Keyword Consolidation: Merging Pages That Compete With Each Other

TL;DR: When two or three of your pages target the same query, none of them wins. They split impressions, links, and relevance signals, and they usually stall on page two. The fix is unglamorous — pick one page, merge the others into it, redirect, and rebuild internal links. We’ll show you the problem on our own site, including the numbers, because it’s the clearest example we have.

Table of Contents

• How competing pages form on purpose
• The Search Console query that exposes it
• Our own three-page overlap, with the numbers
• Which page to keep
• Merge, redirect, and relink mechanics
• What recovery looks like

We’ve written about diagnosing keyword cannibalization before — see our comprehensive guide to resolving keyword cannibalization and recognizing and eradicating keyword cannibalization. Those cover identification. This one covers the part that’s harder to write about honestly: the merge itself, and what it looks like when the site with the problem is ours.

How Competing Pages Form On Purpose

Nobody publishes two pages about the same thing by accident. Each one made sense when it was written.

Year one. You publish a guide on a topic. It ranks reasonably.

Year two. A new writer, a new content plan, or a new agency doesn’t know the old post exists — or knows and judges it dated. They write a fresh, better one. The old one stays live, because deleting content feels wrong and nobody wants to lose a page that gets some traffic.

Year three. Someone notices the topic is valuable and commissions “the definitive guide” to it. Now there are three.

Each decision was locally reasonable. The aggregate is three pages splitting one query family.

This is worse on sites with high publishing cadence, and it’s near-universal on sites that have changed content providers. It also isn’t a penalty and nothing is broken. That’s why it survives so long — the site looks healthy, individual pages get impressions, and no error report flags it.

The Search Console Query That Exposes It

Fifteen minutes, no tooling beyond Search Console.

1. Open Performance → Search results, set the date range to the last three months.
2. Click the Queries tab, sort by impressions descending.
3. For each of your top 20 queries, click it, then switch to the Pages tab.
4. You’re looking for queries where more than one of your URLs appears.

That’s the whole diagnostic. When two of your pages both show impressions for the same query, they’re competing.

Two supporting signals confirm it:

High impressions, near-zero clicks, position 11–30. The classic signature. Google has decided your site is relevant to the query but can’t decide which page, so neither gets the position that earns clicks.

Position instability. Watching a query where the ranking URL flips between two of your pages week to week is Google telling you plainly that it can’t pick.

Do the same pass on your Pages tab in reverse: pick your most important pages, look at the queries each ranks for, and see whether two pages claim the same query set.

Our Own Three-Page Overlap, With the Numbers

We ran this on pushleads.com while planning this quarter’s content. Here’s what we found, unedited.

Three published posts target the same question — how many keywords should a page target:

The query family they’re splitting, from our own Search Console over a 28-day window, includes: how many keywords per page seo (1,624 impressions), how many seo keywords per page (1,601), how many seo keywords should i use (1,493), how many keywords for seo (1,296), how many keywords should i use (912), how many keywords per page (895), how many keywords should i use for seo (862), how many keywords should a website have (853), plus eight more variants.

Totals across the sixteen variants: roughly 13,000 impressions in 28 days. Average position between 16 and 26. Clicks: effectively zero.

Thirteen thousand impressions is real demand. Position 16 to 26 is page two and three, where click-through is close to nothing. We have three pages, considerable topical coverage, and we are earning nothing from any of it.

That’s not a content problem. We have too much content. It’s a consolidation problem, and it’s worth more than any new post we could write this quarter.

We’re publishing the fix as it happens. The point of showing you our own numbers rather than a client’s is that you can verify the pages exist and check the outcome yourself in a few months.

Which Page to Keep

Four criteria, in order.

1. Backlinks. Check Search Console → Links → Top linked pages. The page with the most external links usually wins, because that’s the hardest thing to move. You can rewrite content; you can’t easily re-earn links.

2. Current best position. Whichever page Google already prefers has some accumulated signal. Fighting that preference costs you time.

3. URL quality. A clean, keyword-relevant, non-dated URL is worth keeping. /how-many-keywords-per-page/ beats /how-many-seo-keywords-per-page-2026-guide/, because the second one is stale in five months and you’ll want to redirect it again.

4. Content quality and length. Least important, because you’re going to merge the best material from all of them anyway.

When criteria conflict — the best URL has the fewest links — keep the page with the links and, if the URL is genuinely bad, change the slug on that page and redirect the old slug. One extra redirect is cheaper than abandoning link equity.

Merge, Redirect, and Relink Mechanics

Do it in this order. Skipping step five is the most common way this goes wrong.

1. Pick the survivor using the criteria above. Write it down.

2. Build the merged page. Open all versions side by side. Take the best explanation of each sub-topic from whichever page did it best. The result should be genuinely better than any of the originals, not just longer. Add anything all three missed.

3. Cover the full query family. You now know all sixteen variants from Search Console. Make sure the merged page actually answers each phrasing — as sections, as FAQ entries, as natural body copy. That’s the point of merging: one page comprehensive enough to satisfy every variant.

4. Publish the merged version on the surviving URL. Update the publish or modified date honestly.

5. 301 the retired URLs to the survivor. Direct, one hop. Then confirm they resolve with a single redirect, not a chain.

6. Rebuild internal links. Crawl your site for links pointing at the retired URLs and repoint every one at the survivor. Relying on redirects for your own navigation wastes the signal and slows the page. This step is where most consolidations lose half their value.

7. Update your sitemap and resubmit.

8. Request indexing of the survivor in URL Inspection.

9. Record the baseline. Export the combined impressions, clicks, and average position for the query family before anything moves. Without this you’ll never know whether it worked.

Two things not to do. Don’t noindex the losers instead of redirecting — you keep the duplication and throw away the link equity. And don’t delete them outright, which strands every backlink they had.

What Recovery Looks Like

Realistic expectations, based on how this typically behaves:

Weeks 1–2. Volatility. The merged page may temporarily rank worse than the best of the originals did. This is normal and it’s the part that makes people panic and revert. Don’t.

Weeks 3–6. Google settles on the single URL. Position for the query family typically improves as signals consolidate. Impressions may drop slightly — you had three pages collecting impressions and now have one — while clicks rise. That’s the trade you wanted.

Weeks 6–12. The real result. If the merged page moves from position 18 to inside the top ten, clicks change by an order of magnitude, because click-through rate at position 8 is dramatically higher than at 18.

How to judge it. Compare the query family before and after, not the individual URLs. The correct measurement is total clicks for those sixteen variants, before versus after. Anyone reporting this as “the old page lost traffic” is measuring the wrong thing.

When it doesn’t work. Sometimes the merged page stalls anyway, because the query is genuinely too competitive for your site’s authority, or because the search intent wants a different format than you produced. That’s useful information too — it tells you the problem was never cannibalization, and you can stop spending on the topic.

Quick Recap

• Competing pages form through reasonable individual decisions across years and providers.
• Find them in Search Console: for each top query, check the Pages tab for more than one of your URLs.
• The signature is high impressions, near-zero clicks, position 11–30, and an unstable ranking URL.
• Keep the page with the backlinks, then the best position, then the best URL. Merge the best content from all versions.
• Redirect one hop, then repoint every internal link — that step is where consolidations lose value.
• Judge the result on total clicks for the whole query family over 6–12 weeks, not on individual URLs.

Frequently Asked Questions

How do I know the difference between cannibalization and just having related pages?
Related pages target different queries and rank for different things — that’s healthy topical depth. Cannibalization means the same query returns multiple of your URLs. Search Console’s Pages tab under a query answers it definitively.

Is keyword cannibalization a penalty?
No. There’s no penalty. It’s a signal-splitting problem: relevance, links, and engagement divided across pages instead of concentrated in one.

Should I merge or just improve one and noindex the others?
Merge and redirect. Noindexing keeps the duplication in your site and abandons any link equity the retired pages had. A 301 consolidates.

What if both pages rank well for different queries?
Leave them alone. If page A owns one query set and page B owns another, they aren’t competing. Consolidate only where the same query pulls up both.

How long should I wait before judging the result?
Six weeks minimum, twelve for confidence. Anything you measure in the first two weeks is volatility, not outcome.

Can this hurt me?
Done properly — one-hop redirects, internal links updated, best content preserved — the downside is small. The risk comes from sloppy execution: chained redirects, deleted pages, or losing content that was earning.