Is enterprise SaaS dead? No. Not exactly.

In fact, I think the enterprise application layer is best positioned to capture the gains AI has to offer. It’s just a matter of how thoughtfully and how quickly enterprise SaaS companies adapt.

We started Esper in 2018 when enterprise SaaS was all the rage, promising to “automate workflows” across every industry imaginable. If we’re being honest (and I always try to be), most companies digitized manual work but never truly made it to the automation stage.

I think the fact that so many enterprise SaaS companies proudly report DAU and MAU as core metrics is evidence of this. If automation is really the goal, shouldn’t humans be spending less time in the software?

AI now gives us the chance to finally deliver on the promise of true automation. Let’s not mince words: far fewer humans will be using enterprise SaaS products directly. But that doesn’t mean there isn’t enormous value left to be captured.

Here’s how I think about it.

Most enterprise SaaS digitized work by encoding rules.

Between 2015 and 2021, I highly doubt enterprise SaaS automated nearly as much as the pitch decks claimed. I once had a non-technical friend describe enterprise SaaS companies as “digital filing cabinets.” Honestly, that feels pretty accurate for a lot of companies I met in the twenty teens.

Most companies turned messy human processes into structured data, permissions, approvals, and audit trails. Humans still had to do the work, but on the computer instead of in manila folders.

SaaS companies captured so many “rules” about how industries operate. Fast forward to 2025, and the real value isn’t the digitized workflows. It’s the business logic those systems captured along the way.

The new workflows are higher up the stack.

Many of the original workflows enterprise SaaS companies built billion-dollar businesses on are no longer relevant. AI can handle these workflows faster and cheaper now.

But new workflows emerge around exception handling, oversight, training, compliance, and continuous improvement.

Humans move from doing the work to supervising outcomes. When humans did everything, accountability was implicit. When AI acts, accountability has to be explicit. Logs, explanations, approvals, and auditability become first-class features. This is especially true in enterprise and government contexts.

The best enterprise SaaS companies will coordinate this handoff exceptionally well.

AI wrappers won’t cut it

The enterprise SaaS companies best positioned for this shift have:

• deep domain understanding
• embedded workflows and years of captured business logic
• trusted distribution
• years of clean, structured data

They aren’t just slapping AI on top of existing products. They’re re-architecting how decisions get made inside organizations.

Let’s be real: this re-architecture will be expensive, and it needs to happen fast. There are years of technical debt to sift through and old habits that need to radically shift.

A lot of 10 year old enterprise SaaS companies are tired, and many won’t have the energy to grind through another platform shift.

That’s okay. More for the rest of us. 😉