Legacy SaaS

Many companies today no longer buy their software, but rent it instead. The program then doesn’t run on your own computer, but in a data center belonging to the provider, and you pay a monthly fee per user. This model is called Software as a Service, or SaaS for short. Well-known examples are Salesforce for customer data, Workday for human resources management, or Zendesk for customer support. Legacy SaaS is not a distinct technology, but a value judgment: it refers to providers whose product was designed before the AI boom and whose architecture is now considered outdated. The English word “legacy” means an inheritance or something handed down, and in the tech industry it’s used for anything that still runs but originates from an earlier era.

The dispute over the price per user

The term matters above all because a great deal of stock market value hinges on it. Classic rented software charges a fixed amount per employee per month. This model only works as long as companies employ many people who sit in front of a screen typing in data. It is precisely this work that AI programs can now partly take over.

This creates an uncomfortable calculation for the providers. If a customer, with the help of AI, manages to do the same work with fewer staff, they also need fewer licenses. The software provider’s revenue therefore falls exactly when its customers become more productive. Investors call this the threat to the seat model, meaning billing based on the number of workstations.

Still, one shouldn’t rush to judgment. Many of these providers sit on enormous stores of customer data that new competitors cannot access. Moreover, large corporations don’t swap out their HR software within a few months. So whether a provider is truly “legacy” only becomes clear over the course of years.

How to recognize old-fashioned construction

Technically, Legacy SaaS is often software built around forms and tables. A human operates it: clicking through menus, filling in fields, and saving. The software itself barely makes any decisions; it merely reliably manages what has been entered. For years, its value lay precisely in this reliability.

Newer approaches reverse this relationship. There, the user describes their goal in plain language, and the program carries out the necessary steps itself. Such systems are called agents, because they act independently instead of merely waiting for clicks. For an old application, such a conversion is difficult, because the entire operating logic depends on it.

There is a second problem on top of this: the program code itself. Some of these products have been under development for fifteen or twenty years and carry a lot of legacy baggage with them. Every change has to take thousands of existing customers into account, none of whom want to see anything break. A young competing product starts without these constraints and is therefore faster.

The term in quarterly results and headlines

The word is most often read in stock market reports and analyst commentary. When a stock falls despite good results, the explanation given is often that the market considers the company to be Legacy SaaS. What is meant then is not the current situation, but concern about the next five years.

The providers themselves push back against the label. Almost all of them have by now built in AI features and in some cases bill for them separately, for instance per task completed rather than per user. Whether that is enough remains an open question. A typical misconception, by the way, is equating Legacy SaaS with bad software. Many of these programs work excellently; their problem lies solely in the business model behind them.