Title tag length, and writing one that earns a click

The title tag is the line in a page's head that becomes the clickable headline of your search result and the label on a browser tab. In terms of how many people read it against how long it takes to write, it is the highest-leverage sentence on any page — far more people will see it in a results list than will ever read the page itself. It is also, for most small business sites, the element left entirely to whatever the platform generated.

A useful distinction up front: the title tag is not the h1. The h1 is the heading on the page, read by someone who has already arrived; the title tag is read by someone deciding whether to arrive at all. They often say similar things and there is no requirement that they match. The title can be written for a decision made in a results list, where it competes against nine other lines, and the h1 for a reader already committed. Platforms that force them to be identical are removing a choice worth having.

Titles are truncated by available width, and characters differ in width — a title of capital Ms occupies far more space than the same number of lowercase i's and l's. So the real limit is a pixel measurement of roughly 600 pixels on desktop results, and character counts are a proxy that happens to work most of the time.

As a working number, about 60 characters is the limit worth holding to, which is also what an on-page checker will flag against. Titles longer than that are commonly cut with an ellipsis, and the part that disappears is the end. This has one clear implication and it is the only length rule that really matters: the important words go first. A title reading 'Welcome to our website — Sharma Dental Clinic — Dental Implants in Pune' has spent its visible portion on a greeting and put the subject where it will be truncated away. Reversing it costs nothing and changes what the searcher sees. Note also that a truncated title is not a ranking problem — the page is assessed on the whole tag — it is a communication problem, which is a good enough reason to fix it.

The default advice is to append the business name after a separator, which is sound most of the time and worth understanding rather than following blindly. It goes at the end because for most queries it is the least useful part: someone searching for a treatment is choosing between providers and the treatment words are what they are scanning for. Putting the name first spends the front of the title, the part guaranteed to be visible, on something that does not answer their query.

Two exceptions. On your homepage and your about page, the business is the subject, so the name belongs at the front. And where your brand is genuinely well known to the audience you are addressing, the name is a reason to click rather than an afterthought, and moving it forward is defensible. Beyond those, keep it at the end and keep it short, because a long name plus a separator can consume a quarter of the available width. If space is tight, the name is the first thing to drop: someone reading your result can see your domain in the same block, so the name is rarely the only signal of who you are.

Google rewrites titles in results when it judges the tag a poor description of the page or a poor match for the query. Zyppy, examining a large sample of results, found rewrites in a majority of cases, and Google has confirmed the behaviour is deliberate rather than a fault. You cannot prevent it, and the useful response is to avoid the patterns that most reliably invite it.

The first is a title with no relationship to the page's content, which is the strongest invitation of all. The second is a title that is mostly boilerplate — the same twelve-word suffix of business name, tagline and location on every page, with two distinguishing words buried in the middle. The third is a title assembled from repeated variations of the same phrase, which reads as written for a machine and gives the engine little to work with. The fourth is an empty or generic title: 'Home', 'Untitled', 'Products', or the platform's default. In each case the engine is substituting because your tag described the page worse than something it could take from the page's own content, and the fix is to write a title that genuinely describes the page — after which a rewrite is still possible, because the engine may still find a better match for a specific query, and that is not a failure.

A method that works without inspiration. First, write the specific thing this page is about, in the words a customer would use rather than the words your industry uses. Second, add the qualifier that distinguishes it from every similar page elsewhere — the city, the specific variant, the audience. Third, if room remains, add the business name after a separator. Fourth, read the first forty characters alone and ask whether someone scanning a list would know what this is. That last check catches most problems, because it simulates what the searcher actually does, which is scan rather than read.

Two further rules. Every page's title should differ from every other page's in more than a word or two, because near-identical titles across pages are both a signal of generated content and a reliable indicator that two pages are competing for the same search. And the title must describe the page honestly: a title promising a price to a page that never states one produces a visitor who leaves immediately, which serves nobody and wastes the click you worked to earn.

Titles are unusually well suited to self-checking because everything structural about them is in your own markup. You can read every page's title tag out of its source. You can measure the length. You can sort the full list alphabetically and see near-duplicates immediately, which finds both boilerplate and pages competing with each other. You can confirm the important words come first by reading the first forty characters. None of this requires an account or a subscription, and you can re-check it the moment you change something.

What you cannot determine from your own pages is whether a title performed better than its predecessor. That comparison needs impressions and clicks, which comes free from a verified Search Console property and from nowhere else you have access to. Nor can you see, from your markup, which of your titles Google is currently rewriting in results — you can only observe it by searching and looking, one query at a time, in a private window. So the realistic position is that the craft is fully inspectable and free, the outcome needs a credential somebody has to connect, and a rewrite is a decision made downstream of anything you control.