Should You Use Meta Keywords for SEO in 2026?
The same question keeps coming round. A client asks it, a new developer asks it, the relative who built the golf club’s website in 2004 definitely asks it. “Shouldn’t we be filling in the meta keywords?”
No. You shouldn’t.
But the longer answer is more useful than the short one, because the reasons tell you where your time ought to go instead. So here’s what Google and Bing have actually said about meta keywords, why keyword stuffing killed the tag off, whether leaving it on old pages does any harm, and what deserves your attention in its place (spoiler, the meta description is near the top of that list).
A hidden line of HTML that got abused to death
You’ll find it in the head of older pages, looking something like <meta name=”keywords” content=”running shoes, trail shoes, waterproof trainers”>. Visitors never see it. The idea, back in the 1990s, was that site owners could tell a search engine what the page was about at a time when search engines were pretty bad at working that out for themselves.
Some engines listened. That was the mistake.
If you could rank just by listing words in a hidden tag, people listed every word they could think of. Competitors’ brand names, celebrities, the same phrase forty times in a row. Meta keywords turned into one of the first big spam playgrounds on the web, and keyword stuffing in that little tag was happening years before anyone fretted about keyword stuffing in body copy. So search engines learned, fairly quickly, not to trust anything a page said about itself in a spot users couldn’t check.
Keyword stuffing killed it. Simple as that.
Google stopped reading meta keywords years ago
September 2009. Google published a post on its webmaster blog with the gloriously blunt title “Google does not use the keywords meta tag in web ranking”, and Matt Cutts, who ran the webspam team at the time, put his name to it.
That line now lives in Google’s own docs. Its page on meta tags Google supports says the meta keywords tag isn’t used by Google Search and has no effect on indexing and ranking. Not a small effect. None at all.
So if you’ve been sat agonising over which ten words to put in meta keywords on a product page, you can stop now. Put that energy into a decent meta description instead. Google reads your visible text, the title, the headings, links pointing at the page and a pile of other signals. That tag isn’t one of them, and it hasn’t been for well over a decade.
And Bing? Slightly more interesting
This is where people get muddled. In October 2011 Search Engine Land reported comments from Duane Forrester, who worked on Bing Webmaster Tools then, and the gist was that Bing did look at meta keywords, but mostly as a spam signal. A tag stuffed with dozens of unrelated terms told Bing something about the person who wrote it, and not in a good way. Danny Sullivan’s write-up put it in the headline. Only spammers should use it.
Yahoo, for what it’s worth, said in October 2009 that it had dropped meta keywords from ranking too. And since Yahoo’s results have come from Bing for years, that one has sort of answered itself.
So across the big engines it’s consistent. No ranking benefit whatsoever, and a small chance that obvious keyword stuffing in the tag makes you look like a cowboy.
Do old meta keywords do any harm?
Not much, honestly. Five accurate words on an old site won’t get anyone penalised, and nobody sane would call that keyword stuffing. Google just skips past the meta keywords line.
Two small catches, though. The first is that anyone can read it. Open a competitor’s page, hit view source, and if they’ve listed their target terms in meta keywords you’ve been handed their keyword research for nothing (plenty of SEO tools scrape these lists automatically, so it’s not even a secret trick). The second is wasted time. Every minute a writer spends picking tags that nobody reads is a minute not spent on the title, the meta description, or the page itself.
My take? Don’t add meta keywords to anything new. If an old site still has them, strip them out next time someone’s in the templates anyway. I wouldn’t run a special clean-up project just for this. Life’s too short.
The one exception is a messy one. If the old tag shows heavy keyword stuffing (forty variations of “cheap plumber Croydon”, say), tidy that sooner rather than later, because that’s precisely the pattern Bing said it treats as a warning sign.
Why the question refuses to die
Partly because a meta keywords box still turns up in odd places.
WordPress is the big one. Yoast SEO used to offer an optional meta keywords field and then removed it in version 6.3. The “focus keyphrase” you see in Yoast now, and the focus keyword box in Rank Math, are internal checks. They tell the plugin what to grade your content against (and a plugin nagging you to repeat a phrase isn’t an invitation to keyword stuffing). They don’t print a keywords tag into your page, and loads of site owners mix the two up (understandably, given the names).
Older CMSs and some ecommerce platforms still show a “keywords” field in the product editor simply because nobody ever got round to deleting it. Some internal site search tools read it too. If your own search box matches products on meta keywords, that’s a perfectly good reason to keep filling them in. Just be clear that you’re doing it for your search box, not for Google.
Titles, the meta description and the tags that matter
Right. If meta keywords are dead weight, where should the effort go?
Start with the title tag. Strictly speaking it isn’t a meta tag at all, but it’s the most important line up there because it usually becomes the blue clickable headline in results. Main topic near the front, readable, unique to each page. Google rewrites titles it thinks are unhelpful, so keyword stuffing a title tends to backfire.
Then the meta description. It doesn’t directly move rankings, but Google may use it as the snippet under your title, which means it’s basically a two-line pitch to someone scrolling on their phone with nine other results in view. Google doesn’t set a fixed length. Snippets often get trimmed somewhere around 150 to 160 characters, though, so lead with the point. What’s on the page? Why click? A meta description that repeats the same phrase four times reads as desperate, and unlike the old tag, keyword stuffing here is visible to every single person who sees the result. Google won’t always use your meta description either (it often pulls a passage from the page that better fits the search), but when it does, a clear one earns its keep. Write a fresh meta description for every page that matters, not just the homepage.
The robots meta tag decides whether a page gets indexed and whether its links get followed. Get it wrong and pages quietly vanish. A leftover “noindex” carried over from a staging site is the classic, and it’s a properly painful one to discover three weeks after launch. The viewport tag tells phones how to scale the layout, and without it mobile visitors get a shrunken desktop page.
Canonical tags tell Google which version is the main one when duplicates exist (URLs with tracking parameters, for instance). Open Graph tags control how a link looks when it’s shared on LinkedIn, Facebook or WhatsApp (many SEO plugins fill the description from your meta description if you leave it blank). They won’t change Google rankings. They do change whether anyone clicks the share.
Put your keyword effort where people can see it
Thing is, the instinct behind meta keywords was never daft. You do want search engines to understand each page. You just do it in the visible bits now, and without the keyword stuffing that wrecked the old tag.
Do proper keyword research first so you know what real people type; our piece on AI for keyword research has a few decent shortcuts. Get the main phrase into the title, the meta description, the opening paragraph and a heading or two, without tipping over into keyword stuffing, and then write the rest so it genuinely answers the question. Related terms will turn up on their own if you’re writing about the subject properly. Link to the page from other relevant pages with anchor text that says what’s there.
One topic per page, with the title, headings and meta description all telling the same story. That’s it. Not glamorous.
If your older content was written around old habits, an audit is a sensible place to begin, and our guide to SEO copywriting and audits walks through what that looks like. A crawler such as Screaming Frog will also spit out every page’s title, meta description and robots tag in one spreadsheet, which makes duplicates and blanks easy to spot. We’ve compared it with its best-known rival in Screaming Frog vs Ahrefs.
Five minutes before a page goes live
Say you’re checking a new page on a Thursday afternoon before it’s published. Is the title unique, and does it describe the page? Was the meta description written for a human, and is it different from every other page’s (a duplicate meta description across dozens of URLs is one of the most common things a crawl turns up)? Does the robots tag allow indexing? Does the canonical point to itself or to the right main version? Will the Open Graph image look decent in a share, or will it pull some random logo from the footer? And if there’s still a meta keywords line in the template, take it out or leave it empty. Then read the copy out loud once. If it sounds like keyword stuffing to you, it’ll sound that way to a reader.
Five minutes. Covers everything that actually affects how the page looks in search and on social.
So, my honest view. Meta keywords belong to a version of the web that doesn’t exist any more, and every hour spent on them is an hour stolen from the title, the meta description and the content, which are the things people read. Google ignores meta keywords. Bing has mostly treated the tag as a red flag. The big WordPress plugins don’t even print it.
Leave it to the history books. And keep keyword stuffing out of every other corner of the page while you’re at it.
Want a hand getting the basics right across your whole site? Take a look at our SEO services.
