The whole thesis in six points
- Big platforms in their late stage live by one rule: pay or die. They gathered an audience, became indispensable and now charge for what used to be free. Cory Doctorow named this enshittification.
- They cannot give it up: their valuations and promises to investors are built on the squeeze. So the opposite position is vacant: a free and useful core, earning on real value, and the user is never a hostage.
- This is not charity but strategy: less margin per user, a huge audience, and trust the giants cannot copy.
- The window is open now: AI has collapsed the cost and the time of building products, and speed no longer protects the giants.
- The winner is whoever first takes the “we are not milking you” position in each category.
- For investors: the products launch in a strict order, first the cheaper ones that build trust, then the most expensive strike. Early money is cheaper than late money.
The diagnosis
The internet has entered its milking phase.
The big internet today runs on a single principle, and it sounds blunt: pay or die. A platform gathers an audience, becomes a place you cannot go around, and then starts charging for what used to be air. This is not an accident or the greed of a few managers. It is a pattern.
It even has a name. In 2022 the writer and tech critic Cory Doctorow coined the term enshittification, the decay of platforms, and the word proved so apt that it was named word of the year in 2023, then in 2025 grew into one of the most talked-about books about technology.
The logic he describes is always the same, in three beats. First the platform is good to users, to gather and lock them in. Then it makes things worse for users to please the business clients who pay. And finally it squeezes the business clients themselves, to take all the value for itself. After that the platform dies, but it dies slowly, because people have nowhere to go.
Pay or die is exactly that third beat. The late stage. And the late stage, for all its profitability, leaves a huge hole in the market.
What it looks like
Not in theory, but on the invoice.
To avoid empty words, three examples. All three come from real experience, not from slide decks.
Listings boards
Take OLX. What was once a free place to sell anything has gradually turned into a paywall, category by category: first cars, then real estate, then motorcycles, farm equipment, down to a separate charge for a spare-parts listing. In real estate and cars you can no longer post for free at all, and in some categories the price is dynamic, tied to the most expensive item in your portfolio. The cherry on top: if the platform itself removes your listing for breaking the rules, the spent unit of your package is not refunded.
Right next door is Otodom, from the same group that owns OLX, OTOMOTO and the service marketplace Fixly. On 6 May 2026 Poland's competition authority UOKiK opened an inquiry into its new pricing rules. The trigger is telling.
2,380 → 3,010 zł a month
A professional Otodom package that rises after six months. A quarter more, with not a single change to the service itself. This is not a market, this is leverage being used.
Service marketplaces
Fixly and its kind. The model is elegant and merciless: nominally free, but to respond to a request a tradesperson spends credits, and credits run out, and you pay for them again and again. The contractor pays not for a result, but for the very chance to try to win work. Reached the client or not, matched on price or not, it does not matter, the lead is already paid for. It is a permanent subscription to lottery tickets, and the one paying is the most vulnerable link in the chain: the person who simply wants to work with their hands.
Advertising on Meta
Here the hole is biggest, and the numbers say it better than any words. Small businesses either get stuck setting up Business Manager, or do set it up and get banned within the first minutes, by an automated system, with no human, no clear reason and no real way to appeal. People describe it the same way: the account is disabled a minute after the first attempt, the appeal is reviewed by a bot, and there is no live chat.
200M+ businesses vs 10M+ advertisers
Between those two numbers sit tens of millions of companies that could advertise but do not, because the path there is built to be hostile.
Three different markets, one and the same disease. The platform is stronger than the user, and it knows it.
The thesis
A user-friendly platform is not charity, it is strategy.
The obvious takeaway from Doctorow's story is usually the pessimistic one: since decay is inevitable, all that is left is to endure it. We think the opposite. If all the big players are structurally obliged to milk, because their valuations and their promises to investors are built precisely on growing the squeeze, then the position left vacant is exactly the opposite one. A platform that deliberately refuses the third beat. One that keeps its base free and genuinely useful, earns on added value and on the businesses that actually get a return, and never takes the user hostage.
This is not kindness for its own sake. It is entry into the market through the very door competitors have bricked up with their own hands.
Let us say it honestly upfront: such a model brings in less than today's monumental leaders, an order of magnitude less margin per user. But less margin multiplied by a huge and universal audience is still a very respectable and, more importantly, durable business. Because it has what the milkers traded for quarterly profit: trust. When people stay with you not because they have nowhere to go but because you are not milking them, that trust becomes the moat the big ones filled in long ago.
And there is a reason this window opened right now. The giants used to have one ironclad argument: building a full alternative is long and expensive, so put up with our paywall. That argument has fallen apart. The AI boom has collapsed the cost and the time of building products. What took a team and a year yesterday is now done with a small effort in weeks. The milkers are trapped: too clumsy and too dependent on milking to give the base away for free, and no longer able to defend themselves with development speed.
The future of platforms is democratic, user-friendly and free at the core. The winner will not be the one with the better technology, but the one who first takes the “we are not milking you” position in each category.
For this scenario we are building not one platform but a connected roadmap of three projects, built in a strict sequence: first the ones that are cheaper and faster to earn trust, and only then, on that trust, the biggest and most expensive strike. All of them are built by EPAV AI, a studio where engineers work together with AI: that is why a product takes us weeks, not a year.
Platform 01 · live
Skilcrew
Skilcrew is hiring and finding work through your own AI assistant. One-off services, permanent jobs and IT live on one platform, and the assistant, Claude or another, posts a vacancy, finds candidates or finds work in plain language. It starts in the Baltics, Estonia, Latvia and Lithuania, and scales worldwide from there.
The basic functionality is free: post, find, get in touch, at no charge. On top there are paid subscriptions for those who need more. So far this sounds like any freemium, but next comes the move that turns the whole Fixly model on its head.
Instead of charging the tradesperson again and again for the mere chance to see a request, Skilcrew spends part of its own revenue on advertising its service providers on Facebook and Instagram. The platform pays to bring work to the contractor, rather than taxing the contractor for wanting to work. In a classic service marketplace the tradesperson is a cash cow. On Skilcrew the tradesperson is a partner the platform invests in.
Open SkilcrewPlatform 02 · private beta
TrustMeta.pro
TrustMeta.pro closes the hole of 190-plus million businesses. It runs Facebook and Instagram ads under a company's name and points them to any page: a Skilcrew profile, a Shopify store or any other site.
The user does not need to create a Business Manager, figure out the pixel, pass verification and then live for months with the risk of a ban that cannot be appealed. Ads on the platform run from agency accounts that have reputation and stability. For a small business it comes down to essentially one button.
It is in private beta and already runs ads: it is the very mechanism Skilcrew uses to promote its service providers. TrustMeta.pro accepts investment.
Open TrustMeta.proPlatform 03 · in development
The flagship on Skrew Engine
The third project is the most ambitious. It is a democratic marketplace that moves onto the turf of the marketplaces themselves and strikes at the ones we started this conversation with: OLX, Otodom and their like. The same logic as everywhere, but now applied to the fattest and most defended segment, listings boards and classifieds, where today you are charged for every step. It will be built on our own Skrew Engine, the same technological foundation that powers the first two platforms. Our own engine means we launch new verticals faster and cheaper than competitors.
But there is a nuance, and it is central to understanding the whole strategy. This project is deliberately placed last. Not because it is less important, on the contrary it is potentially the largest of the three, but because going head to head with entrenched giants from zero user trust is the most expensive scenario possible. So it launches not now, but after the first two platforms have accumulated trust, audience and brand reputation.
The investment it requires is inversely proportional to the trust Skilcrew and TrustMeta.pro accumulate by the time it launches. Every unit of trust earned now is money that will not have to be spent later acquiring users for the flagship.
Already working · EPAV AI's own product
EPAV Desk: the same principle in customer support
Help desks charge for every resolved conversation and every agent seat: the better the bot works, the bigger the bill. EPAV Desk does the opposite: one subscription for the whole system, AI already included, no fee per resolution and no fee per seat.
It is already live and shows how quickly the same team turns the thesis into a working product.
About EPAV DeskHow it connects
One idea that reinforces itself.
Skilcrew, TrustMeta.pro and the future flagship on Skrew Engine are not a random set of projects. They are three points of entry into one and the same thesis, and they feed one another.
- Skilcrew is itself a client of TrustMeta.pro: it advertises its service providers through the very agency mechanism TrustMeta.pro provides to everyone else.
- A service provider on Skilcrew who wants even more clients can also advertise through TrustMeta.pro.
- A Shopify store owner who cannot launch ads on their own becomes a direct TrustMeta.pro client.
Each platform feeds the other and at the same time serves as living proof that the model works. And both together work for the third project: when its time comes, the flagship starts not from zero but on the trust, audience and reputation that Skilcrew and TrustMeta.pro pass on to it as an inheritance.
The result is a combined approach: a service platform free at the core, plus accessible advertising infrastructure on top of it. It is exactly these combinations that let you win the market now, on cheap development and a tired audience, and keep it later, once trust is already accumulated.
Investment
Why now is the time to put money in.
Let us bring the investment logic together, because here it is unusually favourable.
Money works harder now
The same budget today builds what would have taken a whole team and a year yesterday. Low burn, fast launches, and capital enough for not one but several attempts.
A measurable prize
Not a theoretical trillion. Concrete holes with concrete numbers: 200M+ businesses against 10M+ advertisers on Meta, and packages rising to 3,010 zł a month on classifieds.
Durable and distributed
Lower margin per user, but a huge audience makes a large absolute number. Trust cannot be copied without wrecking the giants' economics, and three platforms spread the risk.
And most important: early money here is cheaper and stronger than late money. Investing in Skilcrew and TrustMeta.pro today is, in effect, prepaying the cost of the future flagship at a discount. Every euro spent now on trust and audience removes a many-times-larger future cost of a frontal assault on the classifieds market. Entering at this stage means buying at once both the lowest valuation before traction and the maximum leverage across the whole roadmap.
The inverse proportion.
As the first two platforms build trust over time, the cost and the risk of launching the flagship fall. The earliest euro buys the most product.
The window is closing
The place on this market is being taken right now.
The pay-or-die era is profitable, but late and therefore fragile, and we are already seeing it crumble from every side at once. At such a moment the market is taken not by the richest or the loudest, but by the one who first takes the “we are not milking you” position in their category. The big players structurally cannot take that position, because their entire economics is built on the opposite.
Skilcrew, TrustMeta.pro and the flagship on Skrew Engine are a bet on exactly this. Not on the hype around AI and crypto, but on the boring, huge and underserved reality: tens of millions of people and businesses who need simple, honest tools instead of paywalls. The roadmap is built so that each next step is cheaper and surer than the last, and the most profitable window to enter is the earliest. Democratic SaaS is not a slogan. It is the next step, and the place on it is being taken right now.