Redirect latency is the time between a user clicking a shortened URL and the destination page beginning to load. This delay directly impacts user experience: even a few hundred milliseconds of additional page load time has been shown across numerous performance studies to lower conversion rates and increase abandonment.
The total latency is the sum of several steps: DNS resolution (looking up the short URL domain, 20-120 ms), TCP connection (establishing a connection to the server, 10-50 ms), TLS handshake (setting up HTTPS encryption, 30-100 ms), server processing (fetching the redirect target from the database, 1-50 ms), and response transmission (returning the 301/302 response, 5-20 ms). A typical total falls between 70 and 340 ms.
Major URL shortening services optimize redirect latency to 50-100 ms through CDN edge caching. If you build your own service, an in-memory cache like Redis can bring server processing below 1 ms.
Redirect chains - multiple sequential redirects - multiply latency. A chain of short URL to another short URL to the final page doubles the delay. Google recommends avoiding redirect chains, and from an SEO perspective a single direct redirect is always preferable.
You can measure redirect latency with curl's -w option (which prints timing for each phase), Chrome DevTools' Network tab (which shows detailed redirect timing), or WebPageTest (a third-party performance tool).