Glossary HTTP
HTTP (Hypertext Transfer Protocol) is the protocol a client, such as a browser or a search engine crawler, uses to request a resource from a web server and receive the response. Every response carries a status code that says whether the request worked.
HTTP is the protocol browsers and crawlers use to request pages: how it works and what search engines read in its status codes.
HTTP stands for Hypertext Transfer Protocol. Its semantics, meaning what the methods, status codes and headers stand for, are set out in RFC 9110, published by the IETF in June 2022. Several versions run on those same semantics: HTTP/1.1, HTTP/2 and HTTP/3.
HTTPS uses the same methods, codes and headers as HTTP, but the connection is encrypted with TLS, so nobody between the browser and the server can read or alter what is sent. Today it is the standard on the web.
Every HTTP exchange is a request and a response. The request names a method, such as GET to read a page or POST to submit a form, the address of the resource and headers with extra information, for example the preferred language or the type of browser.
The response starts with a three-digit status code. Codes beginning with 2 mean success, such as 200; with 3, a redirect, such as 301; with 4, a client error, such as 404 for a page that does not exist; and with 5, a server error, such as 500. Then come the headers, which say, for example, how long the response may be cached, and finally the content.
Versions change how the data travels, not what it means. HTTP/2 sends many requests at once over a single connection, and HTTP/3 runs over QUIC instead of TCP, which cuts waiting times on unstable networks.
For a search engine, the status code is the first thing it reads in each response, and it decides whether the content gets used at all. Google explains that a 301 redirect is a strong signal that a page has moved permanently and a 302 a weak one; that a URL returning 404 or 410 drops out of the index if it was indexed; and that 5xx errors and the 429 code make its crawlers slow down. If a server error persists, the URL is eventually removed from the index too.
An error page that responds with 200 confuses the search engine: Google may treat it as a soft 404. And every redirect in a chain adds one more request before the content is reached. Google's crawlers use HTTP/1.1 and HTTP/2, with HTTP/1.1 as the default, and pick the version that gives them the best crawling performance.
We measured this on zds.es on 13 October 2026. The site responds over HTTP/2 and announces HTTP/3 in the alt-svc header. The variants of an address, with http instead of https, with www, with a trailing slash or in capital letters, return a 301 to the canonical version, and an address that does not exist generally returns a real 404. We found one exception: in German and English, a made-up address that starts like most entries in this glossary, with "was-ist-" or "what-is-", redirects with a 301 to the glossary's main page. A 301 tells the search engine that the page has moved, even though it never existed.
One request, and what Google does with the response.
GET /en/what-is-http HTTP/1.1200successuses the content301permanent redirectstrong signal of a permanent move302temporary redirectweak signal404
410page does not existthe URL drops out of the index if it was indexed5xxserver errorslows down crawling; if the error persists, the URL is removed from the index
A technical audit covers this and everything else in one pass.
Request an auditThe content and the codes are the same. HTTPS is HTTP encrypted with TLS, so nobody between the browser and the server can read or change what is transmitted. Browsers mark pages served without HTTPS as "not secure", and today HTTPS is the standard for any website.
It is the three-digit number every server response starts with. 2xx means success, 3xx a redirect, 4xx a client error, such as a page that does not exist, and 5xx a server error. Search engines use that code to decide what happens to the URL.
A page that tells the user something does not exist but answers the search engine with a 200 success code. Google may spot it from the content and treat it as an error, but in the meantime it spends crawling on it. The right answer is a real 404 or 410.
Google does not name it as a ranking factor, but it usually speeds up loading, because many requests travel at once over one connection. According to Google's documentation, its crawlers use HTTP/1.1 by default and switch to HTTP/2 when that gives better performance.
If the server returns 5xx errors, Google's crawlers temporarily slow down crawling and gradually speed up again once it responds successfully. If the error persists on a URL, Google eventually removes it from the index. A short outage is therefore much less of a risk than an error that lasts for days.