An SEO title is the text in a page's title tag, the one Google usually shows as the blue link in search results and the one that appears in the browser tab. And on most of the large sites I review, nobody actually wrote it: it came out of a template someone set up years ago, says something like "Product | Store" and repeats itself across thousands of URLs.
That's why this article has two parts. The first is what any guide to SEO titles should explain: what it is, where it shows up, how many characters to use, how to write it and why Google sometimes changes it. The second is what pays off most on a website with lots of pages: treating the title as a template decision, measuring each group of URLs against your own site's click curve and testing the change with a control group.
What is an SEO title?
The SEO title (also called title tag or meta title) is an HTML element that sits inside the head of every page and works as a one-line description of that content. It looks like this:
<title>Waterproof backpack for 16-inch laptops | Store</title>
The tag isn't visible in the body of the page. The browser, search engines and social networks read it when someone shares the link.
Something worth knowing: what appears in search isn't always your tag as written. Google builds what its documentation calls a "title link" from several sources.
According to Google's guide to title links, those sources are the title element, the main visible title, headings such as the H1, the og:title tag, other prominent text, the text of links pointing to the page and WebSite structured data. What you write in the tag is the most important source, but not the only one (which explains a few surprises we'll get to below).
Where does the SEO title appear?
In three places, and it plays a slightly different role in each:
-
In the SERP (Google's search results page). It's the link users read before deciding whether to click. There it competes with nine other results, ads and, more and more often, an AI-generated answer.
-
In the browser tab. When someone has twenty tabs open (that would be me), the start of the title is the only thing that identifies your site. If the text is long, it gets cut.
-
On social networks. When a link is shared on LinkedIn or WhatsApp, the card uses the og:title tag if there is one and, if not, the title. If you want a different title for social networks, Yoast SEO and other WordPress plugins let you write a separate one.
How is the SEO title different from the H1 and the meta title?
These three get mixed up a lot, so here's a table:
| Element | Where it goes | Who sees it | What it's for |
|---|---|---|---|
| Title tag | In the head | The user on Google | Deciding whether to click and explaining what the URL is about |
| H1 | In the body | Whoever already arrived | Confirming they landed in the right place |
| meta name="title" | In the head | Nobody | Google doesn't use it; some systems generate it out of habit |
The H1 is the visible title and can be longer and more conversational, because its job starts once the person is already on the page. The title has to work in a single line, next to the competition. What you do want is for both to say the same thing: when they talk about different things, Google tends to take the visible text as its source.
And when someone in a meeting says "meta title", 99% of the time they mean the title tag. Nobody needs to correct them (although sometimes I can't help myself). The meta description is a different tag: the gray text below the title.
What is the SEO title for?
It has two jobs at once. The first is relevance: it helps Google understand what the page is about and which searches it makes sense to show it for, in other words, to rank. The second is persuasion: of all the times your page appears, the title decides how many end in a visit. That ratio is CTR, the percentage of impressions that end in a click (the details of how it's calculated are in what CTR is).
Some guides say that raising that rate improves rankings. It might, but I don't know of data that proves it cleanly, so I prefer to measure it by what we know it does: turn impressions into traffic.
And there's a lot of work left there. In the Search Console data from 173 sites we analyzed for the content audit, the median was that 66% of pages with impressions didn't get a single visit from Google in 28 days. Some sit so far down that no title would save them, but others are on page one with a title that tells nobody anything.
The context has changed too. In a study I did for Ahrefs en Español (in Spanish), AI Overviews appeared in 81% of 2,998 popular searches across five Latin American countries, and in 79% of the ones in Chile. That article also cites an Ahrefs study of 300,000 keywords where the first result had a CTR about 58% lower when there was an AI Overview (a global figure, not a prediction for your site). So today the title competes in the SERP with an answer sitting at the very top, and it has to give a concrete reason to click.
How many characters should an SEO title have?
The recommendation you'll see everywhere is between 50 and 60 characters. The number has its logic: Google shows the title in a space of about 600 pixels on desktop, and 60 characters in a normal font fit roughly there. It's an approximation.
The official documentation says something different and much more useful: there's no length limit for the tag, but the title link gets truncated when it doesn't fit the width of the device. The limit is about display. For rankings, a 75-character title has no documented penalty; the user just won't read the end.
My practical rule:
-
The important part (what the content is about) in the first 50 characters.
-
The secondary part (brand, year, an extra detail) at the end, where it doesn't hurt if it's cut.
-
Check the length on mobile, which brings more than half the visits on many Latin American sites (we saw it in the mobile SEO study).
How do you write a good SEO title?
These are the tips for optimizing a title that actually move something, in the order I apply them:
-
Describe the page accurately. Before thinking about hooks, the title has to say what's behind it. If you sell a backpack for a 16-inch laptop that doesn't get wet, that's what your buyers are searching for. "Designed to go with you on every adventure" doesn't say whether the laptop fits or whether it gets wet.
-
Put the main keyword near the beginning. Search engines don't read left to right with a magnifying glass, but users scan the SERP in a second and need to see their keyword reflected. Just one, natural, without listing variants.
-
Add an honest hook. Numbers, a data point or a concrete promise the content delivers: "with 13 examples", "2026 prices", "step by step". Questions work when the page answers exactly that question. Years only if the content is truly up to date (Google detects old dates and changes them).
-
Brand at the end, short. A separator (pipe or hyphen) and the site name. On the homepage you can say more about the brand.
-
One unique title per URL. If two pages share the same title, neither Google nor the user knows which one to pick.
Some illustrative examples (made up, but very close to what you see on real sites):
| Before | After | What changed |
|---|---|---|
| Adventure Backpack - Store | Waterproof backpack for 16" laptops - Store | Attributes buyers search for |
| Home | Travel insurance for Europe from Chile - Insurer | The homepage says what it sells |
| Blog - Article 34 | How to calculate severance pay in Chile (with examples) | Keyword first and an honest hook |
| Sneakers, running sneakers, cheap trail sneakers | Women's trail running shoes - Store | No more keyword stuffing |
Formulas like "keyword + secondary + brand" work as a starting point. For the SEO writing on the rest of the page, and for the meta description, the criterion is the same: first what the visitor needs.
Why does Google rewrite your SEO title?
It happens often, and there's almost always a reason. Google's documentation lists the cases where it rewrites the title:
-
Half-empty titles. A title that only says "| Store" because the product field was blank. Google fills it in with the visible heading.
-
Obsolete titles. The title says 2024 and the H1 says 2026. The visible date wins.
-
Titles that don't match the content. If it promises something the content doesn't have, Google looks for more faithful text.
-
Text repeated across many pages (micro-boilerplate). Ten URLs with the same title that differ by a detail that doesn't appear. The detail is pulled from the content.
-
No clear main title. Several headings of the same size competing; the first one wins.
-
Different language. The title is in English and the site is in Spanish.
-
Repeated site name, when Google already shows it separately in the result.
Also, when the search is very specific, Google may pick other text from the page that better explains why that result fits the query. There's nothing to fix there: it's how Google adapts the title to search intent.
And a practical note: changes to the title tag can take from a few days to a few weeks to show, because the URL has to be crawled again. If it's urgent, you can request reindexing from the URL inspection tool.
The most common SEO title mistakes
These are the ones that show up most in audits:
-
Duplicate titles. SE Ranking analyzed 80,000 sites and found that 52.3% had titles repeated across pages. On large sites the cause is usually a template that pulls a field that repeats (the category instead of the product, for example). Beyond the duplicate content, Google doesn't know which one to show.
-
Empty field. The CMS shows the page's internal name or something like "Untitled".
-
Too long or too short. "Recipes - Site" tells nobody anything; a 120-character sentence gets cut in half in the SERP.
-
Keyword stuffing. Repeating the same keyword or listing its variants doesn't help you rank better. Google calls it out as bad practice and usually ends up rewriting it.
-
Titles that promise something else. A hook the content doesn't deliver brings visits that leave in two seconds.
To find them on a site with lots of URLs, a crawler like Screaming Frog lists duplicate, empty or overly long titles.
How do you change the SEO title in WordPress and other CMSs?
In WordPress, the usual route is a plugin:
-
Yoast SEO: this WordPress plugin puts an "SEO title" field at the bottom of each page's editor, with a preview and a bar that changes color depending on the length.
-
Rank Math: the WordPress alternative has an equivalent field in its panel, with the same preview logic.
-
No plugin: the theme usually builds the title from the post name and the site name, and sometimes that's enough.
In Shopify, PrestaShop and most ecommerce CMSs there's a title field and a description field in the SEO tab of each product or page. And if the site is custom built, the title tag lives in the template's HTML, which is exactly where the topic gets interesting.
How do you optimize SEO titles on a site with thousands of pages?
On a site with hundreds of thousands of URLs nobody writes titles one by one. They come from a formula per page type: product pages, categories, articles, store locations. So the useful question in that scenario is which template to change first, more than how to write a good title.
This is the method we use to optimize titles at scale (I shared it on LinkedIn a while ago and it's one of the things that pays off most):
-
Group by URL pattern. /product/, /category/, /blog/, /store/. For each group, pull impressions, visits, click-through rate and average position from the Search Console performance report.
-
Build your own click curve by position. What percentage your pages get in position 1, 2, 3 and so on, separating searches that include your company name. Across 131 sites we reviewed for the branded traffic study, the median was 15.4% for branded searches and 0.5% for non-branded; if you mix them, the curve is useless. For a quick calculation there's the CTR calculator (in Spanish).
-
Pick the template where the gap multiplied by impressions is largest. A group with a small gap but millions of impressions can be worth more than one with a big gap and little visibility. That cross tells you where the fastest return is.
-
Change the formula and keep a control group. Update the template on most of the URLs and leave a similar subset untouched. Measure for four weeks.
A real case, anonymized: an ecommerce site's product page template had 4,000 URLs adding up to 600,000 impressions a month, with a 1.2% CTR (about 7,200 visits) and an average position of 6. The site's own curve said that at position 6 the expected rate was 1.8%. By closing only half of that gap, up to 1.5%, the template went to about 9,000 visits a month. That's 1,800 extra visits from a change written once (and nobody had to rewrite 4,000 titles by hand, which is the part the team appreciates most).
How do you know if the new title worked?
This is where most title changes get lost, because nobody wrote anything down. The minimum:
-
Write down the date of the change and which URLs you touched. A spreadsheet is enough.
-
Wait for the pages to be crawled again. Check in the SERP that the new title actually shows (if it was rewritten, you now know what to look at).
-
Compare the same number of days before and after, on the same pages and the same searches, in Search Console: impressions, position and click-through rate.
-
Look at the control group. If the changed group improved and the control group did too, it was most likely seasonality or an algorithm change, more than your title.
Be careful reading the result only in visits: if impressions dropped because demand changed, traffic can fall even if the title works better. That's why it's worth looking at click-through rate at the same position.
Wrapping up
If you have a small website, apply the tips above: write every SEO title around what the visitor is looking for, check that it fits on screen and that it doesn't repeat. If you have a large website, open the performance report this week, group your URLs by page type and look for the template with the most impressions and the worst click-through rate for its position. Chances are the title that could bring you the most traffic was set up years ago and nobody has looked at it since.