Yandex Is Cracking Down on Content Quality Again. And This Isn't a Routine Threat
Yandex made it official: they're updating the Proxima algorithm, and the goal is to tell quality content from garbage more precisely. Priority shifts toward expertise, trust, real value for the user. Spam and auto-generated content will get cut faster. Separate emphasis on YMYL: medicine, health, anything touching people's money and wellbeing.
Sounds like just another update, another "be more expert" post. But here's what matters: this isn't an abstract recommendation, it's a direct official statement from a tier-1 source, flagged with "official" confidence. Which means this is already built into the algorithm — not sitting in some "someday" drawer.
What's actually changing
If you used to get away with rewriting the top 10 competitors and stuffing in the right keywords — that's increasingly not working. Yandex is methodically moving in the same direction Google went years ago: expertise and trust aren't decoration on top of the text, they're a filter at the entrance.
For YMYL this hits especially hard. Medicine, health, finance — topics where a mistake in the text can cause real harm to a real person. And for betting and gambling verticals, this is a direct hit too, even though YMYL isn't always spelled out explicitly there. A user's money, their financial decisions, trust in the operator — that's exactly the zone where a search engine will demand proof of competence, not just a pretty landing page.
What to do right now
First, and most obvious — stop rewriting competitors. This is a fact: original content built on your own understanding of the topic, not a compilation of someone else's texts, is now literally a ranking factor, not a matter of conscience.
Second — authorship. State who wrote the text, what their qualifications are, where the expertise comes from. For core YMYL niches, this can't be a formal "John Smith" byline — it needs to be real proof: experience, certifications, publications. If you're writing about health or financial products, bring in a subject-matter expert for fact-checking. Not for show. For real.
Third — fact-checking. Mistakes in the text that used to slip through unnoticed will now count directly against you. Before you publish, double-check the numbers, the wording, any statement that reads like a fact.
Fourth — headlines. Clickbait no longer works as a trick to get the click — it works as a signal of mismatch with the actual content, and that's bad news for an algorithm that's specifically learning to catch broken expectations.
UX isn't decoration — it's part of the score
Separately, Yandex is talking about user behavior and site usability. Core content needs to sit above the ads — not buried under three banners and a popover. Search and filtering need to work the way an ordinary person expects, not the way it was convenient for a developer to build three years ago.
This is where a lot of people trip up: they check their site through the eyes of the person who built it, not through the eyes of someone opening it for the first time who just wants an answer. The gap is huge. Take your own site and walk the user's path from zero — you'll often find things your team simply stopped noticing.
Bottom line
Auto-generated content with no value is first in line to get hit. Not because AI is bad, but because content with no value for the reader is bad, no matter who wrote it — a human or a model. AI is a tool for speed, not a license for meaningless text.
If you're running a content or e-commerce project, a YMYL niche, or a gambling vertical — now's the time for an honest audit: who's the author, where does the expertise come from, are there factual errors, is the site actually easy to use. Not because "it's required," but because the search engine is now literally measuring this.
If you want an independent read on exactly where your content and site structure fall short of Yandex's and Google's requirements — let's talk through an audit and consultation.