How to Fix Keyword Cannibalization on a Small Site

You publish a new page targeting a keyword you already rank for, and a few weeks later both pages slide down the results. That is keyword cannibalization: two or more pages on your own site competing for the same query, splitting signals and confusing search engines about which page deserves to rank.

On a small site it is common because every new post feels like it needs its own URL, and internal links get spread thin. The fix is not "add more content." It is deciding which page should own the keyword, then consolidating or differentiating the rest.

How to confirm it's cannibalization (not something else)

Before you start merging pages, verify the diagnosis. Ranking drops have many causes, and deleting good pages is expensive.

• Search the query yourself. Google the target keyword and look at which of your URLs appear. If two or more of your pages rank on page one or two for the same query, that is a strong signal.
• Check Search Console's Performance report. Filter by the query, then look at the Pages tab. If the same query is pulling impressions from multiple URLs, they are competing.
• Look at position volatility. Cannibalized pages often swap positions over time as Google alternates which one it prefers.
• Rule out the alternatives. A sitewide drop usually means a technical or algorithmic issue, not cannibalization. A single-page drop may just be a competitor outranking you.

If one URL consistently ranks and the others get impressions but never clicks, you have confirmed the problem.

Decide: merge, differentiate, or redirect

Every cannibalized pair has one of three correct outcomes. Pick deliberately.

Merge when both pages cover the same intent

If two pages answer the same question for the same audience, combine them into one stronger page. Move the best sections from the weaker page into the stronger one, then redirect the weaker URL to the survivor with a 301.

This is usually the right call for thin or overlapping blog posts. One thorough page beats two average ones, and the redirect passes the accumulated signals to the page you want to rank.

Differentiate when the intents are genuinely different

Sometimes two pages look like duplicates but serve different searchers. A page targeting "email marketing for restaurants" and a page targeting "email marketing software" are not the same query, even if the copy overlaps.

In that case, rewrite each page so it clearly targets its own keyword, with its own title tag, H1, and opening paragraph. Remove the overlapping phrases from whichever page should not own them.

Redirect when one page has no reason to exist

If a page was created by mistake, is outdated, or was always a duplicate, redirect it to the closest relevant page. Do not leave it live and unlinked.

Fix the on-page signals so one page clearly wins

Once you have chosen the surviving page, tighten its signals so search engines have no ambiguity.

• Give it a unique title tag and H1 built around the target keyword. If two pages share near-identical titles, that is a direct competition signal.
• Rewrite the meta description so it reads as the definitive answer to that query.
• Remove the keyword from the losing pages. If they stay live, they should target something else entirely.
• Fix internal links. Point your internal links at the surviving page using descriptive anchor text. Links scattered across three competing pages dilute all three.
• Check the URL. If the surviving page has a vague slug, a redirect from a cleaner URL is usually not worth the risk. Leave it unless the slug is actively misleading.

A quick pre-publish check to prevent it recurring

Most cannibalization is created at publishing time. Before you hit publish on a new page:

1. Search your target keyword on your own site (use site:yourdomain.com keyword in Google, or your site's internal search).
2. If an existing page already targets it, update that page instead of creating a new one.
3. If the new page serves a different intent, make sure its title tag, H1, and first paragraph use different phrasing.
4. Add at least one internal link from the new page to the most related existing page, and one back.

This takes two minutes and prevents most of the cleanup work later.

What not to do

• Don't delete pages outright if they have backlinks or traffic. Redirect them instead so the signals carry over.
• Don't redirect everything to the homepage. Redirect to the most relevant surviving page.
• Don't stuff the same keyword into every page's title hoping one will rank. That guarantees competition.
• Don't panic and rewrite everything at once. Fix one cannibalized cluster, wait a few weeks, then measure.

How to tell it worked

After you merge or differentiate, give it a few weeks. In Search Console, filter the query and check the Pages tab again. You should see fewer URLs receiving impressions and the surviving page's average position improving. Clicks usually follow, though not always immediately.

If nothing changes, re-check whether the surviving page actually answers the query better than the competition. Cannibalization is often a symptom of thin content, not just duplication.

Run a quick audit before you start

Cannibalization is easier to spot when you look at your whole site at once rather than page by page. Run a free audit to surface overlapping pages, thin content, and the on-page issues that make competition worse.

For related fixes, browse the guides library — internal linking, title tags, and content consolidation all tend to come up in the same cleanup.

More guides · Compare audit tools · Run a free website audit