Software Business Models In The Age of AI

I've written before about the idea of "RaaS" (Results as a Service) whereby instead of paying a subscription to access software, you pay for the actual results that the software delivers.

I still think that's an interesting model, and we see that model out in the real world, in a limited set of use cases.

But, I think there's another, intermediate business model that I think we will also start to see: Work as a Service. Pay for the work the software does.

In many use cases, I think this might make more sense than RaaS.

For RaaS to be effective, there has to be some objective, measurable result that the customer is paying for. A common example is a customer support ticket resolution. It makes sense in that case, because we have reasonable ways to measure whether AI "resolved" a particular support case to a sufficient level of customer satisfaction -- or not. We also understand the value for said result.

But, this model doesn't work as well when the result is subjective.

Let's say you hire A.I. to generate a set of illustrations for a book. Who's to say when the work is "done"? How many "iterations" are reasonable? 10? 100? 1000? Who's to say if the images are "good"? What if the images aren't useful, but it's not completely the A.I.'s fault -- but responsibility is shared by the human?

You get the idea.

RaaS makes sense when the result can be rationally and reasonably measured.

Otherwise, a good intermediate step might be WaaS. This is how life works for people now. Your creative director might hire a junior designer to help with visuals for your sales materials. You probably just pay them a salary, have them do work, and expect that as they learn the business, they'll get better over time.

You don't pay them for the number of images they deliver. That doesn't make sense.

I envision a world where people and AI agents are collaborating to work better together as a hybrid team.

We'll work out the details as to how the value created through A.I. is shared between the software providers and the customer. I think in many cases, SaaS will still make sense, when the value the AI delivers is as a tool for humans to do better, more efficient work.

SaaS is about empowerment and augmentation, not replacement.

I don't think that model is going away anytime soon. We'll see a mix of all three, based on the use case.