The title and the meta description are the two pieces of a page that most people read before deciding whether to read the page at all. They are also the two most frequently written on autopilot, by a template, and never looked at again.

They are not the same kind of thing

The title tag is a ranking signal. It is the strongest single statement of what a page is about, and search engines weigh it accordingly. It is also the headline of your result and the text of a browser tab and a bookmark.

The meta description is not a ranking signal, and has not been for many years. It earns its place entirely by persuading someone to click. Treat it as the ad copy under the headline — measured in click-through rate, not in rankings.

<title>Redirect chain checker — see every hop | Roftr Labs</title>
<meta name="description" content="Follow a URL through every redirect and see the status,
destination and timing of each hop.">

Length is measured in pixels

Character counts are a convenient approximation of the real limit, which is width. Google truncates at roughly 580 pixels for a title and around 920 for a description on desktop, and a capital W is more than three times the width of a lowercase l. A 60-character title in a wide, capitalised style gets cut; a 65-character one in ordinary sentence case often does not.

Truncation is not a disaster — the whole title is still read for ranking, and only the display is clipped. It matters because the clipped part is usually the end, so anything you need a human to see belongs at the front.

Why Google rewrites them

Google replaces a large share of titles in its results, and replaces most descriptions. This is normal and not a fault. It happens for consistent reasons:

  • The title is missing, empty or duplicated across many pages.
  • It is stuffed with keywords, or padded with boilerplate — a brand name, a city and three separators before any content.
  • It does not match the query as well as some other text on the page, so the h1 or a heading is used instead.
  • The description does not contain anything relevant to the search, so a snippet is lifted from the body copy where the answer actually is.

The practical response is not to fight it. A page whose title is being replaced by its own h1 is telling you the h1 is the better title.

Writing them

  • Front-load the specific thing. "Redirect chain checker" before the brand name, because the brand is the part a reader can lose without losing the meaning.
  • One page, one promise. If the title could describe three of your pages, those three pages compete with each other for the same searches.
  • Say what is on the page, not what you wish were. A description that oversells earns a click and then a bounce, which is worse than no click.
  • Write descriptions for the pages that matter and let the rest be generated from the page. A thin, near-identical description on 400 pages is worth less than none: Google would have written a better snippet from the content.
  • Keep the title and the h1 aligned but not identical. The title carries the search context; the h1 can be shorter now the reader has arrived.

The boilerplate trap

Templates like {page} | {category} | {brand} feel consistent and quietly spend a third of the visible width on words repeated site-wide. If the suffix is longer than the part that changes, the template is working against you. On a long page name, drop the suffix.

Checking them

Two questions are worth answering, and they are different. First, per page: is there exactly one title, is there a description, do they fit, and does the heading structure agree with them? Second, across the site: which titles are duplicated, and which pages have none at all? Duplication only becomes visible when pages are compared with each other, which is exactly the check nobody does by hand.

Previewing the result as it will actually appear is the fastest sanity check there is — seeing your own title cut off mid-word makes the fix obvious in a way a character count never does.