The licensing value-chain and how AI is changing it

AI is changing the way we build software, use software, but also the way we license it.

Generally, we see a shift from the classic seat-based SaaS model to usage-based and even to value-based models.

I define the licensing value-chain as follows:

1. Unlimited – Pay a flat rate and have no restrictions on users or usage - Pay for complete freedom

2. Seat-based - Pay for users that has access - Pay for what you can potentially do

3. Consumption-based - Prepaid credits or tokens – Pay for what you plan to do

4. Usage-based – Can be viewed as active-user-based - Pay for what you actually did

5. Completion-based - Pay only for your completed tasks

6. Value-based - Pay only for actual value delivered by usage.

   
   

As AI agents automate entire workflows, the software layer (input screens, tables, dashboards, etc.) becomes unnecessary: users state their intent, and the agent delivers. We see this now in certain domains like coding or marketing, and we'll see it more in other domains, like logistics and procurement.

Remember, average gross margins in the SaaS business typically range between 70% and 85%, meaning vendors have significant room to work with.

Interestingly enough, this is also part of the shift that we started seeing in human-labor in recent years, where in some cases the focus is on outcomes delivered and not hours invested - such as work-life balance, working from home, or unlimited PTO.