Keyword cannibalization happens when two or more pages on the same site are built to rank for the same query. Search engines then have to choose between them, and the choice is rarely stable: one page ranks on Monday, the other on Thursday, and neither ever accumulates the links, engagement signals, or ranking history that a single consolidated page would have earned. This keyword cannibalization checker takes a plain list of URL and keyword pairs, groups them by keyword, and scores how serious each overlap really is.
Arb Digital's SEO team runs this audit at the start of nearly every content engagement, because on sites with more than about eighty published pages, overlap is close to inevitable. Nobody sets out to write the same article twice. It happens gradually, one well-intentioned blog post at a time, and it usually goes unnoticed until rankings start oscillating for no obvious reason.
What This Keyword Cannibalization Checker Does
Paste one line per page, in the format URL, keyword, monthly volume. The tool normalises every keyword — lowercasing it, collapsing extra spaces, and stripping trailing punctuation — then counts how many distinct URLs claim each one. Any keyword claimed by two or more URLs is flagged, ranked by severity, and shown in the breakdown bars with the number of competing pages attached.
The headline number is the count of cannibalized keywords. Beneath it, the supporting grid shows how many pairs were read, how many unique URLs are caught up in a conflict, what share of your keyword set is affected, and how much monthly search volume sits behind the affected terms. That last figure is the one that decides priority. Three pages fighting over a 90-volume long-tail phrase is a rounding error. Three pages fighting over a 22,000-volume head term is a strategy problem.
Everything runs in your browser. Nothing you paste is uploaded, stored, or sent anywhere, which matters when the list you are auditing is an unreleased content plan or a client's full URL inventory.
How to Use It
- Export your data. The fastest source is the Search Console performance report: filter to Queries, and note the top page for each query. Alternatively, export your primary target keyword per URL from whatever content brief or rank tracker you already maintain.
- Paste one row per URL–keyword pair. A page targeting three keywords gets three lines. Separate fields with a comma or a tab; the tool accepts either.
- Add monthly volume if you have it. Volume drives the severity weighting. Without it, every keyword is treated as equally valuable and severity depends only on how many pages are competing.
- Set a minimum volume filter. Raising this above zero hides long-tail noise so the report only surfaces conflicts worth an afternoon of work.
- Read the bars top down. The highest severity conflict is your first consolidation candidate. Work down the list, not across it.
The Formula: How Severity Is Scored
Severity blends two things — how many pages are competing, and how much traffic is at stake:
Severity = min(100, (Pages − 1) × 30 + (Keyword volume ÷ Highest volume in your list) × 40)
Each additional competing page adds 30 points, because the damage compounds: two pages split the signals, three pages fragment them further and make it far more likely that search engines settle on the weakest of the set. The volume component contributes up to 40 points, scaled against the largest keyword in your own list rather than an absolute benchmark, so the score stays meaningful whether you run a national retailer or a local service business.
A keyword with 8,100 monthly searches in a list topping out at 22,000 volume, targeted by two URLs, scores (2 − 1) × 30 + (8100 ÷ 22000) × 40 = 30 + 14.7 = 45 — a real conflict, but not the first one to fix. The same keyword targeted by three URLs scores 75 and moves to the top of the queue. Scores are banded as low (under 25), medium (25–49), high (50–74), and critical (75+).
When Two Pages on One Keyword Is Completely Fine
This is the part most cannibalization advice gets wrong. Overlap is only harmful when the pages serve the same intent. A category page at /shop/running-shoes/ and a buying guide at /blog/how-to-choose-running-shoes/ may both mention "running shoes" hundreds of times, but they answer different questions. One is transactional, one is informational, and search results routinely reward both — sometimes on the same page of results.
The test is not word overlap. It is this: if you showed both pages to someone who typed the query, would they be satisfied by either one? If yes, you have cannibalization. If one clearly answers the question and the other clearly does not, you have a healthy site architecture where two pages happen to share vocabulary. Before consolidating anything, check the actual search results for the query. If the live results are dominated by product listings, your guide was never competing for that term in the first place.
A useful second signal is the impressions split in Search Console. Genuine cannibalization looks like two URLs alternating for the same query across weeks, each picking up impressions the other loses. Coexistence looks like one URL holding the query steadily while the other appears only occasionally, at much lower positions, for a scattering of variant phrasings.
The Four Patterns This Report Usually Reveals
Duplicate intent. Two articles written years apart that answer the identical question. This is the cleanest case: merge them, keep the stronger URL, and redirect the weaker one. Check the redirect target is genuinely relevant — a redirect to a loosely related hub is treated as a soft error, not a helpful signal.
Fragmented coverage. Four thin posts each covering a slice of one topic, none comprehensive enough to rank. Consolidating them into a single substantial page usually outperforms all four combined. Our SEO content length checker is a quick way to see whether the merged result lands in a competitive range for the query.
Template overlap. Location pages, filtered category views, or programmatically generated variants that differ only in a city or colour name. The fix here is usually canonicalisation and internal linking discipline, not deletion.
Accidental targeting. A page optimised for a keyword nobody intended it to target — often because the title tag was written casually and happened to match a stronger page's focus term. Rewriting one title frequently resolves the whole conflict without touching the body copy.
How to Fix a Conflict Once You Have Found It
There are four legitimate resolutions, in rough order of how often they are the right call. Merge and redirect when both pages are genuinely trying to do the same job: combine the best sections into the stronger URL, then 301 the weaker one. Differentiate when both pages deserve to exist but drifted together: rewrite titles, headings, and opening paragraphs so each targets a distinct angle and intent. Canonicalise when the duplication is structural and both URLs must remain reachable, as with faceted navigation. De-optimise when one page only accidentally competes: remove the conflicting term from its title and H1 and link internally from it to the page you actually want ranking.
Whichever route you take, fix the internal links afterwards. If twenty internal links still point to the page you just demoted, you have told search engines the opposite of what you intended. Recalculating your link distribution with an internal link opportunity calculator after a consolidation is a five-minute job that protects the work you just did.
Where Cannibalization Comes From in the First Place
Almost always from a missing shared record of what each page targets. When three writers work from three separate briefs, nobody notices that two of them chose the same primary keyword. The structural fix is a single keyword map — one row per URL, one primary keyword per row, owned by whoever approves content — checked before a brief is written rather than after publication.
The second common source is content refreshes that quietly become new articles. A "2024 guide" gets a 2026 rewrite published at a new URL, the old one is never redirected, and now two near-identical pages compete for the same term. Republishing at the same URL, with the date updated, avoids the whole problem. Google Search Central's guidance on consolidating duplicate URLs covers the canonical mechanics in detail and is worth reading before you make structural changes.
Reading the Score Without Overreacting
A high severity score is a prompt to investigate, not an instruction to delete a page. Deleting content is the most destructive available fix and the hardest to reverse. Before removing anything, check whether the "losing" page earns backlinks, converts visitors, or ranks well for a completely different set of terms that this report never captured — because your input list only contains the keywords you chose to paste, and a page usually ranks for dozens more.
Equally, a low score is not permission to ignore a conflict on a commercially critical term. If two pages compete for the single query that drives your revenue, that is worth fixing regardless of what the arithmetic says about relative volume. Judgement outranks the score, every time.
Arb Digital's SEO team builds keyword maps, runs consolidation audits, and handles the redirect and internal-linking work that makes a merge actually stick — without losing the rankings you already have.
SEO Services Content MarketingCommon Mistakes to Avoid
- Treating every overlap as a problem — pages serving different intents can share a keyword indefinitely without harming each other.
- Deleting the weaker page instead of merging it — you lose its backlinks, its ranking history, and any long-tail terms it quietly owned.
- Redirecting to an unrelated hub page — a 301 that lands somewhere irrelevant is treated as an error, not a consolidation.
- Fixing the page but not the internal links — your own navigation keeps voting for the URL you just demoted.
- Auditing once and never again — cannibalization regrows with every publishing cycle unless a keyword map is maintained.
Related Free Tools From Arb Digital
Once you know which keywords overlap, size up the competition with the keyword difficulty checker, group the survivors into non-competing themes with the keyword cluster generator, and check on-page focus with the keyword density checker. Before you merge, preview how the consolidated result will look in search with the SERP snippet preview, and see whether declining pages need a refresh rather than a merge using the content decay calculator. The full free online tools hub has the rest.
Frequently Asked Questions
It is when two or more pages on the same site are optimised for the same search query, forcing search engines to choose between them. The result is usually unstable rankings, split link equity, and neither page performing as well as a single consolidated page would.
It can, when the competing pages serve the same intent. The damage is indirect: links, engagement, and internal signals get divided across URLs instead of concentrating on one. Pages that share vocabulary but answer different questions generally coexist without harm.
Two competing pages on the same intent is already worth reviewing. Three or more almost always indicates fragmented coverage that would perform better consolidated, which is why each extra page adds thirty points to the severity score in this tool.
Merging is usually better than deleting. Move the useful sections into the page you want to keep, then redirect the old URL so its backlinks and ranking history transfer rather than disappear.
Open the Performance report, filter by an individual query, then switch to the Pages tab. If several URLs appear for that single query and trade positions week to week, you have a candidate conflict worth checking here.
They help when the duplication is structural, such as filtered or parameterised URLs that must stay reachable. They do not fix two genuinely different articles competing for one term — that needs a merge, a rewrite, or a change of target keyword.
Yes. All processing happens in your browser using JavaScript. Nothing you paste is transmitted to a server, logged, or stored anywhere after you close the page.
Severity scores produced here are a prioritisation aid based on the data you enter, not a diagnosis — always confirm a suspected conflict against live search results before restructuring pages.
