Lock-in

Lock-in literally means “being locked in.” It refers to a state in which a customer is theoretically allowed to switch providers but practically can no longer do so. Switching would be too expensive, too slow, or too risky. So the customer stays, even if dissatisfied or if the price rises. An everyday example: anyone who has collected photos, contacts, and purchased apps on a smartphone from one manufacturer over the years loses part of it when switching to a different system. This very feeling of being stuck is the lock-in effect.

Why providers make it hard to leave

For companies, lock-in is not an accident but often part of the business model. A customer who cannot switch does not need to be courted with low prices. That is why entry-level offers are often cheap or free, while prices rise later. Experts call the sum that a switch costs switching costs. The higher these switching costs, the greater the provider’s market power.

For customers, this is a real risk. A company that has moved its entire accounting, customer data, and internal processes onto a single platform is dependent on that platform’s prices and decisions. If the service becomes more expensive, is discontinued, or becomes unreliable, there is no quick alternative. Competition authorities also pay close attention to this: too strong a lock-in can shut small competitors out of the market.

It is important to distinguish this from genuine customer loyalty. Someone who stays because a product is simply the best is not locked in. Lock-in only exists when the customer stays even though they would rather leave.

The mechanisms behind the tie-in

The most common mechanism is data. Anyone who has stored documents, messages, or measurement data with a provider for years would have to export and re-import all of it when switching. Some providers deliberately make this difficult by using proprietary file formats that other programs cannot read. These are then referred to as closed or proprietary formats. Export fees for large amounts of data also fall into this category.

A second mechanism is technical integration. Software is often connected to a service via interfaces—that is, defined pathways through which two programs exchange data. If these interfaces differ from provider to provider, a great deal of program code has to be rewritten when switching. For a large corporation, this can amount to several years of work.

The third mechanism is human. Employees know a system, are trained in it, and hold certifications for it. A switch therefore also means further training, mistakes during the transition period, and resistance within the team. These costs appear in no invoice, but they have a strong effect.

Lock-in with cloud and AI models

Today the term is most often encountered in connection with the cloud. Cloud means that computing power and storage are not kept in-house but rented from a provider. The three major providers—Amazon, Microsoft, and Google—offer hundreds of additional services that are only available from them. Anyone who uses these saves development time but ties themselves in more tightly.

The same effect is now becoming visible with AI. Many companies build their products on the language models of a single provider, such as those from OpenAI or Anthropic. The instructions given to the model, known as prompts, are tailored to that particular model and work worse with a different one. Even if the price per request rises, switching is still cumbersome. This is precisely why some companies deliberately opt for open models that they can run themselves and take with them.

In stock market news, lock-in is usually praise for the company, not for the customer. Analysts then speak of high customer retention and reliable revenues. As a reader, it is worth flipping the sentence around and asking who is actually the one stuck here.