Title and description: how to write page meta tags

What title and description meta tags are

Where a user sees title and description

A user sees the title in two places: in the browser tab’s label and in the heading of a search result. Google calls that heading a title link: the text above the result that the user clicks to reach the page. The description isn’t visible on the page itself. It may appear in search results as the description below the heading.

The informal terms «title tag» and «meta description» mean the same thing. In a page’s code, both fields usually sit next to each other:

<head>
  <title>Bathroom renovation: stages and materials — Site name</title>
  <meta name="description" content="What stages a bathroom renovation involves, which materials to pick for the walls and floor, and what to check before signing off on the work.">
</head>

The heading shown in search results largely determines whether someone clicks through to the site. Google states that users typically decide to visit a page based on the title link. According to Yandex’s help documentation, the information in the title mainly affects the user’s decision to visit the site.

What other meta tags exist, and which ones matter for SEO

Meta tags are utility <meta> tags in the <head> section: they pass information about the page to the browser and search engines. Formally, title is a separate HTML element, not a meta tag, but in SEO both fields are usually called meta tags.

Besides title and description, two other tags most often matter for search:

• The robots meta tag sets the crawling and indexing rules for a page. The syntax and use cases are covered in the article on the robots meta tag and the rules for showing a page in search.
• og:title is an Open Graph markup tag, usually set to control the link preview in social networks and messengers. Google lists it among the sources it may use to generate the title link.

The remaining meta tags (character encoding, viewport, and others) govern how the page functions, not how it looks in search results.

Do you need meta keywords

The keywords meta tag is a list of a page’s keywords in the code. Google states directly that it doesn’t use this tag in Search and that it doesn’t affect indexing or ranking. Yandex’s help documentation on meta tags lists keywords with the wording «may affect how well the page matches search queries.» The documentation gives no further detail.

The practical takeaway: the field is useless for Google; for Yandex, you can fill it with a few phrasings that actually appear on the page. It’s better to spend that time on title and description instead — users actually see those.

How Google and Yandex use title and description

Where a search engine gets the heading shown in results

Google generates the title link automatically. According to its documentation, the sources include:

• the content of the <title> tag;
• the main visible heading on the page;
• headings and subheadings, for example <h1>;
• the content of og:title;
• other prominent text set off by styling;
• the rest of the page’s text;
• the anchor text of links on the page;
• the anchor text of links that point to the page;
• structured data of type WebSite.

For Yandex, the title and description shown in results are formed from the title element, the description meta tag, and the page’s text, including Schema.org structured data. Additional sources include Yandex Market and the anchor text of links to the page from other sites. The heading can also be generated from the page’s content, which is why users see different headings for different queries.

Where the description shown in results comes from

Google generates the description automatically: it pulls in the page fragment that best matches the query. Google uses the meta description in some cases — when it describes the page more accurately than other material on it. That’s why the description for one page can differ across different queries.

Yandex lists description as one of the sources used to form page descriptions in search results. Yandex may use it as the snippet’s description, but it isn’t required to.

The description field doesn’t set the price, publication date, or other extra result elements. Such rich results are usually tied to structured data, and a search engine isn’t required to show them either. More on where rich results in search come from.

Why search results show a different title or description than yours

If search results show a different heading, check the title itself first. Google lists the problems it runs into most often:

• the title is only partially filled in — for example, it contains only | Site name;
• the title is outdated — it has last year’s date or information that’s no longer current;
• the title is inaccurate — it doesn’t reflect the page’s actual content;
• the same boilerplate text appears on many pages — it gives no way to tell them apart;
• the page has no clear main heading — several large headings at the same level;
• the title’s language doesn’t match the page’s language;
• the site name is repeated within the title.

Yandex advises avoiding generic headings like «Home» or «Contacts,» keyword lists, and boilerplate titles. Google likewise calls out overly generic phrasing such as «Home page.»

The same logic applies to the description. Google uses the meta description when it describes the page more accurately than other material on it. Yandex states that boilerplate or near-duplicate descriptions are unlikely to be shown in search results.

What to do: find the page with the wrong heading → compare the title with the page’s content and with the titles of neighboring pages → fix the cause from the list above. You can hide specific fragments from the snippet using nosnippet, max-snippet, and data-nosnippet. That limits what’s shown — it isn’t a way to force a search engine to use your description.

How to write a title

What a title should include: the page’s topic and the primary query phrasing

Google’s and Yandex’s requirements are essentially the same:

• every page has its own title, and it reflects that page’s content;
• short and informative;
• no keyword lists and no keyword repetition;
• only what’s actually on the page: Yandex advises putting into the title only values whose content is actually on that page; according to its documentation, other information can go into the description instead.

In practice, a title is built around the primary query phrasing people use to find the page. The article on how to choose the primary query phrasing for a page covers how to pick it from your collected queries. A title should promise exactly what the page delivers: if the query expects a how-to and the page is a catalog, an accurate heading won’t help. More on what a user expects to find for a query.

Site or brand name: at the start or at the end

Both systems allow either option. Google recommends placing the site name at the start or end of the title and separating it with a hyphen, colon, or vertical bar. Yandex recommends the same: brand at the start or end, separated by a colon, hyphen, or vertical bar.

Google allows extra information about the site in the homepage’s title, but advises against repeating it on every page. In our view, it’s more practical to put the page’s topic first and the site name last: someone scanning search results is looking for the answer to their question first.

Title and H1: the difference, and whether they can match

Title is the text for the browser tab and search results. H1 is the visible heading on the page itself. Google uses H1 as one of the sources for the title link, so a mismatch between the two is one more reason a search engine might show the wrong text.

They can match if the topic is the same. H1 can be made more detailed, while the title stays shorter and includes the site name. The rules for H1–H6 headings in body text, and where in the page’s text to place the primary query phrasing, are covered separately.

Examples: what’s wrong and how to do it better

Title	What’s wrong	How to do it better
Home	Too generic, says nothing about the content	Apartment and bathroom renovation — Company name
Apartment renovation, cheap apartment renovation, apartment renovation price, finishing work	A list of keywords with repetition	Apartment renovation: types of work and cost — Company name (if the cost is actually on the page)
Blog — Site name (on every article)	The same boilerplate text on many pages	How to choose bathroom tile — Site name
Discounts on all services until the end of the month (there are no discounts on the page)	The title doesn’t match the content	A heading about what’s actually on the page
\| Site name	The title is only partially filled in, no topic	Page topic + site name at the end

How to write a description

What to put in a page’s description

The description is a short summary of this specific page’s content. The systems’ recommendations:

• different descriptions for different pages;
• specific information instead of generic phrases. Google gives examples: for a blog article, the author and publication date are appropriate; for a product page, the price, age group, and manufacturer;
• no keyword stuffing. According to Google, a description made of nothing but keywords gives no sense of the page. Yandex asks not to overuse keywords, phrases, and capital letters;
• the key point comes first;
• the description differs from the title rather than repeating it.

An example for a hypothetical article:

Before: Bathroom renovation, bathroom remodeling, bathroom repair, prices, cheap.
After:  What stages a bathroom renovation involves, which materials to pick for the walls and floor, and what to check before signing off on the work.

The «before» version is just queries. The «after» version shows what someone will actually find on the page.

Meta description and a product card’s description are different things

By «SEO product description,» people usually mean the visible text on the product card: specs, benefits, use cases. Meta description is a hidden field in the code, visible only in search results. These are different jobs, and the product card’s text doesn’t replace the description.

How long should title and description be

Google and Yandex don’t set a fixed length for title and description. Google states that the length of the <title> element isn’t limited, and the title link in search results is truncated to fit the device’s screen width. The meta description also has no length limit, but it may be shortened in search results. Yandex: the length of the heading and the length of the description are limited by the width of the user’s device screen.

Numbers like «50–60 characters for the title» and «150–160 for the description» show up in many SEO articles, but they’re absent from Google’s and Yandex’s documentation: they aren’t a search engine requirement. The length thresholds in site-auditing tools are just settings built into those tools.

What to do in practice (this is advice, not a rule):

• put the key point first: the topic in the title, the gist in the description’s opening words. Yandex recommends the same for descriptions;
• a long title isn’t an error by itself, but its end may not fit. Yandex warns that overly long titles can be shortened in search results;
• after publishing, check how the title and description look in search results on both phone and desktop.

Title and description templates for bulk pages

How to build a template from a page’s data

For a catalog, product cards, articles, and similar service pages, it’s more practical to build title and description from a template. A template inserts a specific page’s own fields into the text: name, category, specs. Google recommends generating descriptions for such pages programmatically, based on the page’s own data. If writing descriptions for every page at once isn’t feasible, Google advises starting with the main ones — the homepage and the most popular pages.

Templates are usually set up by page type and site section. The article on site structure covers how to identify page types and site sections.

Template or duplicate: where’s the line

A template is useful as long as it inserts fields that distinguish pages from each other. If the same value gets inserted everywhere, you end up with duplicates. In Yandex’s example, the identical heading «Catalog» on an online store’s catalog pages makes it hard for a visitor to tell one page from another. Yandex also advises against headings where only part of the phrase changes, such as «‹Product name› — description, prices, specs, photos, reviews.» Words like «reviews» and «photos» should only be inserted if that content actually exists on the page.

In a similar situation, Google sometimes supplements the heading itself. Its documentation gives a TV show example: if season pages share the same title, Google may find the season number on the page and add it to the title link. It’s more reliable not to count on this and to put the distinguishing field directly into the template instead.

Page type	Which fields to build the title from	Which fields to build the description from	How pages differ
Catalog category	{category name} + {section's main feature} + {site name}	{what the category includes}, {main brands or types}, {terms, if they're on the page}	by the category’s name and contents
Product card	{product name} + {model or key spec}	{purpose}, {2–3 specs}, {availability or price, if they're on the page}	by model and specs
Blog article	{article topic} + {site name}	{what the article is about and what question it answers}	by topic
Service page	{service name} + {clarification: who it's for or what's included}	{what the service includes}, {for which tasks}	by the service’s contents

Draft titles and descriptions can also be prepared automatically, including by AI. They still need to be checked against the same rules: uniqueness, matching the page, no keyword lists.

When meta tags need a systematic rewrite across hundreds of pages, along with what those pages are for, that becomes work on the site’s content as a whole. See how we work with content and page purpose.

How to find empty, duplicate, and boilerplate title and description tags

Yandex Webmaster

In Webmaster, duplicate titles and descriptions are visible in several places:

• on the «Summary» page — the count of identical titles and descriptions;
• in «Indexing» → «Titles and descriptions» — the percentage and count of duplicates. The «title» tab groups pages with identical headings, and the «description» tab groups pages with identical descriptions;
• if there are many duplicates, a warning appears in «Site optimization» → «Site diagnostics.»

If titles are generated automatically, Yandex advises making them more varied — for example, by adding a page number or product specs. That’s a direct sign of a template that inserts too few distinguishing fields.

Google Search Console and search results

The URL Inspection tool in Search Console shows information about a page’s indexed version and lets you check its published version. The tool’s documentation doesn’t list title and metadata as a separate check item. So check the actual appearance of the title and description directly in search results. Keep in mind it may look different for a different query.

Crawler

A crawler is a program that follows a site’s links and collects data about its pages. It exports the title and description of every page it finds and flags the empty and duplicate ones. The «too long» and «too short» filters depend on the thresholds set in the program. More on how to find empty and duplicate titles with a crawler.

Checking metadata is a routine part of a site audit. A separate article covers how metadata is checked as part of an SEO audit.

Checklist:

1. Every indexable page has its own title.
2. The title reflects the page’s topic and the primary query phrasing.
3. The title has no keyword lists and no repetition.
4. The site name sits at the start or end, not in place of the topic.
5. The description describes this specific page and doesn’t repeat the title.
6. The template inserts fields that distinguish pages, and only information that’s actually on the page.
7. Title and H1 don’t contradict each other.
8. After edits, recheck in Webmaster, with a crawler, and in search results.

Common mistakes in title and description

Mistake	Risk	How to fix it
Empty or partially filled title	Google will build the heading from other text on the page	Fill it in: page topic + site name
The same title on many pages	Pages aren’t distinguishable in search results; Yandex notes that users find it harder to navigate	A unique title for each page, or a template with distinguishing fields
A template where only one word changes	It’s hard to tell from the headings how pages differ	Add specs, model, or page number to the template
A list of keywords	Yandex and Google advise against such headings	One primary phrasing and a clear topic
A generic title: «Home,» «News,» «Catalog»	The heading gives no sense of the content	The section’s or page’s specific topic
Title or description doesn’t match the content, or is outdated	Google may replace that heading	Rewrite it to match the actual content, remove outdated dates
Description is a list of keywords	Gives no sense of the page; Google uses the meta description when it’s more accurate than other material on the page	Two or three phrases about what’s actually on the page
Identical or boilerplate descriptions	Yandex most likely won’t show such descriptions	A unique description, or a template with specific fields

Frequently asked questions

Can you stop a search engine from changing a page’s heading in search results?

Google’s documentation has no such setting. Google states that it doesn’t manually change individual sites’ title links: they’re generated automatically. You can only address the underlying causes — they’re listed in the «Why search results show a different title or description than yours» section.

Why does the same page have different headings in Google and Yandex?

Each system generates the heading using its own rules and its own set of sources. For Yandex, the heading can also depend on the query. If neither system is showing your title, start by checking the title itself.

Do you need to write a description for every page on a site?

Both systems advise making descriptions unique. If there are many pages, Google recommends starting with the homepage and the most popular pages, and generating descriptions programmatically for similar, templated pages.

Can AI be used to write title and description?

As a draft, yes. Each version still needs to be checked: whether it matches the page’s content, whether it’s repeated on other pages, and whether it contains a keyword list or information that isn’t actually on the page.