Amazon Keyword Research Using Seller-Authorized Reports

Summary

Client-scoped Amazon keyword-research framework using seller-authorized search-term reports, eligible Brand Analytics, and supported product material. It excludes external Amazon-derived datasets, cross-seller pooling, fixed refresh promises, and undisclosed ranking formulas.

A client-scoped process for reviewing Sponsored Products search terms, eligible Brand Analytics reports, listing coverage, and documented follow-up decisions.

Written by Maksym Lazuto. This guide is part of the BFarm Academy and connects to services, proof pages, and the free audit flow.

Explore Services See Case Studies Read Academy Book a 30-min call Get the free audit

The practical starting point for Amazon keyword research is the seller's own Amazon reporting. Sponsored Products search-term reports and, where available, Brand Analytics show account-specific query performance without introducing data from another seller. The objective is to identify relevant shopper language, connect it to the listing, and document the source and reporting period for each decision.

This guide is an operating framework, not a statement about an undisclosed Amazon ranking formula. Amazon systems and report fields change. Current Seller Central evidence and Amazon documentation take precedence over fixed rules or outcome promises.

Use seller-authorized Amazon sources

Sponsored Products search-term report. Review queries, spend, attributed sales, orders, and match type for the seller account that authorized the work. Separate relevant converting queries from irrelevant or inconclusive traffic. Do not combine the report with another seller's data.

Brand Analytics. Eligible Brand Registry accounts can use Amazon-provided reports such as Search Query Performance and Top Search Terms. Availability and fields vary by account and marketplace. Record the marketplace, reporting period, applied filters, and any assumptions used in the analysis.

Listing and catalog evidence. Compare relevant phrases with the title, bullets, description, attributes, and backend search terms visible in the seller's authorized account. A missing phrase is not automatically a recommendation to add it; relevance, policy, readability, and product accuracy come first.

A four-step review

  1. Define the decision. Decide whether the work concerns listing coverage, advertising structure, or both.
  2. Document the source. Save the Amazon report name, seller account, marketplace, date range, and filters.
  3. Group relevant intent. Organize queries by the product need they express. Keep ambiguous or low-evidence phrases separate instead of forcing a conclusion.
  4. Review and measure. Propose changes in writing, obtain client approval where required, and assess them using a suitable post-change period. No result or timing is guaranteed.

Boundaries that matter

  • Use only data for the seller who authorized the work.
  • Do not pool, resell, license, or expose Amazon data from multiple sellers.
  • Do not ask the seller for a Seller Central password or client-owned API credentials.
  • Do not use an external vendor that sells Amazon-derived datasets as an input to the service.
  • Do not treat an observed correlation as proof of an Amazon algorithm rule.

From research to action

Keyword research is useful only when it leads to a reviewable decision. Listing recommendations should preserve factual accuracy and readability. Advertising recommendations should explain the proposed bid, budget, negative-keyword, placement, or campaign-structure change and its basis. BFarm applies advertising changes only after client agreement.

For related operating guidance, see Amazon SEO vs PPC: what to fix first, the Amazon SEO service, and the managed advertising service. To discuss an account-specific review, contact BFarm.