How Google and Yandex Decide What to Show at the Top

If you've heard SEO acquaintances toss around phrases like "we need to account for Spam Brain 2026," "get there through E-E-A-T," and "close YMYL for the medical cluster" — forget all that for an hour. In reality it's simpler, and search engines have long worked on clear principles that are easy to explain without jargon.

Let's figure out exactly how a search engine decides what to show at the top for your query.

The four ranking blocks

Any algorithm (Google, Yandex, Bing — doesn't matter) gathers candidates internally and sorts them by the sum of their signals. There are hundreds of signals, but they group into four big blocks.

Block 1: Relevance

Does the page text match what the user searched for?

The most basic thing. The search engine looks at your page and judges: does it even talk about this topic at all. Signals:

• The page title (Title) contains the query's keywords or their synonyms.
• H1 and the H2/H3 subheadings cover the query's topic.
• The page text mentions the query and its variants at a natural density.
• The URL contains a readable slug, not "product?id=12345".
• Structured data (Schema.org) gives the search engine an explicit hint about what type of page this is (product, article, recipe, event).

Relevance is your "entry ticket to the game." Without it, the page won't even become a candidate. If yours is lower than a competitor's, the competitor ranks above you.

Block 2: Content quality

Does the page satisfy the user's intent well? Does it cover the topic in depth?

This one is trickier. A search engine doesn't "read" text the way a human does, but it has ways to gauge how much deeper you go than competitors:

• Length and structure — long, well-broken-up articles are usually better than short ones.
• Coverage of related subtopics — do you address the questions a user will ask next.
• Freshness — for topics that change, the date of the last update matters.
• Uniqueness — not a copy of other pages on the internet.
• Multimedia — images, diagrams, video, tables (especially in Google — a big weight).
• Fully satisfying the intent — if the user searched "how to fix," the page should teach them to fix it, not talk them into hiring a repairman.

In Google this is packaged into the concept of E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness). In Yandex — into the "complex quality factors" behind ICS. The essence is the same: content must be useful, not just "optimized."

Block 3: Domain authority

What's the site's reputation overall?

Search engines assess how trustworthy your domain is in general. Signals:

• Domain age — older = more trust.
• Number and quality of external links — who links to your site.
• Brand mentions — do people mention you without a link (Yandex sees this).
• Social signals — social-media posts with a link to you.
• Site size — many pages on the topic = a specialist in the topic.

This is the slowest part. Authority isn't built in a month. If a domain is 2 months old, you won't rank for competitive queries no matter how well the content is written. That's why new sites aim at the long tail and mid-frequency queries — where authority isn't critical.

Block 4: UX signals

Did people stay on the page after clicking? Or leave right away?

This is the most "alive" part. The search engine watches how users interact with your page after they click it in the results:

• CTR in the results — do people click your snippet (depends on title + description + URL).
• Pogosticking — did the user bounce back to the results within 5 seconds (bad).
• Time on page — how long they spent (but this is a weak signal).
• Scroll depth — did they read to the end (for articles).
• Conversions and micro-conversions — did they click into the product card, share it, add it to favorites.

UX signals rank above all the others, because they're direct: the search engine isn't guessing, it sees the real behavior of real people.

Why you don't control authority but can control the rest

How your influence breaks down:

So 2 of the 4 blocks you can close at 90-100%, and that alone gives you a huge advantage over sites that churn out content just to tick a box. Authority is the cherry on top, but you can live without the cherry, especially on mid-frequency queries.

Yandex vs Google differences in 2026

The algorithms are 70% alike, but there are nuances.

The practical takeaway: for Yandex, it's more important to nail the behavioral signals (a great snippet, a direct answer in the first screen, a convenient mobile view). For Google — cover the topic more deeply and demonstrate expertise (dates, authors, sources, updates).

What definitely doesn't work in 2026

A few things that worked 5-10 years ago and still surface in old SEO guides, but are now useless or harmful:

• Cramming keywords into the text ("keyword density of 3-5%"). Used to work, now it's a filter for over-optimization.
• Buying links from marketplaces. These links are recognized in a second, a filter is guaranteed.
• Faking behavioral signals with bots. Yandex is especially harsh: a ban is almost instant.
• Hidden text, white on white. Spotted instantly, a dedicated filter.
• Cloaking — showing bots one kind of content and people another. Also an instant ban.
• EMD domains ("buy-cake-moscow.com"). Used to give a boost, now it's a "spam site" signal.

Modern SEO is about usefulness and depth, not about tricking the algorithm.

Next article: how many pages a website needs for SEO.