Define the job before naming another category
Before adding a category, write the question its archive will help a reader answer. “Articles about websites” is too broad to explain why another collection is necessary. A useful brief names the audience, the selection rule and the next action after browsing. Ask an editor to identify an existing collection that could already perform that job.
This is a planning exercise, not a diagnosis of a ranking penalty. Similar category names or shared articles do not establish harmful competition. Two collections can legitimately serve different needs. The decision is whether the proposed archive provides a distinct, maintainable way to find resources, rather than merely another label for the same list.
Build a comparison that includes the membership rule
Record each candidate archive URL, language, title, introductory text, included resources and inclusion rule. Compare the full collection, including later pages where applicable, rather than just the first screen. A changing publication date can make two equivalent collections appear different on one visit. Save the comparison date and any inventory limitations.
Add a reader-task column and an owner column. For example, a troubleshooting collection may accept only diagnostic procedures, while a maintenance collection includes recurring prevention tasks. Some guides can belong to both. Ask whether an editor could consistently place the next five proposed articles without guessing. If the rule cannot guide future decisions, clarify it before building another archive.
Choose among keeping, refining and consolidating
Keep both archives when their tasks and selection rules are useful and meaningfully different. Refine labels and introductions when the distinction exists but is poorly communicated. If the collections duplicate the same job and material, choose a maintained destination and document how the redundant route and its existing links will be handled. Preserve access to useful resources throughout the change.
Google describes canonicalization for duplicate or very similar pages; it is not a substitute for designing distinct categories. For an actual duplicate being retired, consider a permanent redirect to its equivalent destination. Keep internal links and canonical signals consistent with the chosen URL. Do not point every different archive at one broad category solely because they share some articles.
Illustrative example: two names for one collection
Imagine a small agency has “Site care” and proposes “Website maintenance.” Both would list the same twelve practical maintenance guides, with the same audience and no different selection rule. The editor records that the new archive would add a synonym, not a new browsing task. The proposal is withdrawn and the existing introduction is clarified instead. These are fictional planning details, not a reported customer result.
Now suppose a proposed “Before a redesign” collection selects only preparation resources across several categories. Its task is preparing a handover, not routine upkeep, and its membership rule is documented. The team can retain that distinct collection even if three guides overlap. The decision rests on reader purpose and curation, not an arbitrary maximum overlap percentage.
Validate the decision and prevent its recurrence
Ask a colleague to choose an archive for two concrete tasks without seeing the internal brief. If the labels lead to the wrong collection, revise them and repeat the exercise. Where routes changed, check the old destination, new destination and links from navigation or articles. Confirm that each expected resource remains reachable and that language editions have not been accidentally combined.
Save the final decision, owner and inclusion rule in the editorial inventory. Review new category requests against that inventory before publishing them. A crawl or SEO audit can help locate archive links and technical inconsistencies, but editorial judgment is still needed to evaluate purpose. This process produces a defensible structure; it does not promise a ranking increase or immediate Google canonical selection.