Two of your blog posts have been swapping places in the rankings for six months, neither ever cracking page one, and you've been blaming "the algorithm" instead of your own content calendar. You can find keyword cannibalization in Google Search Console for free, using data you already have access to, without opening a single third-party tool. Most founders never check, because the fix feels like it requires something expensive. It doesn't. It requires about fifteen minutes and knowing exactly where to look.
How to Find Keyword Cannibalization in Google Search Console: Step by Step
To find keyword cannibalization in Google Search Console, filter the Performance report by a specific query, then check how many URLs appear for it and how impressions split between them. If two or more of your own pages show similar impressions, alternating positions, or both stuck below position 10, you're likely watching cannibalization play out in real time.
Here's the exact sequence:
- Open the Performance report in Search Console and click "Search results."
- Add a filter for "Query" and enter a keyword you care about, ideally one tied to a page's main topic.
- Look at the table below the chart and switch the grouping to show pages, not just totals.
- Note every URL that appears for that query, along with its clicks, impressions, and average position.
- Flag any query where two or more of your pages show up with comparable impression counts.
- Check the date range selector and compare the last 3 months against the previous 3 months to see if positions have been swapping.
- Repeat this for your 15-20 highest-impression queries, since that's where cannibalization does the most damage.
- Cross-reference flagged pairs against your sitemap to confirm they're meant to target different intents, not the same one.
That seventh step is the one people skip. Checking one keyword feels productive. Checking your top 20 is what actually surfaces the pattern, because cannibalization rarely shows up as an isolated incident. It shows up as a habit, usually tied to how you've been naming and structuring articles over the past year.
What Cannibalization Actually Looks Like in the Data
It's not always two pages ranking at once. Sometimes it's one page quietly losing the fight without you noticing. Here's the tell nobody talks about: a page that used to rank in position 4 slides to position 9 over a few weeks, and instead of falling further, it plateaus there. That plateau is suspicious. It usually means a newer page on your site has started absorbing part of the relevance signal for that query, and Google is now splitting its confidence between the two instead of committing to one.
The clearest signals in the query report:
- Two URLs appear for the same query with impressions within roughly 20% of each other.
- Average position for both hovers between 8 and 15, good enough to show up, not good enough to win.
- Click-through rate is lower than expected for the combined impressions, because searchers see two similar-looking results and hesitate.
- The pattern started right after you published a new post covering an overlapping subtopic.
That last point matters more than people give it credit for. Pull the publish date of your suspected second page and line it up against the point where position data started fluctuating. If they match within a week or two, you've got your answer.
The Page Report: Your Second Data Source
The query report tells you which searches are contested. The Pages report, sitting right next to it in Search Console, tells you which URLs are contesting them, and comparing the two is where most people stop looking too early. Click into the Pages tab, sort by impressions, then click on any page you suspect is involved. Search Console will show you every query that page ranks for. If page A and page B both show the same top three queries, in a similar impression range, you're not looking at two related articles, you're looking at one article split across two URLs, and Google is trying to guess which half deserves the click.
This is where a lot of founders get the fix wrong before they even get to the fix. Two pages ranking for the same query isn't automatically a problem — a comparison page and a how-to guide can legitimately share a keyword because they serve different intents, and Google is smart enough to keep both around. The problem is when the pages are functionally interchangeable: same intent, same angle, same target reader, just written six months apart because you forgot you'd already covered it.
The Mistake That Makes This Invisible
Most founders check cannibalization by searching one keyword at a time, and that's exactly why they miss it. Cannibalization compounds across dozens of queries simultaneously, and a one-off check only catches the obvious cases. The real damage sits in your long-tail queries, the ones with 50-200 monthly impressions each, spread across three or four blog posts you wrote over an 18-month span without ever cross-referencing your own archive. Individually, none of those look alarming. Added up, they're the reason your organic traffic has flatlined despite publishing consistently.
There's a second edge case worth naming: seasonal or trending queries can fake cannibalization signals. If two pages both spike for a query during a short window — a product launch, a news cycle — that's coincidence, not cannibalization. Check whether the overlap persists for at least 60-90 days before you act on it. Acting on a two-week blip wastes time you don't have.
Fixing It Without Guessing
Finding the overlap is the diagnostic step. What you do next — merge, redirect, differentiate, or leave alone — depends on search intent, page authority, and how much unique value each page actually adds, a decision covered in full in how to fix keyword cannibalization.
This is also where manually scanning query reports starts to break down. If you're running this check across 40 or 50 published posts, doing it by hand in Search Console means dozens of filtered views, spreadsheet exports, and cross-checking dates by eye. WriteGap's cannibalization detection pulls Search Console data automatically and flags competing URLs for the same query without you filtering one keyword at a time. Instead of spot-checking your top 20 queries, you see every conflict across your full archive at once, including the ones sitting in your long tail that never show up when you're guessing which keywords to check manually.
Should You Fix Every Cannibalization Issue You Find in Search Console?
No, not every overlap is worth touching. If two pages target genuinely different intents (a comparison post and a pricing page, for example) and each ranks reasonably well on its own, leave them alone. Reserve fixes for cases where both pages are stuck below position 8, splitting near-identical impressions, and covering the same angle for the same reader. WriteGap's fix-this-page recommendations can tell you which of the two is worth keeping once you've confirmed the overlap is real.