The Complete Guide to SEO Meta Tags

Most meta tags do nothing. This guide covers the ones that genuinely affect how your pages are crawled, indexed and displayed — and how each of them actually gets used.

Updated: 2 August 2026Read: 12 minRuns: 100% in your browser

Which meta tags actually matter

There are dozens of meta tags in circulation and most of them do nothing. Here is the honest breakdown.

Tag	Impact	Notes
`<title>`	High	A genuine ranking factor and the largest element in the snippet
`meta description`	Indirect	Not a ranking factor; drives click-through rate
`meta robots`	High	Controls indexing and snippet behaviour outright
`link canonical`	High	Consolidates duplicate URLs
`link hreflang`	High	Selects the right locale version
`meta viewport`	High	Required for mobile usability
`meta charset`	High	Prevents the page rendering as garbage
Open Graph / Twitter	Indirect	Controls how shared links appear; drives social traffic
`meta keywords`	None	Ignored by Google since 2009
`meta author`, `revisit-after`, `rating`	None	No search engine uses them

Generate a complete, correct head block with the meta tag generator.

Writing title tags

The title is the strongest on-page signal you control and the biggest text in the search result. Get it right and everything else is easier.

Length is measured in pixels, not characters

Google truncates at roughly 580 px on desktop and 920 px on mobile. Since characters differ in width, “60 characters” is only a rough guide: a title full of capital W’s truncates far sooner than one full of i’s. Check the real rendering in the SERP preview tool.

Structure

Primary Keyword - Secondary Context | Brand

Text to Binary Converter - Free Online Tool | TextToBinary.net
Best Running Shoes for Flat Feet (2026 Tested) | RunLab

Front-load the term people search for. Truncation removes the end, and users scan the first few words.

Google rewrites titles

Since August 2021 Google rewrites the displayed title on a substantial share of results, drawing from your H1, anchor text pointing at the page, or on-page headings. It does this most often when the title is too long, keyword-stuffed, boilerplate across the site, or a poor match for the query. The defence is a concise, accurate, unique title — not a longer one.

Common mistakes

Identical titles across many pages — every page needs its own.
Keyword stuffing: Shoes, Buy Shoes, Cheap Shoes, Shoes Online.
Leading with the brand on every page, wasting the most valuable position.
Leaving the CMS default, such as Home | WordPress Site.

Writing meta descriptions

The meta description has not been a ranking factor since 2009. It matters because it is advertising copy in the search result, and click-through rate affects traffic directly.

Google rewrites descriptions even more often than titles — studies consistently find 60–70% of displayed descriptions are generated from page content rather than the tag. It does this when the page text answers the query better than your description does, which is usually reasonable behaviour.

Write one anyway. When Google does use it you control the message, and when it does not you have lost nothing.

Aim for 140–155 characters; mobile truncates around 120.
Front-load the value proposition.
Include the primary term naturally — matched query words are bolded, which draws the eye.
Add a concrete differentiator: a number, a date, a free/no-signup claim.
Never duplicate across pages. A duplicate description is worse than none.

Count characters precisely with the word counter, which has a meta description preset.

Robots directives

The robots meta tag controls what search engines do with a page they have already fetched. This is a different question from robots.txt, which controls whether they fetch it at all.

<meta name="robots" content="index, follow">              default
<meta name="robots" content="noindex, follow">            do not index, follow links
<meta name="robots" content="noindex, nofollow">          exclude entirely
<meta name="robots" content="max-snippet:-1, max-image-preview:large">

The classic mistake: blocking a page in robots.txt and adding noindex. Because the crawler is not allowed to fetch the page, it never sees the noindex, and the URL can still appear in results as a bare link. To remove a page from the index, allow crawling and use noindex. Once it has dropped out, you can block it if you want.

For non-HTML files — PDFs, images, CSVs — there is no head to put a tag in, so use the X-Robots-Tag HTTP header instead. Build robots.txt rules with the robots.txt generator.

Canonical tags

A canonical tag tells search engines which URL is the preferred version of duplicate or near-duplicate content, consolidating ranking signals onto one URL.

<link rel="canonical" href="https://example.com/product">

Five rules:

Self-reference every page. Even unique pages should declare themselves canonical. It protects against parameter and scraper duplication.
Use absolute URLs. Relative canonicals are legal but far more error-prone.
Be consistent. Pick one protocol, one host and one trailing-slash convention, and make canonical, internal links, sitemap and redirects all agree.
Point to a 200 page. A canonical to a redirected or 404 URL is ignored.
Never canonicalise a paginated series to page one. Each page should self-canonicalise; otherwise deeper items stop being indexed.

Canonical is a hint, not a directive. Google weighs it alongside internal links, sitemap contents and redirects, and will override it if the signals conflict. Consistency across all of them is what makes it stick. Verify the URL forms with the URL parser.

Open Graph and Twitter Cards

Open Graph, introduced by Facebook in 2010, is now read by LinkedIn, Slack, WhatsApp, Discord, Signal and X. A page with no OG tags gets a bare link with whatever text the platform scrapes, which measurably reduces clicks.

<meta property="og:type" content="article">
<meta property="og:title" content="Page title for sharing">
<meta property="og:description" content="One or two compelling sentences.">
<meta property="og:url" content="https://example.com/page">
<meta property="og:image" content="https://example.com/share.jpg">
<meta property="og:image:width" content="1200">
<meta property="og:image:height" content="630">
<meta property="og:site_name" content="Example">

<meta name="twitter:card" content="summary_large_image">

The image is what people actually notice. Use 1200×630 — a 1.91:1 ratio — keep it under about 5 MB, use an absolute HTTPS URL, and remember that text in the image gets cropped differently on each platform. Work out dimensions with the aspect ratio calculator.

X falls back to Open Graph when Twitter-specific tags are absent, so twitter:card is usually the only Twitter tag you need. Preview all of it in the social card debugger.

The complete head checklist

<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">

  <title>Unique, front-loaded, under 580px | Brand</title>
  <meta name="description" content="Unique, 140-155 characters.">
  <link rel="canonical" href="https://example.com/page">
  <meta name="robots" content="index, follow, max-image-preview:large">

  <meta property="og:type" content="article">
  <meta property="og:title" content="...">
  <meta property="og:description" content="...">
  <meta property="og:url" content="https://example.com/page">
  <meta property="og:image" content="https://example.com/share.jpg">
  <meta name="twitter:card" content="summary_large_image">

  <link rel="alternate" hreflang="en" href="...">   if multilingual
  <script type="application/ld+json">...</script>   structured data
</head>

Charset must come within the first 1024 bytes. Viewport is required for mobile usability, which is a ranking consideration. Everything else can be ordered freely.

Frequently Asked Questions

Are meta keywords still used?

No. Google announced in 2009 that it ignores the keywords meta tag entirely, and Bing has said it may treat stuffing as a negative signal. Remove it.

How long should a title tag be?

Under roughly 580 pixels on desktop, which is about 55–60 characters depending on letter widths. Google truncates by pixel width, not character count.

Does the meta description affect rankings?

Not directly, and it has not since 2009. It affects click-through rate, which affects traffic, so it is still worth writing well.

Why does Google change my title?

It rewrites titles that are too long, keyword-stuffed, duplicated across the site, or a poor match for the query, substituting your H1 or anchor text. A concise, accurate, unique title reduces the chance.

What is the difference between noindex and robots.txt?

robots.txt stops a page being crawled; noindex stops an already-crawled page being indexed. Blocking a page in robots.txt prevents the crawler seeing the noindex, so the URL can still appear in results.

Do I need canonical tags on unique pages?

Yes. A self-referencing canonical protects against parameter variants, session IDs and scraped copies competing with the original.

What image size should I use for Open Graph?

1200×630 pixels, a 1.91:1 ratio, served over HTTPS at an absolute URL. That works across Facebook, LinkedIn, X and Slack without awkward cropping.

Sources & further reading

Google Search Central: Meta tags Google understands — the definitive list of what is and is not honoured
Google: Title links — how titles are generated and why they get rewritten
The Open Graph protocol — the specification for social sharing markup
Google: Canonicalization — how canonical signals are weighed against each other