Buy proven businesses instead of launching new products from scratch.

Luca Ferrari, the co-founder and chief executive of Bending Spoons, the software company that buys and runs digital businesses, was explaining why he almost never launches new products from scratch. Asked why he does not just build new products, he gave his rule in plain terms:

"Instead, we buy product market fit, companies that have already succeeded."

By product market fit, he means a business that already has customers who want what it sells. Ferrari argues that trying to create that from scratch makes little sense for a company of his size. Bending Spoons, assuming it closes its acquisition of the collaboration company Miro, has about $4 billion in annual revenue. A new launch would need to reach at least half a billion dollars in sales to move the needle.

The math, in his telling, is brutal. Copying an existing leader rarely works, he says, because the leaders Bending Spoons wants to own have moats, or durable advantages, that protect their share. If it were easy to take that share, he would not want to be in that market anyway.

That leaves radical invention, something entirely new. Ferrari puts the chance that any given startup reaches that scale at roughly one in a thousand, and says even a winner can take 10 years to get there. In 10 years, he wants to be much bigger, so waiting a decade for a long shot holds no appeal.

He also argues that startup success stories hide the cost. Talented founders and early employees work for almost no money, chasing a dream with what he calls an awful expected value. People see the few winners, not the huge amount of failed time and effort behind them. The exception, he says, might be a company like Apple, which can sell new products to an existing base of loyal customers through a built-in distribution channel. With a diversified set of properties, he argues, "launching yet another completely different one" would offer poor expected returns, because it does not have a single base of loyal customers to sell into.

So Ferrari says he would rather pay what looks like a high price for a proven business, and then earn it back by operating it better. By his estimate, probably 80% of the companies Bending Spoons bought were growing when acquired, and kept growing afterward, a point he made to counter the claim that it buys only dying companies. But he would happily buy shrinking ones, too, if the return was right. If someone offered him shrinking companies forever at a 65% levered internal rate of return, while investing $5 billion to $7 billion a year for 20 years, he said he would sign in a heartbeat.