Outcome-based Billing
With outcome-based billing, a customer doesn't pay for using software, but only for an actually achieved outcome. With AI providers, for example, every resolved customer case costs money, not every question asked.
Software is normally paid for based on time or volume. You pay a fixed amount per month, or you pay for each use. With outcome-based billing, it’s different: the customer only pays when a previously agreed outcome actually occurs. A provider, for example, only charges money when a customer inquiry has been fully resolved without a human needing to intervene. If the outcome doesn’t materialize, the bill stays empty. This shifts the risk from the buyer to the seller.
What the model changes for AI providers
Many companies buy AI tools and afterward don’t know whether they achieved anything. A monthly price per employee is due even if the program is barely used. This is exactly where outcome-based billing comes in. It answers the uncomfortable question about value before the customer has to ask it.
For providers, this is a strong selling point, but also a gamble. They bear the cost of the computing power immediately, while the money comes only later. If the AI performs poorly, the provider works for nothing. That’s why the model only works if the technology is reliable enough.
On the stock market, this topic is being closely watched. Investors are increasingly asking AI companies whether customers keep paying long-term or are just trying things out. Revenue from outcome-based contracts is seen as proof that a product creates real value. At the same time, it’s harder to predict than fixed subscriptions, because it fluctuates with performance.
From success criterion to invoice
The hardest part is defining the outcome. Both sides must agree on a measurable quantity. This can be a completed support case, a booked appointment, a detected fraudulent payment, or a correctly filled-out invoice. This quantity must be unambiguous, otherwise there will be disputes over every single booking later on.
After that, a measurement both parties trust is needed. Usually the software logs every process and marks whether the criterion was met. There are often additional exceptions as well: if a user abandons contact or explicitly demands a human, the case doesn’t count. At the end of the month, the invoice is generated from these counts.
A common misconception is that outcome-based billing is always cheaper. The price per outcome is often significantly above the pure operating costs, because the provider prices in its risk. Someone who uses the AI intensively and successfully may end up paying more than with a flat price. The advantage lies not in the price, but in the predictability.
Related models and where they occur
The idea isn’t new. Sales representatives have always worked on commission, and online advertising is often billed per click or per purchase. Some lawyers in the US only charge a fee if they win the case. The AI industry is applying this well-known principle to software.
In news coverage, the term usually appears under its English name, outcome-based pricing. Providers of AI assistants for customer service made it well known by charging per resolved conversation instead of per seat. Similar approaches now exist in accounting, sales, and recruiting.
This model should be distinguished from usage-based pricing. There, you pay for consumed computing power, for example per processed text token, regardless of the benefit. Outcome-based billing goes a step further and ties payment to success. Many contracts mix both: a small base fee plus an amount per outcome achieved.