301 Redirect Generator

Chains are the thing that quietly costs you

A chain is /a → /b → /c. Every browser follows it, so nothing looks broken, and that is exactly why it survives three migrations. Google follows up to about ten hops before giving up, so the page still gets found. What you lose is time and certainty: every hop is another request, another chance for a rule to be dropped, and another layer between the link someone built to /a six years ago and the page that link is meant to be pointing at.

The fix is not to delete the old rules. It is to rewrite each one to point at the final destination, keeping every source URL alive. Tick "flatten chains" above and this tool does that for you. /a → /c and /b → /c, both still there.

The four errors worth finding before deploy

• A loop. /a → /b and /b → /a. The browser gives up after about twenty hops and shows an error page. This is a hard outage on those URLs and it usually happens when two people edit the same list.
• A duplicate source. The same old URL listed twice with different destinations. Every server takes the first match and silently ignores the second, so half your rules do nothing and you have no idea which half.
• A URL redirecting to itself. An instant loop, and the most common cause is a trailing slash the server normalises away before the rule is even evaluated.
• A trailing-slash-only change. /services/ → /services looks harmless. If your server or CDN also normalises slashes in the other direction, the two fight each other and you get a loop that only appears in production.

Where redirects actually leak value

Google has been clear for years that a 301 passes essentially all ranking signals, so the old "15% loss per hop" folklore is not a thing to plan around. The real leaks are duller. Redirecting everything to the homepage instead of the closest matching page — Google treats a mass redirect to an unrelated page as a soft 404 and the signals go nowhere. Forgetting to redirect the URL variants: http, www, trailing slash, uppercase. Leaving redirects in place for a year and then removing them because the list "looks old", after the CMS has stopped serving the old URLs anyway.

Redirects are permanent infrastructure. Keep them for at least a year, and if the URL had links pointing at it, keep them forever.

What each format actually matches

These are not equivalent and the differences bite. Apache's Redirect is a prefix match, so Redirect 301 /blog /news also moves /blog-archive. Nginx's location = is exact but ignores the query string entirely. Netlify matches the path and passes the query string through by default. Cloudflare Bulk Redirects need absolute URLs on both sides. This tool warns when a rule you have written is going to behave differently from how it reads.