Berlin · buyer notes
Berlin has more studios per square kilometre than any other German city, and almost no two of them sell the same thing. These notes are for the person on the buying side: what the local market actually looks like, what the money buys, and where the mismatches happen.
Berlin absorbed two decades of startup money and a steady inflow of engineers from across Europe, and the studios grew around that. The result is a market with an unusual shape: a small number of large agencies serving corporates, a long tail of five-to-twenty-person studios that formed around a product team that dissolved, and a wide layer of freelancers who assemble into a team for one project and disperse afterwards.
That shape matters when you buy. In a city with three big agencies you compare like with like. Here the same brief comes back as a fixed-price proposal from one supplier, a monthly team rate from another, and a day rate for four freelancers from a third. The numbers are not comparable until you make them comparable.
Most studios in the city work in English internally, and many have no German-speaking developers at all. This is fine until it meets a German-language stakeholder, a regulator, or a contract that has to be enforceable here. Decide early which language the contract, the documentation and the day-to-day run in, because they need not be the same and it is cheaper to name them than to discover them.
The practical version of this question: who writes the acceptance criteria, and in which language will a disagreement about them be settled?
The recurring mistake is buying capacity when the problem is a decision. Teams arrive with a brief that describes a solution and ask four studios to price it. All four price it. None of them says that the described thing solves the wrong problem, because that conversation is unpaid and adversarial in a tender.
The fix costs little: pay two or three suppliers for a short, scoped discovery, and compare what they come back with. The one that reframes the problem usefully is often worth more than the one with the lowest build number.
The cheapest proposal and the most expensive proposal usually describe different projects. Most of the work of buying well is making them describe the same one.