Consolidation

When a lot of money seems to be up for grabs in a new business field, a great many companies spring up first. After a few years, only a few of them remain. Some are bought up by larger companies, others merge, and still others give up. This exact process is called consolidation: many providers become a few large ones. The market itself doesn’t disappear in the process – on the contrary, it often keeps growing. What changes is only who divides it up among themselves.

Why a hundred start-ups eventually turn into five corporations

Consolidation is not an accident but the normal case. At the beginning of a technology, nobody knows which idea will prevail. So many companies try out many variants. As soon as it becomes clear what works, it’s no longer the idea that counts, but size.

That’s due to the costs. Training an AI model, i.e. having it learn from huge amounts of data, devours hundreds of millions of euros in computing time. A small start-up can’t afford that, but a corporation can. Whoever owns the data centers, the data, and the customer relationships can spread their costs across far more users. Small providers then prefer to sell themselves before they run out of money.

For investors, the decisive question is who ends up among the survivors. A historical example: around 1900, there were hundreds of car manufacturers in the USA; a few decades later, three corporations dominated the market. The same curve later appears with search engines, social networks, and cloud providers. Whoever bets early on the wrong provider loses money even if the technology itself is successful.

Acquisition, merger, and the quiet sell-off

The most common route is the acquisition: a large company buys a small one and pays its owners with money or its own shares. Less common is the merger, in which two roughly equally sized companies combine into one. In both cases, one provider drops out of the market, and the buyer takes over the technology, customers, and staff.

In the AI industry, a third variant has established itself that is formally not an acquisition. A corporation simply hires the entire leadership team of a start-up and additionally pays for a license to its technology. The company then continues to exist on paper, but is practically empty. This is called acqui-hiring, that is, buying by hiring. The advantage for the buyer: antitrust authorities, which can review and prohibit mergers, do not view such a construction as an acquisition.

It’s important to distinguish this from a similarly sounding term. On the stock market, consolidation also refers to a quiet phase in which a price moves sideways after a strong rise. This has nothing to do with corporate mergers. Which meaning is intended follows from the context.

How to recognize an ongoing wave of consolidation

In business news, the word usually crops up once an industry has passed the peak of new company formations. Typical signals are falling valuations in funding rounds, layoffs at mid-sized providers, and several acquisitions within a few months. Phrases like “the market is consolidating” mean, in practice: the number of providers is shrinking.

As a user, you often only notice this later. A service you liked to use suddenly belongs to a corporation and is discontinued a year later. Prices rise because there are fewer alternatives. Sometimes, though, the effect is also pleasant, because the technology of a small provider migrates into a product that millions of people already use anyway.

A common misconception is that consolidation is a sign of a technology’s failure. Usually the opposite is true. What gets bought up is what has proven to be valuable. The real risk lies elsewhere: if only three providers control the most important AI models, prices, access, and rules for very many other companies depend on them. That is precisely why competition authorities in Europe and the USA are currently watching especially closely.