Meta tags sit in the head of every HTML document and communicate information about the page to browsers, search engines and social platforms. They are simple to implement, which is part of the problem: because they are easy, a great deal of folklore has accumulated around them, and a significant portion of what is still recommended online has been irrelevant for over a decade. Getting meta tags right is worthwhile. Getting them wrong is common, and adding tags that do nothing is even more common.

This guide separates the tags that genuinely matter from the ones that do not, and shows the correct markup for each.

How We Handle On-Page Fundamentals

We audit meta implementation on almost every site we take on and we almost always find the same handful of issues. At AAMAX.CO we treat titles and descriptions as conversion assets rather than checkbox fields, because they determine whether a ranking position actually earns a click. Our SEO services cover full technical and on-page implementation including templated meta generation for large sites, structured data, canonical strategy and social markup, and because we build sites as well as optimise them we can implement changes properly at the template level rather than patching them page by page. We work with clients internationally and we validate every change against what search engines actually render.

The Tags That Matter

Title Element

Strictly this is not a meta tag, but it is the single most important element in the head and belongs in any discussion of meta implementation.

<title>Primary Topic | Brand Name</title>

Keep it within roughly fifty to sixty characters to avoid truncation in results, though the actual limit is pixel width rather than character count. Lead with the substantive topic rather than the brand, unless brand recognition is your primary draw. Make every title on the site distinct. Write it to be clicked, not merely to contain a keyword, because search engines increasingly rewrite titles they judge unhelpful and a well-written title reduces the chance of that happening.

Meta Description

<meta name="description" content="A specific, accurate summary of what this page provides.">

Descriptions are not a ranking factor and have not been for many years. They are a click-through factor, which makes them worth writing well for entirely commercial reasons. Aim for roughly one hundred and forty to one hundred and sixty characters. Describe what the page actually delivers, include the terms someone would have searched, and give a reason to choose your result over the adjacent ones. Unique per page, always. If you leave it blank, search engines will generate one from page content, which is sometimes acceptable and often not.

Viewport

<meta name="viewport" content="width=device-width, initial-scale=1">

This one is genuinely non-negotiable. Without it, mobile browsers render the page at desktop width and scale it down, which fails mobile usability assessment outright. It is a single line and its absence causes disproportionate damage.

Character Encoding

<meta charset="utf-8">

Place it as the first element in the head. Incorrect encoding produces mangled characters, which affects both rendering and how text is parsed.

Robots

<meta name="robots" content="index, follow">

Worth understanding precisely because misuse here causes the most severe damage of any meta tag. The default behaviour is to index and follow, so this tag is only necessary when you want to deviate from that. Use noindex to keep a page out of results, nofollow to stop link equity passing, and noindex, follow for pages like internal search results or filtered views that should not rank but sit within your link structure.

An accidental site-wide noindex, usually left over from a staging environment, is one of the most damaging single errors possible. Check it before and after every deployment.

Canonical

<link rel="canonical" href="https://example.com/preferred-url">

A link element rather than a meta tag, but functionally part of the same job. It tells search engines which URL is authoritative when the same or similar content is reachable at multiple addresses. Essential for sites with parameters, filters, pagination, session identifiers or protocol and subdomain variants. Always use absolute URLs and make every page self-canonical unless you deliberately intend otherwise.

Open Graph And Social

<meta property="og:title" content="Page title for sharing">
<meta property="og:description" content="Description shown when shared">
<meta property="og:image" content="https://example.com/image.jpg">
<meta property="og:url" content="https://example.com/page">
<meta property="og:type" content="article">
<meta name="twitter:card" content="summary_large_image">

These do not affect rankings directly. They control how your pages appear when shared, which affects click-through and therefore the traffic and engagement that do matter. Use an absolute URL for the image and size it appropriately, commonly around twelve hundred by six hundred and thirty pixels, or shares will render as unappealing bare links.

Hreflang

<link rel="alternate" hreflang="en-gb" href="https://example.com/en-gb/page">

Only relevant for multilingual or multi-regional sites. Every version must reference every other version including itself, or the implementation will be ignored.

Tags To Stop Using

Meta keywords carries no weight with any major search engine and has not for well over a decade. It does nothing except publish your keyword targeting to competitors.

Meta revisit-after was never supported by any major engine. Meta author, meta generator and meta rating have no search effect. Meta refresh for redirects should be replaced with proper server-side redirects, which are handled correctly and pass authority.

Removing these tags will not improve rankings, but it reduces clutter and removes the false impression that something useful is in place.

Where To Place Them

All of these belong inside the <head> element. Order mostly does not matter, with the exception of charset, which should come first so the parser knows how to interpret everything after it. Tags placed in the body are ignored, which is a surprisingly frequent implementation error in templated systems.

Implementing At Scale

Hand-writing meta tags works for a handful of pages and fails immediately beyond that. For anything larger, generate them from templates that pull page-specific variables, with sensible patterns per page type and manual overrides available for high-value pages. Product and category pages in particular benefit from templated patterns that incorporate genuinely differentiating attributes rather than repeating the same phrasing with a name substituted.

Whatever the system, the requirement is the same: every indexable page needs a distinct, accurate title and description.

Verifying The Implementation

View source rather than trusting the content management interface, because plugins and themes frequently output duplicate or conflicting tags. Check the rendered output rather than the raw source on JavaScript-heavy sites, since tags injected client-side may or may not be seen. Crawl the full site to find missing, duplicate or truncated values. Test social markup with the platform validators, and confirm in search console that the pages you expect to be indexed actually are.

Keeping Perspective

Meta tags are foundational rather than transformative. Correct implementation does not lift a page above better content, but incorrect implementation can suppress a page entirely, and a poorly written title can waste a hard-won ranking position. Get them right, keep them unique, remove the legacy clutter, and then spend your remaining effort on the things that genuinely move positions.