S1E11: Agent Talk #11 - Nick Mehta (Gainsight) - Most SaaS companies will die in the AI revolution

Nick Mehta, Gainsight’s CEO, didn't mince words in our latest conversation. He says the entire enterprise software model we've built over the past 20 years is fundamentally broken.

What doesn’t work about SaaS anymore?

The model that made Salesforce, Workday, and countless others into tech giants? According to Nick, it doesn't work anymore:

"We bring in data from disparate systems, we use some kind of ETL to combine it, we have rules that say if this, then that, we have workflow for humans to take action, and then reports for managers. I argue that model's totally broken."

Why? Because:

* Users have to take manual action

* Ops teams define rigid rules that can't adapt

* Leaders review reports on low-quality data that users entered

* Nothing happens autonomously

Nick framed it in a way that we love, because we feel the same way.

"The whole SaaS industry, and maybe all of humanity, is going from 0 to 1 again."

The Three Phases of AI Transformation

What's replacing the old model? Nick outlined three evolutionary phases for how AI transforms SaaS:

* AI as an Assistant: Like an executive assistant for employees – helping write emails, summarizing meetings, prompting within existing workflows

* AI as an Analyst: A layer that analyzes all communication in real-time, reveals patterns, and gives leaders visibility they never had before

* AI as an Agent: Fully automated workflows that take action without human intervention

Most companies are stuck at Phase 1 because they're afraid to cannibalize their existing seat-based business models.

Classic innovator’s dilemma!

The vibe revenue problem few are talking about

Perhaps the most eye-opening part of our conversation was when Nick revealed what he's seeing in the market:

"If you're an AI company, you think you have customers. They think they're doing a trial."

He explained the disconnect happening across the industry: AI startups claim big enterprise deployments, but when you talk to those same customers, they describe it as "just a proof of concept" or "playing around with it."

Even more shocking: "Some of them have massive churn... like 70% churn!"

We call this vibe revenue!

The innovator's dilemma in action

We asked Nick how established companies like Gainsight navigate this transition, and his answer referenced Clayton Christensen's famous book:

"Almost every time there's a discontinuous change in business, most companies don't make it because they're not willing to say, 'our old stuff doesn't make sense anymore.'"

He's taking his own medicine too. Rather than spreading his new CTO across all of Gainsight, he's focused him exclusively on their new agentic initiatives - acknowledging they need separate focus to avoid being hampered by legacy thinking.

Will Your Headcount Survive?

When we pushed Nick on what happens to customer success teams as these agents take over, he was surprisingly candid:

"Short-term, you're replacing atomic tasks, which is easier to digest... Long-term, yeah, you probably end up replacing seats."

"This is going to be awesome for the customer."

Nick believes we'll see dramatic consolidation of customer-facing roles. No more need for separate CSMs, TAMs, support people, and account managers - the agentic layer will allow fewer humans to cover more ground.

The surprising answer to AI deployment failures

Nick's final insight shocked us. When asked why so many AI deployments fail despite the amazing technology, his answer had nothing to do with models or features: "Human change is really hard. We're so inertial."

His prediction: companies deploying AI in enterprises will need consulting practices to help re-engineer workflows:

"If you're an enterprise company deploying one of these AI tools, if you don't have somebody, some advice and consulting resources, you're going to buy the AI tool and you're going to churn it - guaranteed."

Your takeaway from this episode should be that most SaaS companies won't make this transition. They'll keep trying to fit AI into their existing paradigms rather than fundamentally rethinking what software even is.

If you want to succeed, do what Nick says: Understand that stuff don’t make sense anymore, and be honest with your employees and customers about it.

What do you think? Is your SaaS company ready to make this jump? Leave a comment below 👇

Watch the full podcast here or wherever you listen to podcasts:

👉 https://podcasts.apple.com/us/podcast/agent-talk-podcast/id1792748956?i=1000705561812

👉 https://open.spotify.com/episode/2hd0evHQIByeAbIAT5sIYN?si=MCL2qytsT1GExCCX-oIb5g

👉 https://www.youtube.com/watch?v=HgePQ9sgSXg

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit agenttalk.substack.com

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.