Fast page, slow content, missing rankings

Fast page, slow content, missing rankings

A jobs board scored 98 on performance audits and couldn't rank for a single job title. The team was baffled — the site felt instant.

The data told a split story: the HTML document arrived in 180ms, gloriously fast. But the actual job listings appeared 4.5 seconds later, after a chain of three client-side API calls fired in sequence.

The investigation watched the render timeline. Google's rendering has a budget. When it snapshotted these pages, the first paint had arrived but the listings hadn't — the third API call in the waterfall was still pending. The 'fast' score measured the empty shell; the content that mattered showed up after the snapshot.

Turns out a high performance score on an empty document is a lie when the content depends on a serial fetch chain. Speed-to-blank-page isn't speed.

The fix: collapse the three sequential calls into one server-side data fetch, ship listings in the initial HTML, and hydrate from there.

Result: listing pages went from 0 ranked queries to 1,900 ranking terms in eight weeks, organic applications up +210%. The performance score barely moved — but now it measured something real.