Can traditional per-user/usage-based SaaS pricing survive in the age of GenAI when different outcomes are being delivered?

2.4k viewscircle icon3 Comments
Sort by:
IT & Strategy Advisor || Digital & Enterprise Architecture Consultant in Consumer Goods4 months ago

I second the perspective by Didier Godot. Usage based SaaS pricing will need to evolve in the age of GenAI. Probably it would be a good idea to have outcome-based pricing. Hybrid pricing models with a mix of traditional licensing, usage-based and outcome-based could offer a balanced solution. The elements of traditional model can be simple to manage, providing predicability, fairness of usage-based pricing, and value alignment of outcome-based pricing.

Lightbulb on2
Chief Information Technology Officer in IT Services5 months ago

Traditional per-user or usage-based SaaS pricing will struggle in the GenAI era, where outcomes vary significantly. Models must evolve to factor in compute intensity, value-driven pricing, and hybrid structures that balance baseline costs with variable AI-driven usage. The focus will shift to cost transparency, adaptive pricing, and smarter governance to ensure sustainability while delivering AI-driven value.

Lightbulb on2
Head of Transformation in Government5 months ago

Gartner published a very good research article some 10 years ago on different ways of contracting, from time and means through to split-profit, and business value outcome is one of those deal frameworks. I found an update of the article here: https://www.gartner.com/document-reader/document/4872231?ref=solrAll&refval=454860257

But I think that the age of GenAI will definitely challenge SaaS. Remember that by 2013, we customers had completely optimised our negotiation power with the major software vendors (SAP, Oracle, Microsoft inter alia) and had options for arbitrage to a mature third party maintenance market (Rimini inter alia).

I didn't hesitate to push for a full cloud SaaS strategy back in 2013, despite knowing that I would lose all of that leverage, and also lose quite a bit of functionality. For example, 2014's Oracle HCM was definitely no PeopleSoft. Still, the benefit of SaaS outweighed the loss of leverage and arbitrage. I knew somehow the customer would be able to take it back someday, even if I could not imagine how. Still, market forces are market forces. It's the customer's euro sliding across the table, and not the other way around.

I'm relieved to see AI provide the answer. It give us many options that SaaS vendors will have to react to. In some cases per-user or usage-based models will work. In other cases we will arbitrage the market and shift compute and automation workloads over to a T&M model with a built-it-yourself approach, and dump the SaaS. This will likely cause a transition to value-sharing or outcome-based pricing that the Gartner research article argues are options.

In summary. Long live the customer! What a wonderful time we are living in.

Lightbulb on1

Content you might like

Source code24%

Model weights71%

Training data sets41%

Something else - share in comments

View Results

Established AI governance framework with defined policies and oversight52%

Currently developing governance models and risk controls67%

Relying on existing security/compliance frameworks (no AI-specific policy)26%

No formal AI governance approach in place

View Results