SaaS Isn’t Dead as Services-as-Software Takes Shape

The long-predicted “SaaS apocalypse” looks increasingly overblown. In a recent interview with CNBC, venture capitalist Orlando Bravo argued that AI is not a threat to software companies but “an enormous tailwind.”

“AI is an enormous tailwind for software companies.”

The case for software’s resilience is showing up in the numbers. In May, Salesforce reported $11.1 billion in quarterly revenue, up 13% year over year. The company also moved to strengthen its software business with a $3.6 billion acquisition of customer service software company Fin, formerly known as Intercom. Even IBM, despite what the source describes as a brutal quarter overall, posted 5% growth in its software business.

According to an analysis by Saurabh Gupta and Phil Fersht of HFS, the real shift is not the death of software-as-a-service but its evolution into a hybrid model they call services-as-software. Their argument is straightforward: the idea that AI will quickly replace SaaS and IT service providers depends on enterprises being able to deploy AI at scale, and most still cannot.

Recommended reading

Chinese AI models aren’t a security threat, Arcee argues

What services-as-software means

Gupta and Fersht say enterprises remain held back by technology, data, process, and talent debt. Until those issues are fixed, AI is likely to stay stuck in pilots and proofs of concept instead of reshaping operations at scale.

That creates room for a blended model. In their view, services firms are becoming software businesses, while software vendors are moving deeper into implementation and business transformation. The result is convergence around outcome-based economics rather than pure software licensing or traditional services billing.

They point to IBM as an example of that shift, describing it as a company that has become “a software and AI business that happened to own a consulting arm” rather than a consulting company trying to sell AI.

Five ways enterprises should adapt

Gupta and Fersht recommend five practical moves:

• Treat enterprise debt as an up-front business issue and fund it with the same discipline as capital investments.
• Focus on business outcomes, not AI pilots that fail to show commercial impact within 90 days.
• Buy outcomes instead of effort, measuring suppliers by P&L impact rather than licenses, tokens, or full-time equivalents.
• Align AI and services partners so enterprises do not pay for disconnects between vendors.
• If you’re an AI-native firm, prove value in deployments before pushing for scale.

Their bottom line is that software and services companies remain deeply entrenched. These firms, they write, generate tens of billions of dollars in annual revenue, serve major enterprises under multi-year contracts, grow at roughly 5% a year, and post 15% to 20% operating margins.