Writing Meta Descriptions for Every Page (Yes, Every One)

When I audit websites, the single most common finding is also the least glamorous: missing meta description. It shows up so often that GazeSite has a canned finding for it. Site owners see it in a report and shrug. It feels like being told to floss. Everyone knows they should, nobody quite believes it matters, and the pages that lack descriptions keep working anyway.

So let me make the case for doing this chore, and for doing it completely, on every page, because the completeness is where the value hides.

First, the mechanics, plainly. A meta description is a short piece of text in the head of an HTML page. It does not appear on the page itself. Its main public appearance is in search results, where it often becomes the gray text under your blue link. It is not a ranking factor; search engines have said so for years and their behavior matches. Writing a great description will not move you from position eight to position three. What it changes is what happens once you’re on the results page at all: whether the person scanning ten links picks yours. It is ad copy for a free ad you have already won.

There is a catch that trips people up. Search engines treat your description as a suggestion, not a contract. If they think a snippet pulled from your page’s body answers the query better, they will use that instead, and they do this a large fraction of the time. People hear this and conclude descriptions are pointless. That’s backwards. When you write no description, you delegate the choice to a machine choosing under time pressure, and the failure mode is ugly: cookie-banner text, navigation labels, the first sentence of a legal disclaimer. When you write a good one, you win the cases where it’s used and lose nothing in the cases where it isn’t. It is a bet with no downside except effort.

But the effort argument is the one I actually care about, and it points somewhere unexpected. To write a meta description, you have to answer, in roughly a hundred and fifty characters, two questions: what is this page, and why would a stranger click it? That sounds trivial. Try it on a real site, page by page, and something interesting happens. Some pages are easy. The pricing page, the flagship article, the product page for the thing you’re proud of; the sentence writes itself, because the page has a job and you know it. Then you hit a page where you sit there stuck. You cannot say in one sentence what the page is for, because it isn’t for anything. It’s a tag archive with two entries. It’s an old landing page from a campaign nobody remembers. It’s a near-duplicate of another page, and any honest description you write for one would fit the other exactly.

This is the real reason to write descriptions for every page. The exercise is a page-by-page interrogation of your site, disguised as a formatting task. A page you can describe crisply is a page with a purpose. A page you can’t describe is a page you should merge, redirect, or delete. Most sites have never undergone this interrogation, because nothing else forces it. Analytics tells you which pages get traffic, not which deserve to. Redesigns restyle every page without asking whether it should exist. The humble meta description, done exhaustively, is the cheapest content audit there is.

As for how to write them, I have little patience for formulas, but a few observations hold up. Write for the person mid-search, not for a robot; they have a question, and your two lines should signal that this page answers it. Be specific, because specificity is what generic competitors can’t fake: “Pricing for teams of 5 to 50, no per-seat fees” beats “Flexible plans for every business” every time someone is actually comparing. Front-load, because search engines truncate long descriptions, typically somewhere past a hundred and fifty characters, and your point should survive the cut. And never paste the same description across pages. Duplicated descriptions are worse than none, because they tell both users and search engines that your pages are interchangeable, which is a strange thing to advertise about your own site.

The counterargument I hear is scale. Fine for a fifty-page site, but what about five thousand product pages? For genuinely templated pages, generate templated descriptions from real attributes: name, price, the one property that distinguishes this item from its neighbors. A generated description built from honest data is still a description; it still beats a machine’s guess. What doesn’t scale is the excuse. Most sites claiming to be too big to describe have a few hundred pages, of which a few dozen matter, and the owners have never written descriptions for even those.

Flossing turns out to be a decent analogy after all, but not for the reason people think. The point of flossing isn’t each individual pass of the thread. It’s that the habit forces daily contact with the state of things, so decay gets noticed early. Writing a description for every page forces contact with every page. Sites whose owners have that contact feel different to use. You can tell someone, at some point, went through the whole house and asked of each room: what is this for? Do that, and the gray text under your links in search results will improve too. But that will be the smaller of the two effects.