How Yandex Zen Became What It Is

Yandex Zen launched in 2015 with a compelling pitch: a recommendation engine that studies your interests and serves you exactly the content you want. An algorithmically curated personal feed. Read articles, and over time, the algorithm learns what you like — making your feed more relevant with every interaction. Beautiful idea. The reality turned out to be anything but.

By 2018, Yandex had sweetened the deal with monetization for content creators. This is where things went off the rails completely. Yandex decided to pay for views — not quality, not expertise, not usefulness. Just raw views. When you pay for views, people optimize for views. It does not take a behavioral economist to predict what happens next: the most aggressive clickbait wins, serious work drowns, and the entire information ecosystem tilts toward garbage.

By 2024, Zen completed its transformation into a walled garden. External links are algorithmically punished with reduced reach — this is official policy. Authors are incentivized to write articles that go nowhere, link to nothing, and provide zero utility outside the platform itself. The perfect incubator for information waste.

The Economics of Crap: Why Authors Produce Garbage

Let us look at the numbers. The average payout per 1000 views on Yandex Zen ranges from 20 to 80 rubles — roughly 20 to 80 cents USD. An article that reaches 10,000 readers earns its author between 200 and 800 rubles. Eight hundred rubles for an article read by ten thousand people. Let that sink in. Writing a quality piece — with actual research, fact-checking, unique data, and decent prose — takes days. And it pays less than a food delivery driver earns in an hour.

The economics are brutally simple. The less time you spend on an article, the higher your effective hourly rate. Conversely, the higher the quality, the less money you earn per unit of time. This is not a flaw — it is a feature of Zen's monetization design. The platform created conditions where producing quality content is financially irrational. The market does not select for experts. It selects for spammers.

\u{201c}

I left Zen when I realized my best articles — the ones I spent three days on — earned ten times less than a clickbait headline about "how I lost 30 kg in a month." The platform does not reward quality. It rewards the ability to trick a reader into clicking.

Throw in the referral bonus systemBonus payments for bringing new users to Zen and you get a full-blown industry of content sludge production. People run dozens of channels, each pumping out five to ten articles a day — written with zero effort or auto-generated through ChatGPT. The algorithm does not distinguish between an expert and a hack. It only sees numbers: clicks, scroll depth, time on page.

The Read-Through Coefficient: The Metric That Killed Meaning

One of Zen's core quality metrics is the read-through rate — the percentage of readers who scroll all the way to the bottom. The theory: if someone reads the whole article, it must have been engaging. Authors quickly learned to exploit this. How? By inflating the introduction so the reader scrolls endlessly without finding the point. Or, even simpler, by burying the answer promised by the headline at the very end of the text.

Classic example. Headline: "These 5 Foods Are Destroying Your Stomach. You Eat Number 3 Every Day." The reader clicks to find out what food number three is. To get there, they must scroll past half the article — past ad blocks that count as impressions and generate revenue. By the time the reader reaches the actual content, the author has already pocketed their pennies, and the reader has consumed nothing but filler.

This metric also birthed an entire genre of "infinite articles" — pieces where the author circles the same point endlessly, repeats identical ideas in different words, and deliberately avoids specificity. The goal is not to inform. The goal is to make you scroll. Longer scroll equals higher read-through equals more ad impressions equals more money.

Clickbait: The Engine of the Platform

Headlines on Zen are a unique art form — or rather, an anti-art form. Here are the standard templates driving the most clicks:

• "You Won't Believe What..." — the classic bait promising a revelation where none exists
• "X Things They Are Hiding From You..." — conspiracy-bait, even when the topic is cactus care
• "I Tried It And Was Shocked..." — fake personal experience, usually fabricated from scratch
• "Never Do This..." — forbidden fruit sells, even when "this" means using a specific toothpaste brand
• "Doctors Are Shocked..." — immortal classic, needs no commentary

The problem is that Zen's algorithms actively reward these headlines. They generate high CTR — and CTR is the algorithm's proxy for quality. A vicious cycle forms: clickbait headlines get more clicks, the algorithm marks them as high-quality, promotes them further, they get even more clicks. Honest headlines — informative, accurate, calm — lose the CTR battle and sink to the bottom of the feed, never to surface again.

The inverse correlation between headline honesty and headline effectiveness is nearly perfect. A platform designed to spread quality content has become a headline spam engine.

Who Actually Writes for Zen: Portrait of the Typical Author

Let us be frank: serious journalists, domain experts, and professional writers steer well clear of Zen. Why? First, money — you cannot earn a livable wage on Zen without resorting to outright spam. Second, reputation — publishing on Zen is viewed by the professional community as a mark of amateurism. Third, zero editorial standards — your work sits next to conspiracy theories about flat Earth and garlic curing cancer, and that adjacency affects your brand whether you like it or not.

So who fills the platform with content?

• Students and teenagers — writing for pocket change, often paraphrasing Wikipedia or other people's articles
• Stay-at-home parents and retirees — the platform actively marketed itself as a "work from home" income opportunity. They write about recipes, parenting, gardening, and politics
• Content farms — hiring cheap copywriters or using AI for mass article generation
• Neural networks — since 2023, the platform has been flooded with fully AI-generated articles from ChatGPT and its clones

A typical Zen author's day: open Yandex Wordstat in the morning, check which search queries are trending, and write an article on a hot topic. Then another. Then a third. By evening, five to seven articles are done — all shallow, all with clickbait headlines, all designed to harvest views and be forgotten. No fact-checking. No editing. An assembly line.

Algorithmic Degradation: Why the Smart Feed Gets Dumber

Remember the promise of a "smart feed" that learns from your interests? Forget it. In practice, Zen's algorithm follows a primitive scheme: show users what already has lots of clicks. This is the Matthew Effect in action — the rich get richer, the poor get poorer. Popular posts grow more popular, while new and potentially interesting material drowns without ever surfacing in anyone's feed.

Here is how the algorithmic sausage is made. A post gets published. The algorithm shows it to a small test group — say, a hundred users. If ten of them click, that is a 10% CTR. The algorithm marks the post as promising and expands its reach. But what if those first hundred users were bots, or random people who were tricked by a deceptive headline? The algorithm does not care. It does not check whether the user bounced back to the feed two seconds later. It only registers the click.

This creates a feedback loop: clickbait headlines get more clicks, the algorithm marks them as quality, shows them more, and they get even more clicks. Meanwhile, honest articles — the ones that do not scream, do not promise miracles, do not lie — get low initial CTR and are algorithmically executed. They simply stop being shown.

The result is predictable: after a month of active Zen use, your feed becomes an undifferentiated torrent of identical garbage. You will not find a single article you would have voluntarily chosen to read. The algorithm forced them on you, having decided that a loud headline equals quality content.

Moderation: Or the Lack Thereof

The moderation system on Zen deserves special attention. Technically, it exists. Articles go through pre-moderation — algorithmic and, supposedly, manual. In practice, moderation operates on a "anything goes unless it has swear words" principle. The platform blocks explicit profanity, calls to violence, and blatantly illegal content. But text quality, factual accuracy, and fact-checking all fall outside the scope.

I have personally seen articles on Zen claiming that COVID-19 is curable with garlic, that the Earth is flat, that vaccines contain 5G microchips, that Elon Musk is an alien, and that drinking hydrogen peroxide cures cancer. All of these passed moderation. Not because the moderators are morons (though who knows), but because Zen's moderation does not verify facts. It only checks compliance with the platform's formal rules. And the rules do not require information to be true.

\u{201c}

Yandex Zen is a disinformation factory. The platform's algorithms cannot distinguish truth from fiction, and the monetization system rewards sensationalism over accuracy. Millions of people consume content daily where a significant portion is outright falsehood.

Zen vs. The World: How It Compares

To understand the scale of the disaster, let us compare Yandex Zen with other content platforms:

Zen loses on virtually every dimension. Its only "advantage" is the low barrier to entry and instant monetization. But that advantage is also its fatal flaw: the barrier is so low that anyone with anything floods the platform. No qualification requirements. No expertise verification. No minimum quality threshold. Just sign up and publish — multiplication tables, soup recipes from axes, conspiracy theories, whatever.

What Comes Next: A Degradation Forecast

Where is Yandex Zen heading? The forecast, unfortunately, is grim. The platform sits in a classic quality trap: worse content generates more clicks, more clicks generate more ad revenue for Yandex, and more revenue removes any incentive to change anything. It is a self-reinforcing degradation loop.

The arrival of accessible AI has made things exponentially worse. You no longer need to know how to string words together. Give ChatGPT a headline, and in sixty seconds you have a complete article. Dumb, faceless, stamp-ridden — but formally meeting the platform's minimal requirements. Zen's moderation does not penalize AI use (and how could it, when Yandex's own recommendation engine runs on AI?). The result: content volume grows exponentially while content quality plunges.

Some analysts predict that Zen will follow the path of Yandex's now-defunct Narod hosting service — popular once, gradually turning into a wasteland, eventually shuttered. Others argue the platform is too important to Yandex's ecosystem (it keeps users inside company services) to be killed. But both camps agree on one thing: content quality will continue to decline.

How Authors Game the System: Engagement-Hacking Techniques

Since we are discussing degradation, let us examine the toolbox authors use to game their metrics. None of this is secret knowledge — it is discussed openly in themed chats and forums across the Russian internet.

• Clickbait headlines — we covered these, but they are so fundamental they deserve a second mention
• Text inflation — repeating the same idea in different words, over and over. Starts with filler, ends with filler, filler in the middle
• Dummy paragraph breaks — technically increases "reading time" because the reader spends time scrolling between artificially short paragraphs
• Provocative questions with no answers — the headline promises to reveal a secret, but the text circles endlessly without reaching any conclusion
• Read-through manipulation — burying key information mid-text or at the end, forcing the reader to scroll
• Copy-pasting and rewriting — taking someone else's popular article and tweaking it minimally. The algorithm does not check uniqueness against other Zen publications

All of these techniques are, in essence, algorithmic exploits. Authors are not creating value — they are finding vulnerabilities in the ranking mechanisms and exploiting them. The platform periodically updates algorithms, closing some loopholes, but authors immediately find new ones. It is an endless arms race, and quality content is the primary casualty.

Why Zen Will Not Die

Despite everything said above, Yandex Zen is unlikely to disappear anytime soon. The reasons are banal:

1. Massive audience. Millions across Russia and the CIS use Zen daily — simply because it is baked into the Yandex homepage and Yandex Browser. No need to go anywhere. The feed shoves itself in your face.
2. No alternatives. The Russian-language internet has no other recommendation platform of comparable scale. Mail.ru's Pulse never took off. Zen News is the only game in town.
3. Habit. People got used to scrolling the feed with morning coffee or on public transit. It became a ritual — and rituals are sticky.
4. Advertising budgets. Yandex pours serious money into Zen because the platform generates ad revenue and keeps users inside the ecosystem.

Conclusion: What a Reader Should Do

If after reading all this you want to uninstall Yandex Browser and forget Zen like a bad dream — I understand. But let us avoid extremes. Here are practical recommendations:

• Do not use Zen as a primary information source. It is an entertainment platform, not a news outlet. Treat it accordingly.
• Verify facts. Anything you read on Zen that seems important — cross-check it through other sources.
• Do not click on clickbait. Every click on a screaming headline is a signal to the algorithm: "Show me more of this garbage."
• Subscribe to specific authors. If you find a quality channel, subscribe and read it directly — skip the recommendation feed entirely.
• Use alternatives. Telegram channels, Substack, Medium — there, authors earn based on quality, not click volume.

Yandex Zen is not a platform problem. It is an attention economy problem. As long as advertisers pay for views instead of engagement, as long as algorithms measure quality by CTR instead of accuracy, as long as authors earn more from clickbait than expertise — nothing will change. Not on Zen, not on any platform running the same model.

Frequently Asked Questions