ACCIÓ figures from March 2026 sketch a recognisable map: 28,122 companies in Catalonia’s digital economy, 218,598 people employed and €41.495 billion in turnover. And one detail I find the most informative of all: 46% of those companies were founded in the last decade.

In a market like that, you almost never need to invent a category. What you have to decide is how to enter one that already exists.

The three doors

There are only three. Buy a company that already does it. Build it from scratch. Or come in as a partner of someone already inside. Everything else is a variant of those three, and each buys something different.

  • Buying buys time. And, if it goes well, customers. It is the fastest route and the one hiding the most risk: what you pay for is a past, and what you inherit is a culture nobody has audited.
  • Building buys fit. It comes out exactly as you want because you made it. It takes twice as long as planned, always, and only makes sense if it adds something the market does not offer.
  • Partnering buys information. The most underrated of the three. You get in, you learn how that business actually works, and you decide afterwards whether you want more.

The question that rules out two of the three

Before looking at numbers, we ask one question that orders the decision: does this reinforce something we already have, or open a new front?

If it reinforces something existing — if it can use the technology we already share, or sell to customers who are already ours — building usually wins, because the marginal cost is low and the fit is total. If it opens a new front, building is almost always an expensive mistake: we have neither the knowledge nor the network, and both take years to acquire.

What we have learned buying

That financial due diligence is the easy part. The numbers get reviewed in three weeks and rarely surprise anyone. What always surprises is the rest: how decisions get made inside, who really runs things, what happens when somebody makes a mistake.

A company where nobody argues with the founder is a company that cannot be integrated into a group, however well it invoices. That appears in no report and shows up in the first meeting, if you know what to look for.

And what we have learned building

That the most common mistake is not technical. It is starting without having answered who is going to sell it. A product born in an engineering team that goes looking for a salesperson afterwards arrives late to its own market.

That is why none of our initiatives now moves past prototype without a named person accountable for the first sale. Not for the business plan: for the first sale.

The bias to watch

Groups that know how to build tend to build every time, including when buying would be cheaper. It is a bias of professional pride and it is very expensive. Controlling it is simple: force yourself to write down, for every decision, why the other two doors were ruled out.

In a market where almost half the companies are under ten years old, the advantage is not having the best idea. It is knowing which of the three doors applies each time.