Meta description length, and what the tag really does

A meta description is a short piece of text in a page's head that offers search engines a summary of what the page contains. It is not a ranking factor. Google has said so repeatedly and unambiguously, and the words you put in it do not make the page more or less likely to be found for a query. This surprises people who have been told to work their keyword into it, and the surprise usually turns into the wrong conclusion: that the tag does not matter and can be left empty.

What the tag actually influences is whether someone who has already found your result decides to click it. The description is most of the text under your blue link, and the difference between a sentence that answers the searcher's question and a sentence that repeats your business name three times is a real difference in how many of those people arrive. So the honest framing is not 'this helps you rank' but 'this decides how much of your existing visibility converts into visits'. Those are different jobs, and only the second one is yours to influence here.

Google treats your description as a suggestion and frequently substitutes its own text, usually pulled from the part of the page that matches the query. This is not a penalty and not a sign you wrote a bad one. It happens because the engine has information you do not: it knows what the person actually typed, and it can choose a passage from your page that speaks to that specific query rather than the general summary you wrote months earlier.

The rate is high enough to be worth internalising. Ahrefs, analysing just over 192,000 pages, found Google displayed something other than the supplied meta description about 63% of the time — roughly two out of three. That figure gets quoted as an argument for not bothering. It is better read the other way: the remaining third is still an enormous number of impressions where your sentence is the one doing the work, and you have no way of knowing in advance which queries fall in which group. Writing the tag is cheap. Discovering afterwards that a commercially important query was in the third where yours was kept is not something you can go back and fix.

The practical limit is a pixel width rather than a character count, because a line of capital Ms is wider than a line of lowercase i's, and the space available differs between a phone and a desktop. Character counts are a usable proxy anyway, and the working number most people settle on is about 155 characters, which is the same limit an on-page checker will hold you to. Under that, a typical description survives intact on desktop; well over it, the tail gets cut with an ellipsis.

Mobile is the tighter constraint and the one worth designing for, because the visible area is narrower. The consequence for how you write is specific: put the part that answers the question first, and treat anything after roughly the first hundred characters as material that may never be read. This is the opposite of how people naturally write a summary, which is to set up context and then deliver the point. A description that opens with context and closes with the answer is a description whose answer is the part that gets truncated away.

The first pattern is the business name. 'Sharma Textiles is a leading supplier of...' spends the most valuable words in the tag on something already visible in the URL and the title, and 'leading' is a claim no reader believes from the business making it. The second is restating the title almost verbatim, which doubles a sentence the searcher has already read a line above and adds nothing to their decision. The third is the keyword laid out for the engine rather than the reader — a phrase repeated in slightly different forms, which does not help ranking because the tag is not a ranking factor, and reads badly to the human it does reach.

The fourth is the vague invitation: 'Learn more about our services and how we can help your business grow.' It could sit under any result in any industry. Every one of these fails the same test — read the sentence, and ask what a searcher now knows that they did not know from the title alone. If the answer is nothing, the tag is decoration. The fix is usually to state the page's actual answer: what the treatment costs, how long the process takes, which documents are needed, whether you deliver to their city.

The reliable method is mechanical. Read the page and write down, in one sentence, the specific thing a visitor now knows that they did not before. That sentence is your description, possibly with a second clause added if there is room. It comes from the page, which means it is true, and it is specific, which means it distinguishes your result from the four around it.

Two constraints keep it honest. Do not put a number in the description that the page itself does not state — a figure in a summary should be traceable to the body, and a description that promises a price the page never mentions produces a visitor who bounces immediately. And do not write a description for a page you have not just re-read; the most common cause of a stale, misleading description is a page that changed while its summary stayed where it was. A rewrite that fits the limit is also where claims tend to drift, because the honest qualifier is always the longest part of a sentence and the first thing to get cut under length pressure.

You cannot see your own click-through rate from the page's markup, and no on-page checker can tell you whether a new description performed better than the old one. That comparison needs impression and click data, which comes from Search Console — free, accurate, and available only to someone who has verified the property and connected it. Without that connection, a description rewrite is a change you made for good reasons and cannot evaluate.

Even with it, read the numbers carefully. Click-through rate on a query containing your own business name means something quite different from the same rate on a generic query, because the first group already intended to reach you. Average position is an average across many impressions and moves for reasons unrelated to anything you wrote. And because Google substitutes its own text most of the time, a change in performance after a rewrite may be measuring a query mix that shifted rather than a sentence that worked. The tag is still worth writing. It is just not a lever with a clean readout attached to it.