Dealer Principle

The dealer principle is a sales strategy with two phases. In the first phase, users get something for free or almost for free. They are meant to get used to the offer and organize their everyday lives around it. In the second phase, the same thing then costs money, or at least costs more than before. The name comes from drug dealing, where the first dose is allegedly free and all further ones become expensive. In business, the principle is entirely legal and extremely widespread, especially in software and AI services.

Why free offers end up becoming expensive

The decisive point is not the low entry price, but what happens afterward. Anyone who has used a tool for six months has invested time. They have stored data there, aligned workflows around it, and trained colleagues on it. Switching then costs effort, even if the competitor is objectively just as good. Experts call this hurdle switching costs.

It is precisely these switching costs that give the provider pricing power. They can raise fees without immediately losing all customers. For some cloud services, prices have risen many times over after the introductory phase. Many customers stayed anyway, because migrating their data would have been more expensive than the increase.

For investors, the principle is therefore interesting, but also risky. A company can post losses for years and still be highly valued. The bet is: at some point, the user base gained can be turned into profit. If the bet doesn’t pay off because users jump ship at the first price increase, the money given away is lost.

The two phases in detail

In the attraction phase, the provider subsidizes its own costs. For AI services, this is easily measurable. Every response from a language model consumes computing time in a data center, and that costs money. If a chatbot delivers millions of free responses per day, the operator loses money every single time. This is deliberately calculated this way and is financed with investor money.

The switchover phase rarely takes the form of a blunt price increase. More often, the free offer gradually becomes less attractive. There are daily limits, slower responses, older model versions, or advertising. The most useful features move behind a paywall. This pattern is called freemium: a free basic version, a paid full version.

It is important to distinguish this from a pure advertising model. With classic social networks, the service remains permanently free, and payment is made in attention and data. With the dealer principle, by contrast, the later flow of money is planned from the outset. Both models can mix, but they are not the same thing. A common misconception is also that every free offer is a trick. Some companies give away software permanently because they want to set a standard with it.

Free AI chatbots as a textbook case

The principle can currently be seen most clearly with AI assistants. The well-known chatbots were initially freely accessible and usable without limit. Today there is almost everywhere a paid tier with better models and higher usage limits. Coding assistants and image generators follow exactly this same path as well.

In business news, the term usually appears in a critical context. It comes up when a provider suddenly raises prices after going public. Or when authorities examine whether a company is using dumping prices to push competitors out of the market. That is precisely the legal boundary: giving things away is allowed, deliberately squeezing out competitors with below-cost pricing can be prohibited.

In everyday life, it is worth asking a simple question before signing up. How hard would it be to get out of this again? Anyone who can export their data and does not sign any contracts is not sitting in a trap. Anyone who, on the other hand, collects everything in a single service for years, hands the provider pricing sovereignty.