Jevtown

Jevtown is a social network where people write and 10,000 AI personas read. You post a text, a listing, a product or a headline, and within seconds the town reacts: most scroll past, some like, repost, block, write to the seller or buy. The purpose of the product is to let a writer see how an audience would respond before publishing anywhere that actually matters. Every reader is a computed persona with a name, an age, a job and a city, and the English-speaking town is described as having 10,002 residents. Jevtown is aimed at anyone who writes for an audience — someone drafting a social post, a seller writing a listing, a maker describing a product or a writer testing a headline — and it requires no sign-in to use. The problem Jevtown addresses is the uncertainty that comes before publishing. Once a post, a listing or a headline is out in the world, the reaction is already public: readers scroll past, or they stop, or they block, or they buy, and there is no way to take that back. Jevtown moves that moment of truth earlier. Instead of guessing how a piece of writing will land, you can watch a simulated town react to it in seconds and then decide whether the text is worth publishing at all. The product puts it plainly: a weak text dies for half a cent and a good one reaches everyone in 14 seconds. The same logic applies to a marketplace listing, where the difference between two ways of writing the same item can decide whether buyers write to the seller or smell a scam. Jevtown accepts four kinds of input, each selectable from a single composer: Post, Listing, Product and Headline. The composer allows up to 2,000 characters, shown as a live counter labelled 0 / 2000. When you submit, the English-speaking town reads it — 10,002 residents — and you will see who likes, reposts or blocks it. The site also demonstrates a Ukrainian town: the homepage opens with a saved example labelled "A saved example · the Ukrainian residents" alongside a 01 / 02 selector, and residents in the sample feed carry Ukrainian names and cities such as Dnipro and Zhytomyr. Each resident is identified by a first name, an age, an occupation and a city — for instance Zoe, 20, a student in New York, or Yurii, 56, a nurse in Zhytomyr — and their individual reactions are listed next to their profile. The core mechanic is a wave system that decides how far a piece of writing travels. A post first goes to the 600 residents it should matter to, and it reaches the next 1,500 only if more readers were glad than annoyed. From there it can keep climbing: the feed shows posts that were seen 600 times, 2,100 times, 5,100 times and up to 10,001 or 10,002 times. A listing in the saved example went into a second wave and reached 2,100 personas after readers responded well to it. This staged distribution means that a weak text stops early while a good one reaches everyone, and the counts of how many people saw, stopped, were glad, were sorry, reposted, clicked or bought are reported for each post. Every stage is measured in a small, consistent vocabulary of reactions. Readers can scroll past, stop, be glad, be sorry, repost the post, click a headline, write to the seller, smell a scam or buy the product. The saved iPhone listing shows the pattern: 2,100 saw it, 142 wrote to the seller and 6 smelled a scam, with a Replay control and a visibility setting offering "To the public feed" or "By link only". Because the residents are computed personas rather than anonymous visitors, Jevtown can break their behaviour down by interest, job, age, city and budget, and tell you who stopped, liked, reposted or blocked your text. In the sample feed you can see individual reactions attributed to named residents, such as Iryna, 37, a copywriter in Dnipro who bought it. Jevtown's unique approach is to make the audience itself the product. Beyond feed reactions, a listing or product can be tested with a price ladder: prices can be named in ₴, $, € or £, and every resident who stops at the product is asked for the highest of those prices they would pay. The result shows how many buyers each price gets and which one earns the most — described as a demand curve over a product's price ladder. For listings, Jevtown also reports the questions buyers would ask first. The whole experience runs on the web without sign-in, costs as little as half a cent for a text that dies early, and can show a finished result in 14 seconds. The public feed itself can be sorted by Latest or Travelled furthest, and at the time of writing displayed 148 texts seen 260,005 times. The benefit for a user is a decision made before publication instead of after. Instead of publishing and waiting, you get an immediate reading of whether the text connects and who it connects with: the people who stopped, the people who were glad, the people who reposted and the people who blocked. Sellers learn the questions buyers would ask first, which can be answered inside the listing itself, and they get a view of which price point earns the most. Headlines can be compared by how many people stopped and clicked versus how many were sorry. Because each experiment is cheap and fast, a writer can afford to test several versions of the same text and keep the one the town responds to. The clearest use case shown on the site is a marketplace listing. The same iPhone 13 listing is written two ways — one version offering full details and payment on inspection, the other demanding advance payment only — and the site shows how the town reacts to each, including the number of personas who wrote to the seller and the number who smelled a scam. Another use case is the headline: the feed shows headlines such as "New research solves cancer" that were seen 600 times, with hundreds of residents stopping and dozens clicking. Products can be tested with the price ladder to compare demand at different prices, and general posts — opinions, jokes, technical write-ups, ideas — can be run through the town to see whether the audience stops, is glad, reposts or simply scrolls past. Jevtown is for people who write for an audience and want a fast read on it before going public: sellers writing marketplace listings, marketers writing headlines, makers writing product descriptions and anyone posting to a social feed. The product runs in the browser, requires no sign-in and costs as little as half a cent for a text that dies early. The sample content on the site covers both an English-speaking town of 10,002 residents and a Ukrainian town, and prices for the price ladder can be expressed in hryvnia, dollars, euros or pounds. The site mentions no third-party integrations and does not describe a technical stack, so the experience appears to be entirely self-contained. Jevtown is a sandbox for writing in public without the consequences: a town of 10,000 computed residents who read your post, listing, product or headline in seconds and tell you — by scrolling past, stopping, liking, reposting, blocking, writing to the seller or buying — whether the text is worth publishing. A weak text dies for half a cent; a good one reaches everyone in 14 seconds.