Most sites don’t choose their titles; they inherit them. At launch a developer sets a default in the template: the company name and a slogan in the title, one paragraph about the company in the description. Then the catalogue grows. Products arrive from 1С in batches of hundreds, filter and sorting pages multiply, a blog appears, regional versions go live. Nobody writes meta tags for any of it, so every new page takes whatever the template gives it. Crawl the finished site and you find two thousand URLs and forty distinct titles. In search results the pages look the same: the same name, the same sentence about «a reliable supplier since 1998». The site is in the index, but no page tells the searcher what is on it.
WHAT COUNTS AS EMPTY OR DUPLICATE
An empty title or description is the simple case: the tag is missing or contains no text. The search engine then writes the result itself, using the H1, link text or whatever fragment of the page looks relevant. Sometimes that works. Often the snippet is assembled from breadcrumbs, a menu item, a cookie notice or the words «Price on request».
Duplicates are harder to spot, because every page has a title, just not its own. Typical cases:
- The same title on every page: the brand name, «Home» or «Catalogue».
- One description inherited by the whole site from a single root setting.
- A template whose variable part is empty, so the title reads «— buy at a price of rub.».
- Dozens of pages whose titles differ by one word, or not at all, and share the same long boilerplate tail.
Templates are not the problem in themselves. A short formula where the changing part carries the meaning works: «Bag filters — wholesale and retail | Brand» and «Cartridge filters — wholesale and retail | Brand» are both clear. Trouble starts when the fixed part is most of the text. Google’s guidance calls out long titles that vary «by only a single piece of information», and says identical or similar descriptions on every page «aren’t helpful when individual pages appear in search results».
It also matters which pages carry the duplicates. Twenty product cards with one title are a meta-tag problem. Twenty URLs of one category with different sorting are a different problem: the pages themselves are copies, and the identical tags are only the symptom.
The distinction that matters: the title tells the search engine what the page is, and the description tells the searcher why to click. A duplicate fails at both.
WHY IT COSTS YOU TRAFFIC
The title is one of the clearest signals of what a page is for. When twenty pages share one title, the search engine has to work out from other signals which of them answers the query. It often picks the wrong one, or alternates between them. Pages end up competing with each other for the same queries, and none of them holds a stable position. When the titles, the descriptions and most of the content match, search engines are also more likely to treat the pages as duplicates and show only one of them.
Yandex is explicit about it: identical titles and descriptions worsen a site’s visibility in search. Webmaster counts them on the Summary page, lists them under Indexing → Titles and descriptions and raises an alert in Site diagnostics.
The cost shows up in clicks as well as rankings. The snippet is the only advertisement you get for free, and the searcher compares snippets, not sites. Take two results for «polyester bag filter»:
- «Company — official website. A reliable supplier of filtration equipment since 1998.»
- «Bag filter FRM-200, polyester, 0.5 µm — price from 1,200 rub., in stock, delivery in 2 days.»
The second one gets the click even from a lower position. Pages with generic snippets lose traffic at positions they have already earned.
There is also a loss of control. When the title is missing, generic or boilerplate, Google writes its own title link from the H1, anchor text, og:title or other text on the page. Yandex can build its own snippet in the same way. What appears in the results is then a breadcrumb trail or a random fragment, not the wording you would have chosen.
The damage reaches beyond search. Messengers and social networks build link previews from og: tags and fall back to the title and description. A product link sent in Telegram shows the company slogan instead of the product.
AI-driven search adds another layer. AI Overviews, Yandex Neiro and Perplexity usually label the sources they cite with the page title. A citation reading «Catalogue — Company» tells the reader nothing about why the page is worth opening.
WHY THIS MISTAKE KEEPS HAPPENING
Defaults fill the gap. When no SEO field is filled in, a CMS template outputs the site name, the H1 or an inherited value. In 1С-Bitrix, a title or description set in the properties of the root section is inherited by every page below it. One entry made at launch quietly becomes the description of the whole site.
The catalogue grows faster than anyone writes. Products come from 1С or a supplier feed with a name, a price and an article number, and nothing else. Meta-tag templates for the catalogue are either not configured or rely on fields that are empty for half the products.
URLs multiply on their own. Filters, sorting, pagination, tags and UTM-tagged links all create new addresses, and each one inherits the title and description of its parent category. One category becomes forty URLs with identical meta tags.
New pages start as copies. A landing page or a regional version is cloned from an existing one, and the meta tags are the last thing anyone edits. Thirty city subdomains end up titled «… in Moscow».
Migrations leave the fields behind. The new site goes live with the content transferred and the SEO fields lost, because on the old site they lived in a separate module or plugin that nobody exported.
And nobody owns the meta tags. Copywriters write the text, developers build the templates, and the SEO audit happens once a year. Titles and descriptions fall between all three.
THE MOST COMMON FORMS IN PRACTICE
The brand-only title. Every page says «Company — official website». The homepage ranks for the brand name, and the product and category pages rank for nothing.
The inherited description. One paragraph about the company, written at launch, sits on thousands of pages. The best case is that search engines ignore it and build snippets from the page text. The worst case is that they show it.
The broken template. «{Name} — buy at a price of {price} rub.» with an empty name or a zero price produces «— buy at a price of 0 rub.». It looks like an error to the searcher, because it is one.
The pagination and filter clone. /catalog/bag-filters/, ?PAGEN_1=2 to ?PAGEN_1=40, ?sort=price and /filter/material-polyester/apply/ are all titled «Bag filters | Brand». Here the tags are the smaller issue. The real question is whether these URLs should be in the index at all.
The regional clone. City subdomains or folders share one title and description. Either the city is not substituted at all, or it is substituted in the nominative case, so a Russian title reads «купить в Новосибирск» instead of «купить в Новосибирске».
The identical variants. A separate product card for each size or colour, all named «Filter bag», all with the same description. Twelve pages, one title.
HOW TO FIX IT
Start with an export, not with writing. Crawl the site and export the URL, status code, indexability, canonical, title, description and H1 for every page. Group the pages by title and by description. The groups show which page types produce the duplicates and which problem you actually have.
Separate duplicate pages from duplicate tags. If the URLs are copies of one page (sorting, UTM tags, upper- and lower-case variants, print versions), unique titles are the wrong fix. Close those URLs from indexing or point them to the main version with rel=»canonical», the Clean-param directive for Yandex or 301 redirects. Yandex recommends exactly this for GET parameters and case variants. Unique meta tags are for pages that are meant to rank.
Write by hand where it pays. The homepage, main categories, key product lines, service pages, landing pages and top articles deserve titles and descriptions written individually. To choose them, sort pages by impressions in Search Console and Yandex Webmaster: these are the snippets people already see.
Use templates for everything else, by page type. A good template is a formula with meaningful variables, not one sentence for all pages:
- Category: «{Category name} — {main type or application} | {Brand}».
- Product: «{Product name} {model or size} — price, specifications | {Brand}».
- Product description: «{Product name}: {material}, {key specification}. Price {price} rub., {availability}, delivery to {city}.»
- Pagination: the category title plus «— page {N}». Yandex suggests adding the page number to the title for exactly this case.
- Regions: the city name stored in the right grammatical case as a separate field.
Build in fallbacks. If a variable is empty, the template switches to a shorter version or to the H1, and a zero price is never printed. In 1С-Bitrix these templates live on the SEO tab of the infoblock settings; in WordPress, in the SEO plugin.
Put the meaning first. Search engines truncate long titles, and the cut-off depends on pixel width and device. Keep the part that distinguishes the page within the first 50–60 characters and move the brand to the end. In the description, give the searcher something the title can’t: price, range, stock, delivery time, warranty.
Change the process. Title and description become required fields whenever a page or product is created. Meta-tag templates are part of every catalogue import and every migration checklist. After each release or migration, re-crawl the site and compare the number of unique titles with the previous crawl.
HOW TO CHECK IF YOU HAVE THIS PROBLEM
Crawl the site. Screaming Frog SEO Spider (free up to 500 URLs) has Missing and Duplicate filters on its Page Titles and Meta Description tabs. Netpeak Spider, Sitebulb and the site audits in Ahrefs and Semrush report the same issues. Look at the share of indexable pages with a unique title. If it is well below 100%, find out which page types the duplicates come from.
Check Yandex Webmaster. The Summary page shows the number of identical titles and descriptions. Indexing → Titles and descriptions lists them, and Site diagnostics raises an alert when there are many.
Check Google. Search Console no longer has a report on titles and descriptions; the old HTML Improvements report was retired along with the old interface. Use the Performance report instead: sort pages by impressions and look for pages with many impressions and a low CTR. Then search for their main query and compare the title Google shows with yours. If they differ, Google has rewritten it, usually because yours was empty, generic or boilerplate.
Look at your site the way searchers do. A site: search returns your pages as a list of results. If the first screen is ten identical titles, that is what searchers see too.
And the simplest check of all: take your ten most valuable commercial pages, copy their titles and descriptions into one table and delete the URLs. If you can’t tell which row belongs to which page, neither can the search engine, and neither can the person deciding whether to click.