Keyword cannibalization check: decide whether overlapping pages need fixing

Open the Search Console Performance report, choose a date range and search type, and apply a query filter for the phrase you want to investigate. Inspect the Pages dimension to see the associated URLs. Keep the country and device scope consistent when comparing observations.

1. Choose a query or closely related group with a clear reader task.
2. Record the pages that appear for it and the exact report scope.
3. Compare page-level observations over comparable periods rather than one isolated day.
4. Open each live page and write down what it helps the visitor do.
5. Record a hypothesis about the overlap before choosing an edit.

A change in which URL receives impressions can be worth investigating, but it does not establish the cause by itself. Search context, small samples, and different meanings within a query group can affect what you observe. The report also has aggregation and visibility limits; it is not a complete list of every search appearance.

Copy the fields or download the Markdown file for each investigation. Fill in the reader tasks and the evidence for your decision. Leave an observation unverified if the report or source is unavailable rather than treating the template as a diagnosis.

# Keyword cannibalization review

Query or query group:
Review date and reviewer:
Search Console property:
Date range, search type, country, and device filters:

## Pages observed
Page A URL and reader task:
Page B URL and reader task:
Clicks, impressions, CTR, and position for each page in the same scope:
Dates or conditions under which the visible page changes:

## Read the pages
Does each answer a distinct reader need?
What important information is duplicated or missing?
Which page should own the primary task, and why?
Relevant internal links and anchor wording:
Canonical or redirect issues requiring a technical review:

## Decision
Keep separate / differentiate / investigate duplicates / consolidate:
Evidence supporting the decision:
Useful material to preserve:
Old-to-new URL mapping, if applicable:
Owner and planned change:
Live checks completed:
Follow-up window and comparable filters:

Imagine a fictional outdoor store with a rain-jacket collection and a guide comparing waterproof and water-resistant clothing. Both can appear for related jacket searches, but one helps visitors browse available products while the other explains a distinction. Their shared terminology does not make either page redundant.

Read the opening, main sections, product recommendations, and next step on both pages. Different titles can conceal essentially the same answer. Conversely, two titles containing ‘rain jacket’ can describe distinct tasks. Base the decision on what visitors get, with the search data as supporting context.

• Keep separate when each page serves a distinct need and the relationship is clear.
• Differentiate when the intended tasks differ but the introductions, headings, or examples blur them together.
• Investigate duplication when substantially the same content is available under more than one address.
• Consider consolidation when pages answer the same task and a single maintained destination would be more useful.

Do not choose a destination by word count or age alone. Preserve useful evidence, accurate product details, and important explanations. If the pages should remain separate, a clearer scope and relevant internal links may address the problem without moving either URL.

Google’s canonicalization documentation concerns duplicate or very similar URLs. A canonical annotation is a signal about a preferred version, not a general-purpose remedy for two different articles mentioning the same subject. Have a technical reviewer assess canonical and redirect changes against the actual page relationship.

For a content revision, confirm that the title, H1, opening, and examples now support the intended reader task. For a consolidation, record which URL remains, what material was preserved, and how the retired address and internal links are handled. Test the route on the live site after the approved change.

Save the starting observations and return to the same report scope after a suitable follow-up period. Record counts alongside the direction of change. A small before-and-after movement does not prove that the edit caused it, and there is no fixed number of days that guarantees a useful result.

Put the final page assignment in your keyword map and update the content inventory. For contextual navigation, review the internal linking workflow. Those records help the next writer avoid recreating the same overlap.