The internal links that didn't exist until you hydrated

The internal links that didn't exist until you hydrated

A large wiki-style site found Google crawling its pages but barely following links between them. New articles took months to get discovered, despite heavy internal linking.

The symptom: crawl depth was shallow. Google entered via the homepage and a few hubs, then stalled. Deep articles sat undiscovered for weeks.

The investigation looked at the navigation in raw HTML. The main menu and related-article links rendered as <span> and <div> elements with JavaScript onClick handlers that did programmatic navigation after hydration. There were almost no real <a href> tags in the server HTML. Google parses the raw response for links to queue — and found a page that looked like a dead end.

Turns out 'clickable divs with router pushes' are invisible to link discovery. A crawler needs a real href to follow, present before any JavaScript runs.

The fix: render every navigational element as a true anchor with an href in the server HTML, letting the router progressively enhance the click.

Result: average crawl depth doubled, new-article discovery time dropped from ~7 weeks to under 5 days, and indexed pages grew +34% over a quarter as the link graph finally became followable.