Google clarifies URL separators: use hyphens for new slugs, plan changes to old ones

Google logged a clarification to its URL guidance on October 23, 2026, explaining why it recommends hyphens rather than underscores between words. For publishers creating new article addresses, the practical choice is straightforward: use readable words separated by hyphens. For an established archive, changing every address deserves a separate decision.

The Search Central change log describes a documentation clarification aligned with Google's video series, The Non-Canonical Cabinet. It does not announce a ranking-system change or attach a traffic forecast to the advice.

What the separator tells a reader and a crawler

Google's URL structure guidance explains that hyphens make word boundaries clear. Underscores have historically joined concepts that belong together, including function names in programming. The recommendation concerns how an address communicates its meaning.

Consider a hypothetical new article with the slug garden-tool-storage. Its words are separated visibly, unlike garden_tool_storage or gardentoolstorage. This example illustrates a naming choice; it is not a test of ranking performance.

The useful place to act is the publishing template. Check what the content management system produces when an editor saves a new article, and whether imported titles follow the same convention. A written slug rule is more useful when the publishing workflow actually applies it.

An existing URL is a migration decision

Renaming a published address introduces work that a new draft does not. Google's site migration guidance calls for mapping old URLs to new ones, setting redirects and monitoring the move. It also warns that rankings can fluctuate while pages are recrawled and reindexed.

Our editorial recommendation is to separate those decisions. Apply a clear convention to new pages now, then evaluate any archive-wide rename against its operational cost and other reasons for changing the site. The October 23 clarification alone does not establish a measurable benefit for your particular archive.

If a migration is already planned, add separator consistency to its checklist. Before launch, review canonical URL handling so that the preferred address, internal links and redirect plan do not contradict one another.

Check the pattern, not just one slug

Sample a new article, an imported article and a category page. Record which system generates each address and who can change that rule. Keep this check distinct from controlling faceted navigation, where combinations of filters can create a much larger crawling problem.

A successful review can end with no changes to existing URLs. The concrete result is a consistent rule for future publishing and a documented reason for any migration, rather than a bulk rename based on an assumed search reward.