How to Measure SEO Performance for a Topic Cluster

Key takeaways

Define and version the cluster member list.
Compare a fixed cohort apart from new pages.
Trace group changes back to individual pages.

To measure topic cluster performance, define its member pages first. Compare the same group over time, then inspect the pages that explain the change. Keep search visibility separate from on-site engagement and leads. A single total can tell you that the group changed. It cannot tell you which article deserves the credit or whether internal links caused the result.

Create a fixed list of cluster members

Start with the pillar and the published satellites that serve its topic. Record each public URL, role, publication date, and status. Give the list a version date. Do not include every page that mentions the topic. Membership should follow the content plan, not a wish to make the total look larger.

Choose one primary cluster for reporting if a page helps several topics. You can record secondary relationships, but explain them. Otherwise two teams may each claim the same clicks in totals that readers expect to be exclusive. A content relationship and a reporting group do not have to mean the same thing.

Decide which question the group should answer

Write the business question before choosing a chart. You may want to know whether a new cluster has begun to reach relevant searches. You may want to see whether an older group is losing clicks. You may also want to find a weak satellite within a growing topic. Each question needs a different comparison.

Keep the measurement scope clear. Search Console describes search performance. Your analytics setup may describe visits and tracked actions after a click. A lead record may describe a later business outcome. Do not treat those sources as interchangeable or assume that matching totals are required.

How to measure topic cluster performance with a URL group

Use a page filter or a filtered export that matches the approved member list. A folder prefix works only if every URL beneath it belongs to the group. If the folder contains unrelated pages, use an explicit list or a tested pattern. Check both known members and a known non-member.

Keep the group definition beside the data. Note the property, search type, country, device, and date range. Use the same scope for the comparison. A change in filters can look like a change in performance. Do not quietly move pages into or out of the group halfway through the review.

Separate existing pages from newly published pages

Report a fixed cohort when you want to understand how existing pages changed. Add a second view for all current members when you want to understand the full topic today. The two views answer different questions. Keeping both prevents new publication volume from hiding a decline in older pages.

Record the first date each page was live. A satellite published last week has not had the same time to earn visibility as a pillar published months ago. Avoid ranking every member by raw clicks alone. Consider age, purpose, and the searches it is meant to answer before deciding which one failed.

Use totals and rates that match the grouping

Keep clicks and impressions within the same source and aggregation level. Do not add a group total to its component page totals. Do not add overlapping groups and call the result a unique site total. Label the calculation so a reader knows whether it represents pages, queries, or the property.

Calculate a combined CTR from total clicks divided by total impressions within the same scope. Do not average each page’s CTR without weighting. A page with ten impressions should not carry the same weight as one with ten thousand. Keep position as contextual information rather than calling a mixed average the cluster’s fixed rank.

Find which members explain the movement

Sort the member pages by absolute click change as well as percentage change. Large percentages from small bases can distract from the page responsible for most of the total shift. Review both gains and losses. A growing pillar can hide several satellites that need attention.

Then inspect relevant queries for the pages that matter most. Look for a new audience, changed position context, or a different device mix. This is diagnosis of the group’s result, not a reason to rewrite every article. Give each proposed action a page and a clear finding.

Example: growth comes from one new article

Imagine a fictional cluster with a pillar and four older satellites. The group gains 100 clicks after a sixth page is added. The new page supplies 140 extra clicks while the older group loses 40. These invented numbers show why one total can hide two different stories.

The team reports both views. It welcomes the new page’s reach and investigates the older group’s decline. It does not claim that all members improved. It also does not say that the link structure caused the gain without a test that addresses other explanations. The group total is a starting point for decisions.

Connect the result to the content workflow

DEX organizes content around the business’s products and services and brings key Search Console metrics into the workflow. Use that context to connect a weak group to the offer it supports. If you need custom cohort exports or analytics segments, verify the relevant controls in the source tool rather than assuming DEX creates them.

Choose a next task that fits the evidence. A missing answer may need a content review. A changed URL may need a technical check. A relevant new query may suggest a distinct support page, but first check the existing inventory. A measurement finding should not become an excuse to publish a duplicate.

Keep a short recurring cluster review

For each review, retain the member-list version, comparison window, totals, main contributing pages, and next actions. Record any membership change explicitly. Use a review rhythm that gives the group enough data to be meaningful. Do not rebuild the report every day when little has changed.

Close with a statement the data supports: the fixed cohort grew, the full group expanded because of new pages, or a few members drove the decline. Keep lead outcomes separate unless tracking supports the connection. That gives the team a clear basis for its next task without overselling the precision of the report.

FAQ

Should I average page CTR values?

Use total clicks divided by total impressions within the same scope.

Should new pages be included?

Show the current full group and a fixed older cohort separately.

Does cluster growth prove internal links caused it?

No. Other changes may explain the result.

Sources

Official guidance · Google
Performance report overview · Google

Dex SEO

Dex SEO is an AI SEO agent for small businesses. Every article here is planned from search data, drafted by Dex, fact-checked against the sources it cites, and reviewed by the team before it is published.

How this article was made: planned and drafted by Dex from search data, fact-checked by Dex (5 claims checked), then approved by the Dex SEO team.