How a meta refresh works
A meta refresh is an instruction placed in the head of a web page telling the browser to load a different address after a set number of seconds. Set the delay to zero and it happens as soon as the page arrives; set it higher and the visitor sees the first page for a while before being moved.
The crucial difference from a proper redirect is where the decision is made. With a 301, the server answers the request with a redirect and no page is ever sent. With a meta refresh, the server sends a complete, successful page, the browser reads it, and only then does the move happen. The redirect is content, not a response.
That is why it survives in places where nobody can change the server: old content management systems, hosting without redirect rules, marketing pages built by someone with access to the page but not the configuration.
Why meta refresh matters
Search engines can follow a meta refresh and will usually work out what was intended, but they have to load and parse the page to discover it, and the signal is weaker and slower than a server redirect that arrives before any content. Where a permanent move is meant, you are relying on interpretation rather than instruction.
It is also worse for people. A delayed refresh moves the page while someone is still reading, which is disorienting and is one of the recognised accessibility failures — anyone using a screen reader, or reading slowly, or on a connection that made them wait for the first page, gets pulled away without warning. On a slow mobile connection it also means paying to download a page that was never meant to be read.
Common mistakes with meta refresh
The first is using it for a site migration. Hundreds of pages redirected this way still answer successfully, so a crawl report looks healthy while every old URL is quietly asking search engines to guess. Migrations need server redirects and a mapped list.
The second is the delay. If a refresh has to exist at all, a delay serves nobody: it shows a page the visitor cannot use and traps the back button, because going back lands on the page that immediately moves them forward again.
The third is combining it with other hops — a meta refresh that lands on a page that then redirects again. Each stage costs a page load, and on a poor connection that is a real wait rather than a technicality.
What to do about it
Replace it with a server-side redirect wherever you can. In WordPress and most modern platforms a redirect plugin or a hosting rule does the job in a minute, and the result is faster, clearer and easier to audit.
Where you genuinely cannot reach the server, set the delay to zero, point straight at the final destination rather than another redirect, and include a visible link so anyone whose browser does not move them can continue by hand. Then find them all: crawl the site and check for pages that answer successfully but send visitors elsewhere, the same sweep that catches a JavaScript redirect. Clearing them out is small, cheap technical SEO work with a visible effect on speed.
