One missing slash created a 12,000-page redirect chain.

One missing slash created a 12,000-page redirect chain.

A documentation site on Gatsby launched with both trailing-slash and non-trailing-slash versions of every URL. The CDN redirected one to the other. Fine — except internal links pointed at the wrong variant, so every navigation triggered a 301, sometimes two.

The symptom: crawl stats in Search Console showed 60% of Googlebot's requests resolving as redirects, and link equity visibly diluting — pages 3+ clicks deep weren't getting crawled at all.

Turns out the JS router emitted internal links without trailing slashes, the CDN canonicalized to with-slash, and a legacy rule occasionally chained http→https→slash. Three hops on some pages. Googlebot follows redirects but discounts them and wastes budget on the hops. Deep pages starved.

We fixed it at the source: router emits the canonical with-slash form, one CDN rule, no chains. Every internal link points at the final URL.

Redirect share of crawl dropped from 60% to under 4% in three weeks. Crawl depth improved — pages that hadn't been visited in months got fetched. Indexed pages rose 23%. Organic +34% over two months.

A redirect isn't free. Multiply it by 12,000 pages and you've quietly rationed your own crawl budget.