# Keyword cannibalization worksheet

Page decisions and content preservation

Compare two pages before deciding whether either should be retired. Record their reader tasks, map the useful content that must survive and document the next action. Use the editable worksheet first, then consult the six fictional Cedar decisions and full guide.

Pack version 1.0 · 17 September 2026. Worksheet version 1.0.

Guide: https://rankecho.io/resources/fix/duplicate-intent-consolidation

## Your page pair and preservation worksheet

Compare one page pair, map the useful content that must survive, and record a justified next action. The completed example includes six page-pair decisions to review before proposing a URL retirement.

CONTENT CONSOLIDATION WORKSHEET — version 1.0

ONE REVIEWED PAGE PAIR
Page A URL and saved version:
Page B URL and saved version:
Audience, language, region, product/version and buyer task for each:
What would a reader lose if either URL disappeared?
Observed overlap: required answers, examples, tools, downloads, exceptions and next action:
Relationship judgment: same task / different task / retained duplicate view / unresolved
Page inspection and supporting evidence locations:
Queries, dates, country, device and search type if search evidence is available:
Useful visits, referrals, conversions, external links and operational dependencies, with units and windows:
What remains unknown:

CONTENT PRESERVATION MAP — REPEAT FOR EACH UNIQUE USEFUL ITEM
Item ID and supported claim or function:
Source URL, saved version and exact section:
Keep, transfer, correct or intentionally retire — with reason:
Destination URL and exact section:
Evidence that the destination preserves the meaning and conditions:
Responsible reviewer and outstanding action:

DECISION AND SHIPMENT HANDOFF
Decision: keep separate / prepare content / propose consolidation / inspect canonical variant / investigate
Why this decision fits the reader's task:
Selected destination and why it is a relevant replacement:
Old-to-final URL mapping, only if retirement is justified:
No duplicate sources, cycles, redirect chains or unrelated homepage destinations:
Internal links, sitemap entries, fragments, assets and language annotations to inspect:
Target 200 response, useful content, canonical, intended indexing and access checks:
Source permanent redirect and final destination verification, if approved:
Publication owner, approval, deployment record, rollback copy and acceptance checks:
Separate later search, referral and product outcomes from technical completion:

COMPLETED FICTIONAL EXAMPLE — Cedar CSV export guides
All paths, content assessments and evidence pointers are invented teaching inputs.
P01: /guides/csv-export and /guides/export-records complete the same current CSV export task. Choose /guides/export-records after checking its complete draft and stable reader-facing role.
Transfer map: export-permission -> #permissions; date-filter -> #date-range; timezone -> #date-range; row-limit -> #limits; retry -> #recovery.
All five required items are present in the supplied destination record. Proposed mapping: /guides/csv-export -> /guides/export-records. This is a proposal, not an executed redirect.
P02: /pricing/data-export and /docs/export-api serve plan choice and API implementation. Keep separate and link by the relevant next task.
P03: /guides/export-migration lacks legacy-field-mapping and retention-exception from /guides/legacy-export. Prepare the destination before considering retirement.
P04: /docs/report-print remains a useful duplicate print view. Inspect canonical signals; do not remove the view just to reduce URL count.
P05: The automation/scheduled-report pair lacks an inspected relationship. Investigate; no mapping is emitted.
P06: English and French export guides serve different language needs. Keep the localized destinations and review their language annotations separately.
Acceptance: one consolidation proposal, two keep-separate decisions, one content-preparation case, one canonical review and one investigation. Only P01 has a proposed mapping; all other URLs remain outside it.
The evaluator checks supplied judgments and preservation records. It does not crawl pages, infer intent, verify the source documents, implement redirects or establish search/business effects.


The worked site and page judgments are fictional. The worksheet prepares a decision; it does not execute changes.

## What should trigger a keyword cannibalization review?

Start with a specific observation: two guides appear to answer the same question, internal links send readers to inconsistent destinations, or a landing page does not match the visitor's task. Search reports can help prioritize the review when available. Record the exact page pair and the suspected problem rather than treating every repeated phrase as a sitewide defect.

Multiple URLs appearing for related queries may be appropriate. A pricing page, an API reference and an implementation tutorial can mention the same feature while serving different needs. An older page can also retain support value for an older product version. Inspect the actual content, audience and next action before using the term cannibalization as a diagnosis.

For a new site with little search data, complete the content and navigation review now. Mark unavailable performance observations as unknown. The absence of clicks does not prove that a page is unnecessary, and a page should not need a traffic history before its useful role can be described.

- [Map buyer questions to existing pages](https://rankecho.io/resources/find/prompt-gap-analysis)

## When should pages merge or stay separate?

Ask whether a reader arriving at either page could finish the same task at one combined destination without losing important context. Compare the question, audience, language, product version, required answers and next action. A shared topic is broader than a shared task; similar titles can conceal a meaningful difference.

Use consolidation when the overlap is substantive and the destination can preserve useful differences as well-organized sections. Keep separate when combining would force readers through unrelated information or remove an audience-specific requirement. If one page contains stale facts but has a distinct purpose, a content refresh may solve the problem without changing its URL.

| Observed relationship | Appropriate next action | What must remain clear |
| --- | --- | --- |
| Same reader task; complete relevant destination | Prepare a consolidation proposal | Preserved answers, target choice and explicit old-to-final mapping. |
| Different task or language need | Keep separate | Specific titles, page roles and contextual links. |
| Same task; useful material would be lost | Prepare the destination first | The missing content and who will transfer or resolve it. |
| Duplicate view still needed by users | Review canonical signals | Why the variant remains available. |
| Missing capture or unresolved relationship | Investigate | What evidence is missing and the next inspection. |


## What do the six worked page pairs show?

Fictional teaching example: Cedar has six page pairs related to exporting data and reporting. The paths, content assessments and evidence pointers are supplied inputs, not crawled websites or customer observations. The local evaluator applies the declared relationship and preservation rules; it does not infer search intent from page text.

The result is one consolidation proposal, two keep-separate decisions, one content-preparation case, one canonical review and one investigation. Only P01 produces a proposed old-to-final mapping. There is no percentage target for consolidation and no assumption that fewer URLs would make the site better.

| Pair and fictional paths | Reader task or constraint | Decision |
| --- | --- | --- |
| P01: /guides/csv-export + /guides/export-records | Export the current workspace's records as CSV | Propose consolidation |
| P02: /pricing/data-export + /docs/export-api | Choose an export plan versus implement the export API | Keep separate |
| P03: /guides/legacy-export + /guides/export-migration | Move legacy exports to the current workflow | Prepare content: preserve legacy-field-mapping and retention-exception first |
| P04: /docs/report-print + /docs/report | Read the same report instructions in a retained print view | Review canonical variant |
| P05: /guides/report-automation + /guides/scheduled-reports | Relationship unresolved because one page capture is missing | Investigate |
| P06: /en/guides/export + /fr/guides/export | Complete the export task in the reader's language | Keep separate |


## How do you choose the page that should remain?

P01 contains two instructions for exporting the current workspace as CSV. The supplied review chooses /guides/export-records because its destination draft completes that task and has an appropriate continuing page role. The choice is not based on a manufactured authority score or an assumed ranking advantage.

For your own pages, inspect the quality of each current answer, URL stability, existing links, useful visits and conversions, and any support or product dependencies. Define the observation window and units for each metric. A commercially important destination should not be selected from raw traffic alone, and a newer URL should not replace an established one merely because its wording looks cleaner.

Document the trade-off. If the preferred destination cannot preserve a necessary older-version workflow, keep a separate version-specific resource or revise the proposed scope. A destination that is only vaguely related is not an adequate replacement for the retired page.


## How do you preserve useful material before merging?

Inventory the unique useful answers, examples, exceptions, downloads and interactions on both pages. Record where each item lives now and where it will live after the proposed change. Do not treat copied wording as sufficient when a condition, date, source or working function is lost. A reviewer should check the complete destination draft before approving retirement.

P01 preserves five required items. P03 remains a content-preparation case because its proposed destination has the migration steps but lacks the legacy field mapping and retention exception. Renaming a heading or adding a redirect would not restore those missing answers.

| P01 required item | Destination section | Acceptance question |
| --- | --- | --- |
| export-permission | #permissions | Does the guide explain which user can export? |
| date-filter | #date-range | Can the reader select the intended records? |
| timezone | #date-range | Are date boundaries interpreted consistently? |
| row-limit | #limits | Is the supported export size still explicit? |
| retry | #recovery | Can the reader recover from the documented failure? |


## When is a canonical tag the right question?

Canonicalization concerns duplicate or very similar content. Google describes canonical annotations, redirects and sitemaps as signals about the preferred URL; consistent signals support that preference. A canonical annotation does not move a visitor to another page or combine the missing content for them.

P04 keeps a useful print view of the same instructions, so it needs a canonical-signal review rather than automatic retirement. P02's plan-choice and API-implementation pages serve distinct tasks; pointing one at the other would skip the underlying content decision. Use the existing noindex/canonical diagnostic when served signals conflict. Do not use noindex as a substitute for selecting a canonical within the same site.

P06 preserves English and French destinations. Google documents separate language and regional annotations for localized versions; similar subject matter does not eliminate their audience need. Review that implementation separately instead of folding a translation into a general consolidation batch.

- [Review Google's canonicalization methods](https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls)
- [Diagnose conflicting served signals](https://rankecho.io/resources/find/issues/noindex-canonical-conflict)
- [Review localized page annotations](https://developers.google.com/search/docs/specialty/international/localized-versions)

## What belongs in an approved consolidation handoff?

P01's proposal maps /guides/csv-export directly to /guides/export-records. It names a final destination, the completed content transfer, affected internal references, deployment checks and the person approving the change. No other pair contributes a retirement mapping. Keep a saved before version and a recovery plan for an implementation error.

For a permanent URL move, Google recommends a server-side permanent redirect and documents 301 and 308 responses. Choose the implementation for the site's real environment. The instructional worksheet does not install a redirect or authorize a change to any current RankEcho URL.

Google's migration guidance advises direct mappings to relevant destinations and warns against unrelated homepage redirects. Update affected navigation and discovery references, inspect old fragments and linked assets, and check any language annotations. If separate pages remain, connect them using descriptive contextual links that help readers move to the next task.

- [Check permanent redirect behavior](https://developers.google.com/search/docs/crawling-indexing/301-redirects)
- [Review URL-mapping checks](https://developers.google.com/search/docs/crawling-indexing/site-move-with-url-changes)
- [Use descriptive crawlable links](https://developers.google.com/search/docs/crawling-indexing/links-crawlable)
- [Prepare links among retained pages](https://rankecho.io/resources/fix/internal-link-cluster)

## How do you verify the change without claiming an SEO win?

First verify the publication decision and delivered behavior. The old URL should return the approved response, the final destination should return useful content with the intended access and indexing state, and all mapped content should survive. Test direct entry, internal links and any important saved fragments or downloads. A destination returning 200 is insufficient if it no longer answers the original task.

The local exercise recomputed all six decisions and the five-item preservation check. It rejects conflicting retirement mappings, redirect chains or cycles, and retirement of a path another pair keeps or leaves unresolved. Changed-input cases stop the proposal when a required fact, evidence pointer, ready destination or reviewed mapping is missing. These are supplied-record checks, not HTTP, browser, CMS or Google validation.

Then observe search and business outcomes separately. Search Console supports date and search-type filters; use comparable query/page observations and record other material changes. Preserve the old-to-new mapping when reviewing later visits and product actions. Do not add incompatible metrics or treat a short-term change as proof that consolidation caused it.

| Stage | Evidence that answers the question |
| --- | --- |
| Decision | Inspected page roles, preservation map and approved scope. |
| Implementation | Correct source response, useful final destination and preserved content/functions. |
| Discovery | Observed crawl/indexing and canonical state, with capture dates. |
| Outcome | Comparable search observations, identified referrals and product actions with explicit units. |

- [Review Search Console performance filters](https://support.google.com/webmasters/answer/7576553?hl=en)

## How can RankEcho support the page proposal?

Bring the chosen destination, inspected overlap, content-preservation map and proposed copy to the sample fix. Evaluate the paid Fix Engine for organizing a bounded page improvement with human review and manual shipment. The team remains responsible for URL retirement, CMS implementation and redirect verification; the product handoff does not promise an automatic site consolidation.

Free audits check Perplexity and Gemini once per prompt. Paid and trialing accounts add ChatGPT, Claude, and Google AI Overviews when configured for the account, for up to 5 engines. RankEcho does not currently run Microsoft Copilot checks.

When the pair should stay separate, the useful change may be a clearer title, a more specific answer or a contextual link. When evidence is missing, collect it before choosing a destructive action. Both are valid outcomes of the review; there is no requirement to merge a page to complete the worksheet.

- [Inspect a sample fix](https://rankecho.io/sample-fix)
- [Evaluate the Fix Engine](https://rankecho.io/product/fix-engine)
- [Refresh a retained page's facts](https://rankecho.io/resources/fix/refresh-stale-product-facts)
- [Separate discovery and conversion observations](https://rankecho.io/resources/prove/citation-to-session-to-conversion)

## Use this pack

Complete the page inspection and preservation review before approving any retirement. Keep distinct reader tasks and language versions available, and mark missing evidence as unknown. The six fictional pairs yield one consolidation proposal, two keep-separate decisions, one content-preparation case, one canonical review and one investigation. The worksheet does not crawl pages, infer intent, verify source documents, execute redirects or establish search and business effects.

Inspect a sample fix — https://rankecho.io/sample-fix. Evaluate the Fix Engine — https://rankecho.io/product/fix-engine for a bounded page proposal with human review and manual shipment. Your team owns content retirement, CMS changes, redirects and outcome analysis. Full-page writing is a separate content add-on — https://rankecho.io/pricing/content, not included with the visibility subscription.