Using Amazon Brand Analytics to Find Keywords Worth Buying

Amazon ยท October 26, 2026

Most Amazon advertisers optimize from their own data: search term reports, campaign reports, and rank trackers. That is useful, but it only shows you what happened on the traffic you already earned or paid for. Brand Analytics shows something different, which is how you compare with the whole market on the same searches.

The Search Query Performance report is the part of Brand Analytics we use most. Used well, it points to queries where shoppers already prefer your product but rarely see it, and those are often the best places to put the next dollar of ad spend or the next round of listing edits.

What you need to access it

Brand Analytics is available to sellers who own a brand in Amazon Brand Registry, and access is tied to your Brand Registry role and the ASINs under your brand. If you do not see it in Seller Central, check your Brand Registry enrollment and user permissions first. Amazon changes the menu structure and available reports from time to time, so look for it under the Brand Analytics section of your console and check the current options there.

If you are not brand registered, you will not have this data, and the rest of this post will be theory. In that case, enrolling is worth doing for several reasons beyond analytics.

What the Search Query Performance report shows

The report is built around search queries, meaning the exact phrases shoppers typed. For each query you can generally see, for a chosen reporting period:

There are views at the brand level and at the ASIN level, and reporting periods typically include weekly, monthly and quarterly. The exact columns and period options can change, so confirm what your console currently offers.

The important idea is that every stage of the funnel has a market total and your share of it. That lets you compare your share at one stage with your share at the next.

The core comparison: share by funnel stage

For any query, ask four questions in order:

  1. What share of impressions do we have?
  2. What share of clicks do we have, relative to impressions?
  3. What share of cart adds do we have, relative to clicks?
  4. What share of purchases do we have, relative to cart adds?

If your purchase share is higher than your impression share, shoppers who see you are choosing you more than the market average would suggest. That is the signal we look for first. If your clicks share is lower than your impression share, your main image, price or title may be underperforming in that query. If your cart adds share is high but purchase share drops, look at price, delivery promise and anything that might cause hesitation late in the process.

Finding queries where you convert above market but have low impression share

This is the most actionable pattern. Our process:

  1. Export the report for a recent period, by ASIN or for the brand overall.
  2. Filter out tiny queries. Very low volume or very few total purchases means noise.
  3. Sort by the gap between purchase share and impression share.
  4. Check consistency across more than one period.
  5. Sanity-check relevance. A narrow, exactly-fitting query is a better lead than a vague one.

The result is a short list of queries where shoppers already like your product when they see it. Your problem on those queries is visibility, not persuasion, which is a cheaper problem to solve.

Turning findings into changes

A list of queries is only useful if it changes something. We usually sort each finding into one of three buckets.

Listing changes

If a strong query does not appear in your title, bullets or backend terms, that is a likely reason for the low impression share. Work the phrase into the title if it fits naturally and respects the length guidance in our 75-character title post, into a bullet where it helps the shopper, or into the backend search terms if it does not belong in visible copy. Do not stuff. The aim is to describe the product in the words shoppers actually use.

Advertising changes

If the query already appears in your listing but impression share is still low, paid placement can fill the gap. Create or expand exact-match targeting for the query in a well-structured campaign (see our Sponsored Products structure guide), and set bids with your margin in mind using break-even ACoS. Because your conversion on these queries is already above market, you often have room to bid more confidently than on a speculative keyword, though you should still verify against your own ad data once it starts running.

Image, price or offer changes

If your impression share is fine but click share lags, look at the main image, price and rating compared with the results you appear next to. Our main image post covers how to approach that.

Beyond the obvious wins

Check your branded queries too: low share on your own brand name is worth investigating. Queries with high market volume and almost no brand impressions can reveal gaps in your catalog or listing language, and month and quarter views can show seasonality so you can plan stock and budget.

Limitations to keep in mind

Brand Analytics is powerful, but we treat it with caution.

Because of these limits, we use Brand Analytics to generate hypotheses and prioritize, then confirm with campaign data and experiments before making large budget decisions.

Want a second set of eyes?

Reading this report well is mostly a matter of knowing which numbers to trust and which to ignore. If you want us to go through your Brand Analytics data and your campaigns together, request a free Amazon audit and we will show you where we would look first.