that $40,000. Ignore dramatic percentages on tiny products until the main movement is accounted for.

Label each major mover with the metric that best explains it:

  • Units
  • Average selling price
  • Sessions
  • Unit Session Percentage
  • Featured Offer percentage
  • Availability

The goal is a short list, usually no more than a few products.

Minutes 9 to 12: choose the first explanation

For each material ASIN, follow the metric:

PatternFirst check
Sessions down, conversion steadyDemand, organic placement, advertising traffic, seasonality
Sessions steady, conversion downOffer, price, delivery, reviews, listing experience
Featured Offer percentage downInventory, eligibility, total price, delivery, competitors
Units down, revenue steadyPrice or product mix may be hiding weaker demand
Parent steady, child volatileVariation mix or child-level availability

Write the explanation as a hypothesis if the evidence is incomplete. "Likely caused by lost Featured Offer share" is better than presenting a guess as a fact.

Minutes 12 to 15: assign the next move

Every issue that survives the review gets four fields:

FieldExample
ObservationSessions fell 22 percent on two child ASINs
Next checkReview organic placement and ad impressions
OwnerJordan
Follow-upThursday after three complete days

Limit the action list. Ten observations with no owner are less useful than two investigations that close.

What to leave out

The short review is not the place to:

  • Rebuild every report
  • Debate a one-day fluctuation without enough volume
  • Optimize campaigns search term by search term
  • Redesign a listing during the meeting
  • Explain every small ASIN movement
  • Change the comparison window because another one tells a nicer story

Those tasks can follow when the review shows they matter.

Keep a decision log

Over time, the log becomes part of the analysis. It tells you when price changed, when a listing was edited, why a promotion ran, and which hypothesis proved wrong.

The next review should begin with the open actions from the last one. Close the loop before adding more.

Fifteen minutes works because the review follows change instead of wandering through metrics. Find what moved, find where it moved, and decide who will learn why.

  • Business Reports
  • Sessions