Reference
HTTP vs HTTPS: What Is the Difference?
HTTP and HTTPS are the same protocol. HTTP sends every request and response as plain text, which anyone on the network path can read or quietly alter. HTTPS sends exactly the same messages inside TLS encryption, so the contents are scrambled in transit, tampering gets detected, and the server has to prove it really owns the domain.
What HTTP actually does
HTTP, the HyperText Transfer Protocol, is the language your browser and a web server use to talk. Your browser sends a request: a method (GET to fetch something, POST to submit something), a path, and some headers describing what it wants. The server sends back a response with a three-digit status code and, usually, a page.
It is text, literally. A raw HTTP request is a few readable lines you could type by hand, and the reply is the same. The current specification, RFC 9110, was published in June 2022 and defines what those messages mean separately from how each version writes them down, which is why HTTP/1.1, HTTP/2 and HTTP/3 all share the same vocabulary of methods and status codes while shifting bytes around differently.
The protocol grew out of the web that Tim Berners-Lee built at CERN, and the wider story of how it spread is on our page about the internet age. It was designed for sharing physics papers between trusted machines. Nobody was thinking about card numbers on café wifi.
And that is the problem with plain HTTP. Between you and the server sits a chain of equipment you have never inspected: the router in the coffee shop, your ISP, whatever networks the traffic crosses on the way. On HTTP, every one of them can read the page you asked for and the details you typed into it. Any of them can also change the reply before it reaches you, inserting adverts or malicious scripts, with nothing to give the swap away.
What the S adds: TLS
HTTPS is HTTP running inside TLS, or Transport Layer Security. TLS is the descendant of SSL, which is why people still say “SSL certificate” for something that has been TLS for years. The current version, TLS 1.3, is RFC 8446, published in August 2018.
RFC 8446 lists what a TLS connection guarantees, and the items are worth separating because people collapse them all into “it’s encrypted”.
Confidentiality is the obvious one. Once the connection is up, only the two endpoints can read what crosses it, and an observer sees noise. Integrity is the quieter guarantee: nobody in the middle can modify the data without the change being detected, so your bank’s page arrives as your bank wrote it.
Authentication is the one people forget, and it does most of the real work. The server has to prove it is the server for that domain, which is what stops an attacker from sitting in the middle and impersonating the site. Encryption on its own would not give you this, and encrypting a conversation with an impostor buys you nothing.
The proving happens through certificates. Before any page data moves, the browser and server run a handshake. The server presents a certificate: a file saying “this public key belongs to example.com”, signed by a certificate authority. Your browser ships with a list of authorities it trusts, and it checks three things. Does the certificate cover the domain in the address bar? Has it expired? Does the signature chain back to an authority on the trusted list? Fail any of those and you get a full-page warning rather than a page.
Pass, and the two sides use that verified key to agree a fresh shared secret for this connection alone. Everything after that, including the URL path, the headers, the cookies and the page itself, is encrypted with it.
Certificates used to cost money and take a phone call. Let’s Encrypt and the ACME protocol made them free and automatic, which removed the last practical excuse for staying on HTTP.
What does HTTPS mean?
Literally: HyperText Transfer Protocol Secure. The S is just “secure”, and secure here means one specific thing, that the connection runs inside TLS.
In practice, when you see https:// in the address bar, you know that what you send to that server and what it sends back cannot be read or edited by anyone in between, and that the server presented a valid certificate for that exact domain. That is a claim about the pipe. It is not a claim about the company at the other end, which brings us to the padlock.
What the padlock does not tell you
The padlock means the connection is encrypted. It does not mean the site is honest.
This trips up an enormous number of people, and the security industry spent years accidentally teaching them to be tricked. “Look for the padlock” was the standard advice, and the implied promise was safety. What the padlock actually certifies is that you have a private, unmodified line to whoever registered that domain. If the person who registered it is a criminal, you have a private, unmodified line to a criminal.
Phishing sites run on HTTPS as a matter of course. Certificates are free and issued in seconds to anyone who can prove they control a domain name, and control is exactly what a scammer who has just registered paypa1-secure-login.com has. So the padlock duly appears, and it is not lying. It is answering a narrower question than most people think they are asking.
Read the domain, not the icon. Google seems to have reached a similar conclusion: in the 2018 announcement that Chrome would start labelling HTTP pages, it said the eventual goal was to stop marking HTTPS pages as secure at all, leaving the safe case unmarked and warning only about the dangerous one.
What HTTPS still leaks
Encryption hides the contents. It does not hide the fact of the conversation, and it is worth being precise about which parts stay visible.
Start with the domain you visited, which historically has not been protected at all. Your device looks up the address first, and unless you use encrypted DNS that lookup is public. The TLS handshake then names the host it wants to reach in the clear, in a field called SNI, so that a server hosting many sites knows which certificate to present. Encrypted Client Hello is the fix. It encrypts that opening message, hiding the hostname and the protocols being negotiated, and it became a standards-track specification, RFC 9849, in 2026. Browsers have shipped support, but it only helps when the server offers it and the client is also using encrypted DNS, so coverage is partial.
The server’s IP address stays visible regardless. On a large shared host that narrows you down to thousands of possible sites. On a dedicated one it identifies the site outright.
Size and timing leak too. TLS 1.3 conceals neither how much data moves nor when, and traffic patterns can be distinctive enough to guess which page someone loaded. That is an active research area rather than something happening to you on the train.
What is hidden: the path, the query string, the headers, the cookies, the form contents, the page. An observer can often tell you visited a hospital’s website. They cannot tell which condition you read about.
Why every site moved
The pressure arrived in stages, and search was only the first of them.
Google announced in August 2014 that HTTPS was a ranking signal. The company was blunt about the size of it: a very lightweight signal, affecting fewer than 1% of global queries, and outweighed by things like high-quality content. It also said it might strengthen the signal later. Whether it ever did is not something Google has spelled out, and anyone claiming a precise figure today is guessing.
Then browsers got impatient. Chrome 68 shipped in July 2018 marking every HTTP page “not secure” in the address bar, by which point Google reported that 76% of Chrome traffic on Android was already encrypted, up from 42%, and 83 of the top 100 sites defaulted to HTTPS. Site owners could argue with a ranking nudge. A permanent warning sitting beside their brand name was harder to ignore.
The last bit of pressure is quieter and probably did more than either. The HTTP/2 specification permits cleartext, in a mode called h2c, but no major browser implements it, so in practice HTTP/2 in a browser means TLS. HTTP/3 removes the option altogether, because it runs over QUIC, and QUIC builds TLS 1.3 into the transport itself. There is no unencrypted HTTP/3 for anyone to implement. Staying on HTTP means giving up every performance improvement of the last decade.
Mixed content
A page served over HTTPS that pulls an image, script or stylesheet over HTTP is mixed content, and it undermines the whole point: an attacker who can rewrite the HTTP script can rewrite your page.
Browsers stopped tolerating it. Active mixed content, meaning scripts and stylesheets, is blocked outright, so the resource simply does not load. Passive content like images gets upgraded to HTTPS automatically where possible and blocked where that fails. The usual cause is a hard-coded http:// URL in an old template or database row. If a page looks broken after a migration, this is the first place to look.
Moving a site properly
Get the certificate installed and both versions serving, then do two things in order.
First, redirect. Every http:// URL should 301 to its https:// twin, same path, permanently. Skip this and you have published two copies of your site. Crawlers will find both, index both, and split your links and ranking signals across them, and Google Search Console will treat them as separate properties reporting separate numbers. It is a self-inflicted duplicate-content problem, and the fix is one redirect rule.
Second, once you are confident the HTTPS side is solid, add HSTS. The redirect still leaves a gap: the first request goes out over HTTP before it gets bounced, which is the moment an attacker can hijack. HSTS tells the browser to never use HTTP for your domain again, so subsequent visits skip the redirect entirely. Do it last, and mean it, because once a browser has the policy it will refuse to load your site over HTTP at all until the policy expires.
Check any site
Our free HTTP status and redirect checker shows the redirect chain for any URL, so you can see whether http:// actually lands on https:// and whether it does so in one hop or four. It reports the security headers too, HSTS included. Point it at your own site before you assume the migration went the way you think it did.
Frequently asked
What is the difference between HTTP and HTTPS?
They are the same protocol. HTTP sends every request and response as plain text, so anyone on the network path can read or alter it. HTTPS wraps that traffic in TLS encryption, which scrambles the data, detects tampering, and makes the server prove it owns the domain with a certificate.
What does HTTPS mean?
HTTPS stands for HyperText Transfer Protocol Secure. The S means the connection runs inside TLS, the encryption layer that protects data in transit. When your browser shows an https:// address, the pages, form fields and passwords you send are encrypted between your device and the server.
Is HTTP safe to use?
Not for anything you would mind a stranger reading or changing. On plain HTTP, whoever runs the network can see every page you load and edit what comes back, including injecting scripts. Chrome has labelled HTTP pages "not secure" since 2018. Treat an HTTP login form as compromised.
Does HTTPS help SEO?
A little, directly. Google made HTTPS a ranking signal in August 2014 and called it very lightweight, affecting under 1% of queries and outweighed by things like content quality. The real benefit now is avoidance: browser warnings scare visitors off, and a botched migration splits your ranking signals across two sites.
Sources
- RFC 9110: HTTP Semantics, IETF (June 2022)
- RFC 8446: The Transport Layer Security (TLS) Protocol Version 1.3, IETF (August 2018)
- RFC 9849: TLS Encrypted Client Hello, IETF (2026)
- HTTPS as a ranking signal, Google Search Central (August 2014)
- A milestone for Chrome security: marking HTTP as 'not secure', Google (July 2018)
- Enable HTTPS on your servers, web.dev