Sellers Global editorial revision · 16 October 2026. The original contribution was published under NobleAmz42. Sellers Global has replaced unsupported performance anecdotes, share-based accusations, and universal keyword thresholds with the workflow below. The prior version is archived and replies remain. Prepared with AI assistance; official sources were checked on the update date.

Use Brand Analytics to form and test a question

Brand Analytics can help you understand search and shopping behavior. It does not reveal every competitor's campaign, prove manipulation, or guarantee profitable keywords. Choose the dashboard that answers your question, preserve its metric definitions, and test a business change against your own results.

1. Confirm eligibility and select the dashboard

Amazon's Brand Analytics overview says access requires a Professional selling account and a Brand Representative role for a brand enrolled in Brand Registry. Check your actual role, selected brand, marketplace, and available dashboards before assuming that missing data mean there is no demand.

Use these distinctions from Amazon's dashboard guide:

  • Top Search Terms: search frequency ranks and information about leading clicked products for a query, including click and conversion shares.

  • Search Query Performance: query-level performance and your brand or ASIN's share across search stages.

  • Search Catalog Performance: performance associated with products across the search funnel.

Save the dashboard, view, date range, marketplace, and any filter with each export. An ASIN filter and a query's three top-clicked products are not the same as a complete list of the ASIN's traffic sources.

2. Keep share, rank, volume, and rate separate

A share describes part of a total within the dashboard's defined population. A rate relates two different event counts. Search frequency rank is a relative ranking; it should not be treated as a known search count or compared with a volume field as if the units were identical.

For a deliberately simplified example, suppose a report population contains 2,000 clicks and 200 purchases for a query. A product with 200 clicks and 40 purchases would have 10% of those clicks and 20% of those purchases. Its share ratio would be 2. That arithmetic does not prove twice the overall conversion rate of page-one competitors, and it does not establish profitability. Real dashboard definitions and attribution must be checked before using such a calculation.

Do not divide by a zero click share or interpret a rounded, very small denominator as a stable signal. Keep the available counts and repeated periods next to a percentage. If the dashboard does not provide a needed denominator, record that limitation instead of inventing it.

3. Convert an observation into a narrow investigation

  • Impression share changes: Check availability, relevant queries, and reporting scope. This does not establish a competitor's spend or your organic rank.

  • Click share weakens: Review the offer, imagery, price, and competing results. The exact cause remains to be investigated.

  • Purchases lag clicks: Check relevance, delivery, price, and product information. The pattern does not prove that a review or one page element caused it.

  • A term gains frequency rank: Check repeated periods and seasonal context. This does not establish a universal search-volume threshold or profit forecast.

These are hypotheses. A high share held by three products does not automatically mean a market is closed, and a low share does not prove that a keyword is inexpensive. Query relevance and product economics still matter.

4. Verify the offer before changing presentation

Check whether the SKU was buyable during the period and whether pricing or delivery changed. Compare the actual product with the listing. Do not copy a competitor's assets or introduce misleading title symbols because their click share looks strong.

Amazon's product image requirements include main-image and category guidance; sign-in may be required. Evaluate a proposed image against those requirements and the real product before testing it. A report does not override catalog rules.

5. Run a measured test using your own reports

Choose a relevant query or a specific offer issue, save the baseline, and change one material element where practical. Define the review period and what would justify continuing. For advertising, compare target and placement data alongside business results using the Sponsored Products framework. For spend decisions, calculate contribution after advertising.

Keep ad-attributed sales separate from Brand Analytics search events and total business orders unless you have reconciled the definitions. Similar-looking columns from different dashboards need not cover the same population or attribution window.

6. Report limitations and avoid accusations

A sharp change in share can justify checking the data and the customer-facing offer. It does not prove click farming, rank injection, a promotion, or an undisclosed ad campaign. Remove those accusations unless independent evidence supports a specific policy report.

Use repeated, comparable periods to decide whether an observation persists. Note sparse data, changed filters, unavailable products, and incomplete periods. If stock availability changed, use the stockout checklist before interpreting a share change as a content problem. A useful follow-up identifies the dashboard, field, marketplace, dates, and tested change rather than promising a “blue ocean” keyword from one threshold.