Keyword cannibalization

Two of your pages on one search means diluted relevance. Links, clicks and internal anchors split across them. The stronger rival keeps one focused page.

You also confuse Google about which page to rank. Rankings wobble. Your average position swings, impressions rise without clicks, and the wrong page shows.

Example: both /blog/best-crm and /guides/crm-software chase "best crm for startups". Neither holds page one. One clear page could.

1. 1

   Find suspect queries

   Open the Search results report. Set the last 28 days. In Queries, sort by Impressions. Pick a target query you care about.

2. 2

   See all pages that rank for it

   Add a Query filter for that term. Switch to the Pages tab. If more than one URL has impressions, you may have cannibalisation.

3. 3

   Check trends and swapping

   Change the date to Compare last 28 days to previous. Still on the Pages tab, compare which URL gained while another lost. Swapping is a sign.

4. 4

   Look for wrong-page clicks

   Click into a page, switch back to Queries. If the page gets clicks for a query it only mentions in passing, it might be stealing that query.

5. 5

   Scan the SERP

   Google the query in an incognito window. Note the dominant intent: informational list, product page, pricing, comparison. Check which of your URLs best matches it.

• Pick the URL that best matches search intent today. If the SERP is list posts, a long-form guide should win, not /pricing.
• Keep the page with backlinks, internal links and history. You can move content into it.
• If neither page fits, create one that does, then redirect the weaker pages into it.

A fixed page looks like your /guides/best-crm-for-startups covering picks, criteria and FAQs, with the old /blog/best-crm redirected to it.

• Consolidate and redirect: merge overlapping content into the winning URL. Use 301 redirects from the old URLs.
• Reframe intent: retarget the weaker page to a different query. Change title, H1, intro, and internal anchors to match a new intent.
• Use a canonical only when two pages must exist, for example a sorted view. Point the non-preferred to the canonical URL.
• Tighten internal links: update nav, footer and in-article anchors to use the winning page with a consistent anchor like "best crm for startups".
• De-optimise the losing page: remove the head term from title and H1, and link it to the winner with a clear context.
• Prune thin duplicates: delete tag pages or paginated archives that accidentally rank for the head term, and redirect them.

Change only a few things at a time. Recheck the Pages tab for that query after the next crawl and index cycle.

• Do not rely on rel=canonical to fix competing content when both URLs are indexable and distinct. Canonicals are a hint, not a directive.
• Seasonal or date-stamped posts pile up. One "2023" and one "2024" list will fight. Keep a single evergreen URL and update it.
• Homepage vs feature page for brand terms is not cannibalisation. That is normal behaviour for navigational intent.
• Programmatic pages can collide. Use clear facets, titles and on-page copy to separate intents like "compare X vs Y" and "best X for Z".
• Measure after redirects. If signals split across http, https, or with and without trailing slash, fix canonicals and redirects first.