A SaaS pioneer confronts the rise of AI

Tien Tzuo has been around SaaS for much of his professional career, joining Salesforce as employee #11 in 1999, the year the company launched. He later founded Zuora, a SaaS subscription management and billing platform. He understands SaaS as well as anyone, so when he says SaaS companies need to face AI head-on, it pays to listen.

He doesn’t dwell on Wall Street’s 'SaaS is doomed' narrative, and he’s not averse to the shift underway. Instead, he welcomes the challenge and thinks SaaS companies have to see it the way investors do: as a direct threat to their survival.

“Every software company is facing an existential threat right now, including us, right? And you’re better off meeting it and seeing it and navigating through it,” Tzuo told FastForward. "I'm not going to take a defensive attitude that pooh-poohs the change. I'd rather go in the other direction and say, 'Let's assume that we're all under existential threat, and let's work through it.'"

When your customer can vibe code a working prototype in a week, SaaS companies have to ask what truly makes them unique, recognize where their moats are and reassess where their value lies. “There is going to be a new set of rules and a new playbook that you have to build in the age of AI,” he said.

Companies that stand still will get left behind, just as vendors like Siebel Systems, the dominant on-prem CRM vendor in the 1990s, ignored the rise of SaaS and eventually faded. "Can AI versions of an enterprise software category replace a non-AI version? It's absolutely going to happen. How could it not?" he asked.

Pushing the AI advantage

If AI makes developers more productive, executives would be foolish not to take advantage of that efficiency improvement in the same way companies like Salesforce took advantage of the internet to change the way software was delivered.

"We just believed we were that much more efficient, and it was a better way of building software and operating a company than the way we used to do it before the internet," Tzuo said. He sees a corresponding boost from AI, especially when it comes to developer productivity.

"The productivity of an engineer or an IT person inside my customers' companies has just exploded,” he said. “So the productivity of my engineers should also explode. If I don’t adopt these AI capabilities and the developer inside my customer is three times better than my developer, yeah, they should build it themselves. Why rely on me?”

The vendor advantage

But it's not as simple as saying, 'I can build it myself.' Setting aside the many hard parts of building enterprise software effectively — including security, compliance, identity, observability and integration with other software — there is also the data advantage that every mature SaaS company can point to. They all have decades of knowledge that customers can't realistically recreate on their own.

As Alex Wilhelm and I wrote last week in this space on the future of SaaS, these companies have to take advantage of that: "Sure, you could vibe code an SAP replacement, but it won’t have your data, and your LLM won’t have the same data history as the incumbent," we wrote.

Tzuo sees it similarly. Looking at his company as an example, their advantage includes everything they have learned by being a leader in subscription billing. "We need to tap into the collective expertise of being in this space,” he said, pointing to pricing, revenue recognition, quoting and other complex business processes his platform tracks.

As Alex and I pointed out, general-purpose large language models don't have that level of expertise built into a SaaS company’s platform, nor do they have the data SaaS platforms produce naturally in the course of doing business. It's up to SaaS companies to understand and harness the value of that expertise and data to maintain their relevancy to customers as AI tooling becomes increasingly capable.

Vibe coding can only get you so far, though. As we've said previously, building enterprise software is hard. AI might have solved the coding part, but there's a lot more to building a SaaS company than just developing the application. 

"At the end of the day, it's not about writing the software. It's about understanding patterns across the whole community, and building flexible tools that understand those patterns," Tzuo said.