How do you know if two pages are cannibalizing each other?
- Published November 6, 2026
- Linked from 5 answers
- In a cluster of 77
Check Search Console, not your impression of the two pages: cannibalization is one query for which two of your URLs both appear, taking turns or splitting the clicks, while neither holds a strong position, and both pages are trying to answer the same thing. Two pages on related topics that each rank for their own queries are not cannibalizing anything, however similar they look.
Run the check
- Pick the query. In the Performance report, open the query you care about.
- Open the Pages tab. If only one of your URLs appears, there is no cannibalization for that query.
- If two URLs appear, compare their clicks, impressions and average position for that query over the last few months.
- Switch the date view to see whether the URL shown changes from week to week.
- Read both pages and write what each is for in one sentence.
Read the result
| What the data shows | Verdict |
|---|---|
| One URL carries the query; the other appears now and then with few impressions | Not a real problem; the second page is a minor secondary match |
| Two URLs alternate for the query, with unstable positions and split clicks, and both sentences describe the same purpose | Cannibalization |
| Two URLs appear, but each sentence describes a different purpose and each page ranks well for its own other queries | Not cannibalization; the query sits between two legitimate pages |
| Two URLs appear because one is a parameter or duplicate version of the other | A duplicate URL problem; fix with a canonical or redirect, not a merge |
The second row is the one that needs a fix. The third row is the false alarm to watch for: merging two pages that each serve their own readers can lose more than it gains; weigh that risk with the risk of consolidating pages that individually rank.
Why the data step matters
Topic resemblance misleads in both directions. Two pages that look alike can rank for entirely different searches, and two pages that look distinct can quietly compete for one valuable query. Only the query view shows which case you are in, and read it by date, device and country: two URLs that both receive impressions for a query may be shown to different searchers in different places, which is not the same as competing for one result, and only a real loss of clicks or position makes the overlap worth acting on. Once the data confirms one query split across two interchangeable pages, the next decision is whether to combine them or pull them apart; make it with whether to merge cannibalizing pages or differentiate them. On a store, the pair to check first is a category and its products: how to keep category pages from cannibalizing product pages. Confirm the split in the data before changing either page.
Add a comment