Keep these
These tags have a direct, documented effect on crawling, indexing, or social previews.
<title> — The most important tag on the page. Google uses it as the default headline
in search results. Keep it between 50 and 60 characters so it doesn't get truncated.
Write it for the reader, not as a keyword list.
<meta name="description"> — Google often (not always) shows this as the snippet below
the title. Aim for 150 to 160 characters. A clear, specific description sets accurate expectations for the reader.
<link rel="canonical"> — Tells crawlers which URL is the "official" version of a page.
Essential when the same content is reachable at multiple URLs (with/without trailing slash,
HTTP vs HTTPS, www vs non-www, paginated variants). Set it to the exact URL you want
indexed.
<meta name="robots"> — Controls whether a page is indexed and whether its links are
followed. The default behaviour (index, follow) is fine for most pages. Use
content="noindex" on thin or duplicate pages, login screens, and internal search results
you don't want in Google.
Open Graph tags (og:title, og:description, og:image, og:url) — These control
link previews on every major platform: LinkedIn, Slack, Discord, iMessage, WhatsApp. Without
them, platforms guess and often guess wrong. See What is Open Graph
for the full guide.
<meta name="twitter:card"> — Tells X (Twitter) which card layout to use. Set it to
summary_large_image for a full-width image card. Without it, X falls back to a small
thumbnail even if og:image is correct.
Run the free Meta Tag Analyzer to see every tag, its length verdict, and exactly what to fix.
Ignore these
<meta name="keywords"> — Google stopped using this tag in 2009. Bing officially
ignores it. Setting it has no SEO benefit and can reveal your keyword strategy to competitors
who inspect your source.
<meta name="author"> — Not used by any major search engine for ranking or attribution
in search results.
Legacy revisit-after, distribution, rating — Tags from an earlier era of the web. Modern crawlers ignore all of them.
Duplicate OG tags — Some CMS plugins output both og:title and a non-standard variant
like og:site_name as the title. Crawlers read the first valid og:title they find.
Extra copies don't help and can cause confusion when debugging.
How to check
Paste any URL into the Meta Tag Analyzer. It reads the actual
rendered <head> and flags missing tags, overlong titles or descriptions, missing canonicals,
and og:image issues in one view.
Keeping them consistent at scale
One page is easy. The challenge arrives when you have hundreds of pages and different teams (product, marketing, content) each adding tags through different CMS fields. A few practices that help:
- Centralise canonical logic in your framework's layout component so it's automatic, not opt-in per page.
- Template OG images rather than creating them manually. A programmatic approach
(e.g. using
/tools/og-image-generator) keeps images consistent and up to date as page titles change. - Audit on a schedule. Pages drift — title tags get overwritten, canonical conflicts emerge after URL restructuring. Running a periodic crawl with the Meta Tag Analyzer catches regressions before they compound.