I came back to Athens from Krakow and hit the same wall everyone here does — high rent, few options, and a roommate search that felt unsafe
https://idominikos.com/articles/how-myroomie-started
series: The long way round
part 1 of 2
- 1. MyRoomie started as my own problemyou are here
- 2. Fifteen years in marketing, then I started writing code every day
Most founder stories get told backwards. You find the market, you validate the problem, you build the thing. Mine went the other way: I had the problem, badly, and only later found out how many other people had it too.
The wall #
I had been in Krakow. Coming back to Athens meant finding somewhere to live, and what I found was a market that had quietly become impossible. Rents had climbed past what a salary here supports. The obvious answer — share a flat — turned out to be its own problem.
The listings were a mess of classifieds and Facebook groups. Nothing was verified. Nobody was who they said they were, or at least you had no way of knowing. You met a stranger in a flat and made a decision about the next year of your life in forty minutes.
That is not a housing shortage. That is a trust problem sitting on top of a housing shortage, and the second one is the one software can actually touch.
What I built instead #
MyRoomie started narrow: match people who would actually live well together, and verify that both of them exist.
That sounds small. It is not. Compatibility is not one question, it is fifty — sleep schedule, noise tolerance, guests, cleaning, pets, smoking, how you handle a disagreement about the electricity bill. Most of the friction in a flatshare is not about the flat.
Verification is the other half. A listing that has been checked and a person who has been checked are worth more than a hundred that have not, and that is the part the classifieds could never do because it costs money and attention.
What it turned into #
It did not stay a matching app. Once you are the layer people trust for housing, the questions keep arriving: can you handle the lease, can you handle relocation paperwork, can you manage the property, can you do this for a company moving twelve people.
So it became a housing platform rather than a marketplace — flatshare matching, managed properties, corporate housing, relocation support, verification running underneath all of it. Athens first, then wider. Cyprus is the current focus.
The part I keep coming back to #
The problem was mine. That matters more than it sounds, because it means I never had to guess what the frustrating part was. I had felt it, recently, with my own deposit on the line.
Every product decision since has been checked against that: would this have helped me, that month, in that situation? It is a narrow test and it has been right far more often than any market analysis I have run.
topics covered
related reading
- Thinking about going on Dragon's Den? My advice, don'tarticlesWhat the on-screen deal is actually worth, where the advertised money really went, and the one case where going still makes sense1 shared tag(s): lessons learned
- Your vibe-coded database is a mess (mine was too)articlesSchema drift is the tax of AI-assisted coding — why it happens to every vibe-coded project, and the tool I built to squash it1 shared tag(s): lessons learned
- Making your site agent-readablearticlesBuyers increasingly never visit your site — they ask an AI, and the AI recommends whatever it can parse. The full pattern: markdown mirrors, llms.txt, agents.md, and content negotiation1 shared tag(s): lessons learned