Skip to content

Article 16 min read

Title Tag: What It Is, Where Google Gets the Title, and How to Write One

by Daniel Sócrates published

Type “title tag” into Google. We ran that search on September 14, 2026, on Google in Brazil, where we are based. In fourth place was MDN, Mozilla’s HTML documentation, on its page about the title element itself. 1

The blue title came out in Portuguese, with the label “Traduzido pelo Google” (Translated by Google). The word “título” (title) came out broken, spelled “tÃtulo.” The title the page wrote into its own code is in English. 1

So the page that teaches the title tag showed up with a title it never wrote. Google’s documentation allows for both things: a title built by the machine, and a translation. 2 3

In the piece on on-page SEO, we already counted the nine titles of one search. All of them came from the title tag, some cut off and one translated. 4 The piece on meta descriptions ran a bigger test on the snippet, and the description showed up in 21 of 71. 5

Here the question moves. When does Google keep the title you wrote, when does it swap it, and where does the new text come from?

What a title tag is

Picture a book arriving at a library. The publisher printed the title on the cover. The librarian makes the catalog card and almost always copies the cover.

Sometimes the librarian drops the subtitle or the publisher’s name. Sometimes they copy the title page instead, because the cover did not match the book inside.

On a web page, the cover lives in the code, inside the <head>, on a single line: <title>Title text</title>. This is called the title tag, and Google’s documentation calls it the title element. 2 It is one of the first lines a technical SEO audit checks.

The catalog card is the blue, clickable text on each search result. This is called the title link. 2

The same tag also works outside search. Google’s starter guide notes that its text can become the title shown in browsers and bookmarks. 6 It is the name written at the top of each open tab on your screen.

The tag is a request

Google’s page on title links opens with two sentences. Google uses several sources to build the title link on its own. You can signal your preference by following its best practices. 2

The process is described as completely automated. It takes into account the page’s content and the references to the page across the web. 2 And Google writes that it cannot manually change title links for individual sites. 2

That is why I treat the tag as a request, the same way I treat the description. You suggest the text. Google chooses what goes on the card.

The official list has nine sources, in this order. It starts with the title tag, the main visual title on the page, and headings such as the H1. 2

Next come the og:title tag, text made large and prominent through styling, and the rest of the page text. The list closes with the anchor text on the page, the text of links pointing to it, and WebSite structured data. 2

When Google detects an issue on the page, it may try to build a better title link. It draws on anchors, on-page text, or other sources. 2

The swap takes time. To notice a change in those sources, Google has to recrawl and reprocess the page. The documentation says that may take from a few days to a few weeks. 2

The August 2021 change

On August 24, 2021, Google’s blog announced a new system for generating titles. The post was signed by Danny Sullivan, Google’s Search Liaison, who speaks for Search to the public. 7

The system had gone live the week before. Until then, a title could change depending on the search typed. With the new system, that generally stopped happening. 7

The post also announced more use of the text people see when they land on a page. That means the main visual title, the text in the H1, and text that is large and prominent. 7

The obvious question is in the post itself: why not always use the tag? Because the tag does not always describe the page well. The post lists tags that are very long, stuffed with keywords, missing, or repetitive. 7

The examples will sound familiar to anyone who has audited a site. A homepage called just “Home.” Pages with no title at all, or every page carrying only the site name. 7

Two kinds of swap are on record. Google may add the site name where that helps. And with an extremely long title, it may pick the most relevant portion instead of cutting off the end. 7

The tag was still the most used source, more than 80% of the time. 7 On September 17, 2021, after some refinements, Google wrote that the rate had risen to around 87%. 8

In our measurement, one page showed up in two different searches. In both, it carried the same title. 9 That is the behavior the 2021 post describes.

When Google swaps the text

The documentation lists the most common issues, each with an example. 2

The first is the half-empty tag. The site’s template generates only ”| Site Name,” and Google fills in the prominent text from the page. 2

The second is the obsolete tag. A university admissions page says 2021 in its visible title and 2020 in the tag. Google uses the right date, the one in the visible title. 2

Then come the inaccurate tag, repetition across a group of pages, and the lack of a clear main title. In the TV show example, the pages for each season share the same tag. Google adds the season number. 2

The last two cases are a language that differs from the text, and a repeated site name. When the name already shows in the result, Google may drop it from the title. 2

One case happens off the page. Robots.txt can block crawling without blocking indexing, when Google finds the page through an outside link. With no access to the content, Google builds the title from the anchor text of other sites. 2 To keep the page out of the index, the rule is noindex. 2

What we measured across ten searches

Google’s 2021 posts say the tag is by far the most used source. We wanted to see how big that is on a single day, in our own niche.

On September 14, 2026, we ran ten SEO searches on Google in Brazil, where we are based. 9 The searches were in Portuguese, with personalization off, in a browser tab of our own. For each organic result on the first page, we wrote down the blue title. Then we downloaded each page and compared the title shown with the tag, the H1, and the og:title in its code.

There were 85 results. One could not be read, which left 84 to compare. 9

Search (in Portuguese)ReadSame as the tagTag without the site nameTag cut offTag rearrangedTranslatedOther page text
title tag8201050
o que é seo (what is SEO)8131003
seo local (local SEO)9131022
consultoria de seo (SEO consulting)9421101
link building8111140
palavra-chave (keyword)7401002
google search console8431000
marketing de conteúdo (content marketing)9421002
agência de seo (SEO agency)9421200
sitemap xml (XML sitemap)9150201
Total842621961111

In 72 of the 84 results, the title came from the tag. It matched the code letter for letter in 26. 9

The most common swap was dropping the site name, almost always from the end, in 21 results. 9 The name did not vanish from the screen. It shows above the link, on a line of its own, and the documentation calls it the site name. 10

Nine titles were the tag cut off with an ellipsis. Across all the truncated titles in the measurement, the cut fell between 42 and 59 visible characters. Full titles of up to 62 characters showed up uncut. 9

In six, the text was the tag in a different order. In four of those, the brand moved to the other end, with a colon in place of the separator. 9

Eleven titles came translated from English, with the “Traduzido pelo Google” label. Ten translated the tag. One translated the H1, which said something else. 9 Two of the translated titles carried the broken “título,” like the one in the opening. 1 9

In another eleven, Google used other text from the page. Ten times, it was the page’s main visible title, almost always marked as the H1. 9 In all ten, the tag said one thing and the visible title said another.

One of those tags ran 112 characters. Another held only the site name, and the result got the article’s title instead. 9 Those are two cases the 2021 post describes: a tag that is too long, and a tag with only the site name. 7

The last case is the oddest. The title shown came from the alt text on the page’s logo image. 9

The same marketing publisher showed up in seven of the ten searches. In three of them, Google swapped its tag for the H1. 9

Ten searches on one day do not make a law. They put a number and a date on the “most used source” in Google’s posts.

The length Google does not document

Google sets no length for the tag. The documentation says there is no limit, and that the title link is truncated as needed to fit the screen. 2 For the site name, the rule is similar. 10

We read twelve third-party texts that teach the title tag, from the top results on Google in Brazil and in Semrush. Most are in Portuguese. Eleven give a number, from 50 to 66 characters. Four also mention 600 pixels. 11

Several attribute the number to Google, with phrases like “Google cuts” and “on Google, the limit is.” 11 Only one writes that Google does not specify a length. That one’s number comes from its own research on rewritten titles. 11

Where does the yardstick come from, then? From the tools. Screaming Frog flags titles over 60 characters and over 561 pixels. 12

The tool itself calls the pixel limit an estimate, one that Google adjusts frequently. And it lets you change both numbers in its settings. 12 The number that circulates is a tool convention, useful as a warning of where the screen will probably cut.

In our measurement, the cutoff point varied by 17 characters from one title to the next. 9 So the practical rule is a different one. Write the start of the tag as if it were all that shows.

Language and writing system

The documentation asks for the tag in the same language and writing system as the main text. Writing system, in Google’s words, means the script or alphabet of a given language. 2

The official example is a page in Hindi. The tag should be in Hindi too, with no English and no transliteration into Latin characters. If the language or the script does not match, Google may pick another text for the link. 2

One detail trips people up. The HTML lang attribute does not settle the question. Google writes that it detects the language from the page text, without relying on that markup. 3

Take a Portuguese page marked lang="pt-BR" with a title tag in English. The tag is still out of step with the language of the text.

Translation follows a separate rule. When a page is in a language the searcher probably does not read, Google may translate the title link and the snippet. 3 That explains the eleven translated titles in our measurement: English pages, shown to people searching in Portuguese. If you do not want your page translated, there is a rule for that, notranslate. 3

Our company site in Brazil fell under the language rule, back in its WordPress days. The Portuguese homepage served a title tag in English, according to our diagnosis of August 31, 2026. 13 The full case, with the WordPress field that caused it, is in the piece on SEO for WordPress.

What Google asks for in a good title tag

The starter guide sums it up in one sentence. A good title is unique to the page, clear and concise, and accurately describes its content. 6 It can include the business name, the physical location, and what the page has to offer. 6

The title link page goes into detail, with examples. 2 Avoid vague descriptors, like “Home” for the homepage or “Profile” for a person’s profile page. Avoid needlessly long text and the same words over and over.

Also avoid repeated boilerplate text. The official example is a store that titles every page “Cheap products for sale.” Anyone who sees two of those pages in search cannot tell which one to open. 2

Brand with restraint. On the homepage, there is room for a line about the site. On other pages, the site name at the start or the end is enough, set off by a hyphen, a colon, or a pipe. 2

Make it clear which text is the page’s main title, for example in the first visible H1. 2 And keep airfare prices out of the tag. Google probably will not show them, because those prices can change every few minutes. 2

Our own case

On our company site in Brazil, each piece has its own field for the search title. Whoever writes the piece fills in up to 46 characters. The site adds ” | SEO Genome” at the end and builds the tag. 14 Without the field, the site uses the piece’s H1, with the same ending. 14

The 46-character yardstick is a house rule. With the name added, it comes close to 60 characters, the number the tools use. 12 14 Google does not write any number. 2

We keep the yardstick because truncation exists, and a cut title loses its ending. 2

So we measured our own site. In a search for seogenome.com pages, we read the first three pages of results, 30 addresses in all. 15

The ” | SEO Genome” ending did not appear in any title. Of the 26 pages that carry it in the tag, none showed it in search. 15 The company name appears above the link, as the site name. 10 That is the repeated site name case the documentation describes. 2

Three pages came out with the H1 in place of the tag, two of them service pages. On all three, the H1 and the tag said different things. 15

The third was the English page about our method, /en/methodology/. Its tag read “SEO Genome: the method behind the name,” plus the ” | SEO Genome” ending, so the brand appeared twice. Google showed the H1, “SEO Genome: the method behind the company name.” 15 We rewrote that tag on September 14, 2026. It now reads “Methodology: How the Company Works.” 15

The difference alone does not explain the choice. On twenty other pages, the H1 was also different, and Google kept the tag. 15

An old address showed the delay problem. /prompt-gsc/ now redirects to our piece on AI-powered audits. In search, it still showed the title from the WordPress days. 15 The documentation warns that Google has to recrawl and reprocess the page first. 2

The lesson we take from it applies to any site. The H1 is also a candidate for the blue link, so it has to work as a title. 2 15 And the first characters of the tag have to stand on their own without the site name, which will probably disappear.

The subheadings below the H1 have a different job, serving as the address of each passage. That job is covered in the piece on titles as addresses.

The fifteen-minute check

  1. Search for your site. Type site: followed by your domain into Google. Write down the blue title of the pages that sell the most. 9
  2. Open the code. On each page, press Ctrl+U (Cmd+Option+U on a Mac) and look for <title>. Put the tag next to the title shown.
  3. Find the source of the swap. If the text shown is in the H1, the logo, or another passage, look for the case in the list of common issues. 2
  4. Run a crawl. The Page Titles tab in Screaming Frog shows missing tags, duplicates, and tags identical to the H1. Read the character filter as a tool warning. 12
  5. Check the language. Is the tag in the language and alphabet of the main text? 2
  6. Changed something? Ask for a recrawl. Google takes from days to weeks to notice the change. 2

From now on, when the blue title comes back different from what you wrote, you will look for the source. One of the nine got ahead of your tag, and the documentation almost always says why.

Open Google today, search your domain with site:, and put the blue title of your homepage next to its tag.

Frequently asked questions

What is the ideal length for a title tag?

Google sets no limit. The title link is cut to fit the width of the screen. 2 Ranges of 50 to 66 characters come from market articles and from tools. 11 12 Ours, 46 characters plus the site name, is a house rule. 14

Why did Google change the title of my page?

The title link is built automatically from nine sources. 2 The documentation lists the cases for a swap: a tag that is half-empty, obsolete, inaccurate, repeated, lacking a clear main title, or in another language. 2 If you changed the tag recently, Google may take days to weeks to notice. 2

Do the title tag and the H1 have to match?

The documentation does not ask for identical text. It asks for the main title to be clear, for example in the first visible H1. 2 In our measurement, the main visible title went to the blue link in 11 of 84 results. 9

How do I stop Google from translating my title?

The rule <meta name="googlebot" content="notranslate"> tells Google you do not want the page translated. 3 Without it, Google may translate the title link and the snippet when the page is in a language the searcher probably does not read. 3

Notes and sources

  1. Live reading of the Google Brazil results page for “title tag,” September 14, 2026, at https://www.google.com.br/search?q=title+tag&hl=pt-BR&gl=br&pws=0, in the author’s own browser tab. The search ran on Google in Brazil, where we are based, with the interface in Portuguese. In the fourth organic position, the MDN page https://developer.mozilla.org/en-US/docs/Web/HTML/Reference/Elements/title appeared with the title <title> Elemento de tÃtulo do documento HTML and the label “Traduzido pelo Google · Ver original (English)” (Translated by Google · See original). The title tag in the HTML downloaded the same day is <title> HTML document title element - HTML | MDN. In the second position, W3Schools came out with the translated title “Tag de tÃtulo HTML,” from the tag “HTML title tag.” Accessed September 14, 2026.
  2. Google Search Central, “Influencing your title links in search results,” https://developers.google.com/search/docs/appearance/title-link, page last updated December 10, 2025; Portuguese version, “Como influenciar os links de títulos nos resultados da pesquisa,” last updated December 18, 2025. Accessed September 14, 2026. Passages: “Google uses a number of different sources to automatically determine the title link, but you can indicate your preferences”; “Google’s generation of title links on the Google Search results page is completely automated and takes into account both the content of a page and references to it that appear on the web”; “While we can’t manually change title links for individual sites”; the list of nine sources, from “Content in <title> elements” to “WebSite structured data”; “Google has to recrawl and reprocess the page to notice updates to these sources, which may take a few days to a few weeks”; “If we’ve detected an issue on the page, we may try to generate an improved title link from anchors, on-page text, or other sources”; the common issues (“Half-empty”, “Obsolete”, “Inaccurate”, “Micro-boilerplate”, “No clear main title”, “Mismatch of writing system or language”, and “Duplication of the site name”), with the examples of | Site Name, the 2020 and 2021 admissions, and the seasons; “If we don’t have access to the content on your page, we will rely on off-page content to generate the title link, such as anchor text from other sites”; “While there’s no limit on how long a <title> element can be, the title link is truncated in Google Search results as needed, typically to fit the device width”; “Use the same language and writing system (meaning, the script or alphabet for a given language) as the primary content on your pages”, with the Hindi example; “Avoid vague descriptors like “Home” for your home page, or “Profile” for a specific person’s profile”; the example “Cheap products for sale”; “consider including just your site name at the beginning or end of each <title> element, separated from the rest of the text with a delimiter such as a hyphen, colon, or pipe”; “putting the title text in the first visible <h1> element on the page”; and “Avoid including flight price information in <title> elements”, because the price “can change so fast (sometimes every few minutes)”. In the Portuguese version, the term is “link de título” (title link).
  3. Google Search Central, “Meta tags and attributes that Google supports,” https://developers.google.com/search/docs/crawling-indexing/special-tags, page last updated December 10, 2025. Accessed September 14, 2026. On notranslate: “When Google recognizes that the contents of a page aren’t in the language that the user likely wants to read, Google may provide a translated title link and snippet in search results”, and the tag “tells Google that you don’t want us to provide a translation for this page”. In the section on unsupported tags, about the lang attribute: “Google Search detects the language of a page based on the textual content of the page. It doesn’t rely on code annotations such as the lang.”
  4. SEO Genome survey, September 14, 2026, published in this site’s piece on on-page SEO: the nine titles on the first page for “seo on page” on Google in Brazil came from the title tag (five identical, two cut off with an ellipsis, one without the site name, and one translated from English).
  5. SEO Genome survey, September 14, 2026, published in this site’s piece on meta descriptions: in ten searches on Google in Brazil, 71 snippets compared; in 21, the snippet shown was the meta description.
  6. Google Search Central, “SEO Starter Guide,” https://developers.google.com/search/docs/fundamentals/seo-starter-guide, page last updated December 10, 2025. Accessed September 14, 2026. In the section “Influence your title links”: “This title text can also be used for the title that’s shown in browsers and bookmarks”; “a good title is unique to the page, clear and concise, and accurately describes the contents of the page”; and “your title could include the name of your website or business, other bits of important information like the physical location of the business, and maybe some information about what the particular page has to offer for users.”
  7. Danny Sullivan, “An update to how we generate web page titles,” Google Search Central Blog, August 24, 2021, https://developers.google.com/search/blog/2021/08/update-to-generating-page-titles. Accessed September 14, 2026. Passages: “Last week, we introduced a new system of generating titles for web pages. Before this, titles might change based on the query issued. This generally will no longer happen with our new system”; “we are making use of text that humans can visually see when they arrive at a web page”, with the main visual title, the text in <H1> tags, and large, prominent text; the list of tags that are “Very long”, “Stuffed” with keywords, missing, or repetitive, such as homepages called “Home”, pages that are “Untitled” or carry only the site name; “In some cases, we may add site names where that is seen as helpful. In other instances, when encountering an extremely long title, we might select the most relevant portion”; and “content from HTML title tags is still by far the most likely used, more than 80% of the time.”
  8. Google Search Central Blog, “More information on how Google generates titles for web page results,” September 17, 2021, https://developers.google.com/search/blog/2021/09/more-info-about-titles. Accessed September 14, 2026. Passage: “title elements are now used around 87% of the time, rather than around 80% before.”
  9. SEO Genome survey, September 14, 2026. Ten searches on Google in Brazil, where we are based, at https://www.google.com.br/search with hl=pt-BR, gl=br, and pws=0, in the author’s own browser tab, all in Portuguese: “title tag,” “o que é seo” (what is SEO), “seo local” (local SEO), “consultoria de seo” (SEO consulting), “link building,” “palavra-chave” (keyword), “google search console,” “marketing de conteúdo” (content marketing), “agência de seo” (SEO agency), and “sitemap xml” (XML sitemap). In the two commercial searches, the browser’s location brought up results from Brasília, Brazil’s capital. For each organic result on the first page, the title shown was compared with the title tag, the H1s, and the og:title in the HTML downloaded the same day; one page built by script was read in the browser. Result: 85 results, 1 that could not be read, 84 compared. From the tag, 72: 26 identical, 21 without the site name, 9 cut off with an ellipsis, 6 rearranged, and 10 of the 11 translated. From other page text, 12: the main visible title in 11 (one of them translated) and the alt text of a logo in 1. Fourteen titles shown end in an ellipsis, with the cut between 42 and 59 visible characters; full titles ran up to 62. The same page appeared in “consultoria de seo” and “agência de seo” with the same title. Accessed September 14, 2026.
  10. Google Search Central, “Provide a site name to Google Search,” https://developers.google.com/search/docs/appearance/site-names, page last updated December 10, 2025. Accessed September 14, 2026. Passages: “When Google lists a page in search results, it shows the name of the site the page comes from. This is called the site name”; the site name “is different from the per-page title links”; and “While there’s no limit to how long a site name can be, long site names may be truncated on some devices.”
  11. SEO Genome survey, September 14, 2026: twelve third-party texts about title tags, from the first page on Google in Brazil (note 1) and the top ten in Semrush’s Brazil database, read the same day; most are in Portuguese. Eleven give a character count, from 50 to 66; four mention 600 pixels. Examples of attribution to Google: “O Google corta title tags com mais de 55 a 60 caracteres” (Google cuts title tags longer than 55 to 60 characters), https://pt.semrush.com/blog/como-escrever-title-tags/; “No Google, o limite é de 600 pixels” (On Google, the limit is 600 pixels), https://www.mirago.com.br/title-tag/; “Google shows the first 66 characters,” https://www.agconsult.com/en/usability-blog/8-tips-for-the-perfect-title-tag/. Moz’s text, https://moz.com/learn/seo/title-tag, says “While Google does not specify a recommended title tag length” and supports its 50 to 60 range with its own research. Accessed September 14, 2026.
  12. Screaming Frog, SEO Spider, the pages “Page Titles: Over 561 Pixels” (https://www.screamingfrog.co.uk/seo-spider/issues/page-titles/over-561-pixels/) and “Page Titles: Over 60 Characters” (https://www.screamingfrog.co.uk/seo-spider/issues/page-titles/over-60-characters/), and the configuration guide (https://www.screamingfrog.co.uk/seo-spider/user-guide/configuration/). Accessed September 14, 2026. Passages: the pixel filter flags titles over “Google’s estimated pixel length limit”, and “it is an approximation and Google adjusts them frequently”; in the preferences, “This allows you to set your own character and pixel width based upon your own preferences”. The list of filters in the Page Titles tab includes Missing, Duplicate, and Same as H1.
  13. SEO Genome internal record, “Diagnóstico do site seogenome.com” (seogenome.com site diagnosis), August 31, 2026, section 2.6: the title tag served on the Portuguese homepage of the WordPress site was in English. The full case is in this site’s piece on SEO for WordPress.
  14. SEO Genome internal record, production rules for the new site, consulted September 14, 2026: the production brief for the content calendar (section 6) asks for a metaTitle of up to 46 characters; the content importer records that the layout adds ” | SEO Genome” and that the yardstick is 60 characters including the suffix; each piece’s page builds the tag from the metaTitle, or from the piece’s title when the field is missing, followed by ” | SEO Genome”.
  15. SEO Genome survey, September 14, 2026: the search “site:seogenome.com” on Google in Brazil, at https://www.google.com.br/search?q=site%3Aseogenome.com&hl=pt-BR&gl=br&pws=0, first three pages of results, 30 addresses, in the author’s own browser tab, with the title shown compared against the title tag and the H1 in the HTML downloaded the same day. No title shown carried the ” | SEO Genome” ending (26 pages carry it in the tag). The H1 replaced the tag on three pages (services, the local SEO service page, and the English methodology page). The tag of /en/methodology/ was “SEO Genome: the method behind the name | SEO Genome”; it was rewritten on September 14, 2026, as “Methodology: How the Company Works | SEO Genome,” according to the site’s content importer. The address /prompt-gsc/ now responds with a 301 redirect to /seo-tecnico/auditoria-com-ia/ and appeared with the old page’s title, recorded in the crawl of the WordPress site of August 31, 2026. Accessed September 14, 2026.

Want to see SEO Genome more often on Google? Add it as a preferred source.

Daniel Sócrates

Founder and CEO of SEO Genome

Author of AI Search Optimization. Writes about what the company applies in client work.

danielsocrates.com.br

Let us look at what the machine understands about your company

A diagnostic conversation, with no slide deck. You bring your website address. We bring what Google and the AI assistants already know about it today.

Talk to us on WhatsApp