SaaS Pricing

In the past, you bought software once on a CD and then owned it. Today it usually runs on the provider’s servers, and you rent it on a monthly basis. This type of rental software is called SaaS, short for “Software as a Service.” SaaS pricing is the question of how a provider charges for this rental. Does the customer pay per person, per use, or a flat fee? This decision sounds like bookkeeping, but it shapes the entire business.

Why the price determines the company’s fate

With rental software, a customer doesn’t pay once but every month. That’s why such companies calculate with recurring annual revenue, often abbreviated as ARR. Investors look almost exclusively at this number. A pricing model that grows along with the customer’s success lets ARR rise even without new customers.

Conversely, a bad pricing model can hold back a good product. If the entry point is too expensive, no one tries the software. If it’s too cheap, the provider earns nothing from customers who consume a lot of computing power. This is especially dangerous for AI products, because every request to a large language model costs real money.

A common misconception is that the price should be based on production costs. With software, an additional copy is almost free. That’s why the price is based on the value to the customer, not on the effort involved. A piece of software that saves a company 100,000 euros in work time may cost noticeably more than a note-taking app.

The common billing models

The most widespread is the price per user per month. A team of 30 people then pays three times as much as a team of 10. This is easy to understand and easy to plan for. The downside: companies save money by sharing accounts.

The alternative is usage-based billing. You pay for what is actually consumed – such as amounts of data stored, emails sent, or text units processed. Providers of AI interfaces bill per token this way, meaning per small text unit into which a text is broken down. The advantage is fairness, the disadvantage is uncertainty: no one knows at the start of the month what the bill will look like at the end.

On top of that come two building blocks that appear almost everywhere. First, tiering into packages such as Basic, Pro, and Enterprise, with the more expensive tiers including additional features. Second, the freemium model: a stripped-down version is free so that as many people as possible try it out. Money is only made once a small portion of users switch to a paid tier. Typical conversion rates here are only two to five percent.

SaaS pricing in everyday life and in quarterly results

You encounter SaaS pricing more often than you might think. A music subscription, cloud storage for photos, or the paid version of a chatbot all follow exactly this logic. Even the school software used for timetables and grades is usually paid for by the school per person per year.

The term appears in business news when companies present their figures. There is talk of price increases, or of a provider selling AI features as a more expensive add-on package. Microsoft set the example with its AI assistant for office programs, charging an additional fee per user for it.

It’s important to distinguish this from classic license sales. There, you pay a large sum once and own a version permanently. With SaaS, you pay a small amount continuously and lose access as soon as you stop. That is precisely why analysts watch the cancellation rate, known in industry jargon as churn, as closely as revenue itself.