Is SaaS Dead?!?
No. Absolutely not. Nope. (I just want to make sure my position is absolutely clear.)
“SaaS is Dead” is a take so ridiculous it sounds straight from the mouth of a first year investment banker turned VC analyst. Yet it’s been in the zeitgeist the last few months, so as the founder of a company literally called SaaSGrid it feels worth addressing.
The argument goes something like this: AI has so fundamentally changed the nature of software that everything about how SaaS currently operates— the players, the pricing, and the interface— are going away.
I’m not an AI hater. LLMs are a massive leap forward in processing unstructured data, and will lead to a 10x and eventually 100x improvement in the tools we use. But despite how fast the tech is moving, the underlying business model is changing at a more modest pace. So before we blow everything up, let’s break down a few of the arguments of the AI thought boys and see if they have any merit.
Argument 1: “If you’re not using AI, you’re getting left behind! AI native is the future! My AI CRM Cursor wrote for me this weekend is totally going to kill Salesforce!”
Maybe? Of course SaaS companies always need to worry about disruption, but I don’t think pre-AI startups and SaaS behemoths are about to get massacred. Large enterprises were actually some of the first to adopt AI (Agentforce 🤮) because LLMs can just be called by API, and don’t require the platform shift that, for example, moving from on-prem to the cloud did.
Argument 2: “Per seat pricing is dead! We’ll just charge for the work our AI does, or let them hire us as an AI employee!”
First of all, it’s called usage-based pricing. AWS and Twilio invented it like 15 years ago. Second, whoever is proposing having companies “hire” their AI as “employees” has never done an enterprise sales call in their life. Please invite me the first time you pitch that to a grizzled CFO. AI may reduce the number of seats companies need, but it’s often being replaced by even simpler platform-fee based pricing, as enterprises try to understand exactly how much value AI will bring.
Argument 3: “Companies don’t even need SaaS! They’re just going to create their own horizontal AI Agents that will do everything for them!”
LOL. I said I’m not a hater, but this take may make me change my mind. AI is good, but not THAT good (yet). When you think about a complex process like revenue recognition AI can automate many of the repetitive tasks like reading contracts for key terms. But you’ll still need a human in the loop to review exceptions, which will require a GUI, and a whole host of non-AI workflows like assignment rules and permissions. Sounds a lot like SaaS…
AI is changing SaaS, but it’s not killing it. In fact, with all the new use cases (and funding) AI is enabling, SaaS is positioned to become more important than ever. Viva la SaaS.