The whole nav worked. None of it was a link.

The whole nav worked. None of it was a link.

A help-center SPA built its navigation with handlers that called the router programmatically. Worked flawlessly for users. Googlebot found exactly the URLs in the sitemap and not one page beyond them.

The symptom: 1,200 articles in the sitemap, but internal-link reports showed almost no internal links between articles, and 'related articles' pages — the deepest, most specific content — were barely crawled.

Turns out there wasn't a single real in the article navigation. Every link was a click handler on a div. Googlebot doesn't fire arbitrary click handlers to discover URLs — it follows hrefs. The site had no crawlable link graph at all. Pages existed only because the sitemap listed them; PageRank flowed nowhere internally.

We rebuilt navigation as actual anchors that the router intercepts for client-side transitions — same SPA feel, but real links underneath.

Google discovered a proper internal link graph over five weeks. Crawl depth improved, related-content pages got indexed, and internal PageRank started flowing to deep articles. Average position across the help center improved from 21 to 13. Organic +44% over the quarter.

A div with an onClick is not a link. The crawler needs an href to follow.