The difference between HTTP1 and HTTP2 matters to SEO in a narrower way than most hosting sales pages suggest. When Google started crawling over HTTP/2, plenty of people read it as a signal that the protocol would start to count in rankings. It does not. The change was about efficiency on both ends of the crawl.
That still makes HTTP/2 worth having. It just means the case for it rests on page delivery and server load, and you should judge it on those terms.
What Google said about HTTP/2 crawling
The clearest summary of Google’s position is in Search Engine Journal’s report Google will start crawling over HTTP/2, written by Matt G. Southern in September 2020. On the ranking question, Google’s answer was blunt: “No, there is absolutely no ranking benefit.”
The reason Google gave for the change was resource use. In its words: “With h2, Googlebot is able to open a single TCP connection to the server and efficiently transfer multiple files over it in parallel, instead of requiring multiple connections. The fewer connections open, the fewer resources the server and Googlebot have to spend on crawling.”
Two further points from the same report settle most of the questions site owners ask. Forcing Googlebot to crawl over HTTP/2 “is not possible”. Google decides to crawl over h2 only when your server supports it and doing so benefits both sides. And if you would rather it did not, there is an opt-out: have the server respond with a “421 HTTP status code when Googlebot attempts to crawl your site over h2.”
The difference between HTTP1 and HTTP2 in practice
HTTP/1.1 handles requests on a connection one after another. To fetch many files at once, a browser or crawler opens several connections in parallel, and each one carries its own setup cost on both client and server. HTTP/2 multiplexes: many requests and responses share a single connection at the same time.
The table below sets out the points that matter for search, stated generally. Real-world gains depend on your pages, your host and your visitors’ networks, so treat it as a description of behaviour rather than a promise of a speed figure.
| Question | HTTP/1.1 | HTTP/2 |
|---|---|---|
| Parallel file transfer | Needs multiple connections | Multiple files in parallel over one connection |
| Cost of crawling to your server | More connections open | “The fewer connections open, the fewer resources” spent |
| Ranking benefit | None | “No, there is absolutely no ranking benefit.” |
| Can you force Googlebot to use it? | Not applicable | “Not possible”; Google chooses when it benefits both sides |
| Can you opt out? | Not applicable | Yes, by returning a 421 status to h2 crawl attempts |
The supporting question people search, the difference between HTTP/1.1 and HTTP/2, has the same answer. HTTP/1.1 is the version most of the web ran on for years; the step to HTTP/2 is about how requests travel, not what the page contains. Your HTML, your canonical tags and your content are identical under either protocol.
Where HTTP/2 helps SEO indirectly
With rankings ruled out, the useful effects are indirect, and they come in two places.
- Crawling. Fewer open connections means less work for your server per crawl. On a large site where the server is the constraint, that is relevant to how Google treats your crawl budget, because crawl rate responds to how well the server copes.
- Page delivery. Pages that pull in many stylesheets, scripts and images benefit from sending them over one multiplexed connection. Faster delivery is good for visitors, and visitors are what rankings are meant to reward.
Neither effect is a lever you can pull for a predictable result. HTTP/2 does not make a slow application server fast. If the first byte takes a long time to arrive, the protocol carrying it will not fix that; our guide to time to first byte covers where that delay usually comes from.
It also helps to remember how a crawler fetches a page. As our explainer on what a web crawler is sets out, a crawler requests the HTML and then the resources it needs to render it. Those resource requests are where a single shared connection saves work.
How to enable HTTP/2 and check it is working
On most modern hosting, HTTP/2 is already on for HTTPS traffic. Browsers only use it over encrypted connections, so if your site is not yet on HTTPS, that comes first. Check your host’s control panel, or the configuration of your web server, for an HTTP/2 setting.
If you use a CDN, the protocol between the visitor and the CDN edge is set by the CDN, and it is usually HTTP/2 or newer by default. The connection from the CDN back to your origin is a separate setting. Our guide to CDNs and SEO covers what else a CDN changes for crawling.
To confirm what visitors get, open the browser’s developer tools, go to the network panel and enable the protocol column. Requests served over HTTP/2 show as h2. From the command line, curl reports the protocol version in the response:
A first line beginning HTTP/2 means the server negotiated it. For Googlebot specifically, your server access logs show the protocol of each request if the log format records it.
When the 421 opt-out makes sense
Most sites should leave Googlebot’s h2 crawling alone. Google already chooses not to use it where it would not help, so the default is designed to be safe.
The opt-out exists for edge cases: a server whose HTTP/2 implementation misbehaves under crawler load, or an infrastructure team that needs time to investigate a problem. In those cases, returning a 421 to Googlebot’s h2 attempts is the documented way to stop it. Treat it as temporary, fix the underlying issue, and remove the rule.
What to take from Googlebot’s switch
The practical reading is short. Turn HTTP/2 on because it is the better protocol for your visitors and cheaper for your server to be crawled over. Do not expect a ranking change after you do, and do not trust anyone who promises one, because Google has stated plainly that there is none.
If you are deciding where to spend technical effort, HTTP/2 is usually a setting rather than a project. Confirm it is on, confirm your CDN and origin agree, and move on to the things that change how quickly a page becomes useful: server response time, render blocking and image weight.
One limitation is worth stating. Nothing published by Google quantifies how much crawling capacity a given site gains from h2, and the effect will vary with how many resources each page loads and how the server is built. A small site that Google already crawls comfortably may see no difference at all in its logs. Judge it on your own server metrics rather than on general claims.
Frequently asked questions
Is HTTP/2 a Google ranking factor?
No. Asked whether crawling over HTTP/2 brings a ranking benefit, Google answered: “No, there is absolutely no ranking benefit.”
What is the difference between HTTP1 and HTTP2 for crawling?
HTTP/1.1 needs multiple connections to fetch files in parallel. HTTP/2 lets Googlebot open a single TCP connection and transfer multiple files over it in parallel, which uses fewer resources on both sides.
Can I make Googlebot crawl my site over HTTP/2?
No. Forcing it is “not possible”. Google crawls over h2 only when your server supports it and it benefits both sides.
How do I stop Googlebot crawling over HTTP/2?
Have your server respond with a 421 HTTP status code when Googlebot attempts to crawl over h2.
The takeaway HTTP/2 lowers the cost of crawling and helps page delivery, but Google says it carries no ranking benefit. Enable it, check it, and spend your effort elsewhere.
