SaaS is Far From Dead

AI has undeniably become the cool kid in town much, like Software as a Service (SaaS) startups were a few years ago. But today, AI enthusiasts are boldly declaring that SaaS is dead!

In recent online discussions, a debate has surfaced around the viability of SaaS in an era increasingly dominated by AI and economic uncertainty. While some industry insiders have proclaimed the “end of software”, a closer examination reveals that SaaS is not dead but is evolving.

It all started with the Salesforce saga and a report by Upekkha that SaaS startups would struggle in 2024. With its growth slowly declining, Salesforce, one of the most profitable software companies, doesn’t excite the investors anymore. Salesforce missing its earnings guidance for the first time in 73 quarters only added fuel to fire. Some even take the case of Workday, which saw its shares sink more than 15% last week.

On the other hand, Zoho, the bootstrapped unicorn SaaS giant, remains unaffected. The company’s founder Sridhar Vembu, recently took a dig at Salesforce for its disappointing quarterly numbers. Zoho continues to thrive and has not faced any layoffs, claiming to even have zero debt in the balance sheet.

Is AI to be blamed?

Thiyagarajan Maruthavanan, partner at Upekkha, put it rather simply – “Is SaaS dead? Depends on who you are asking.” He explained that investors will put their first fist on the table and say yes as that is where their biggest incentive is – in pointing out the outliers and calling that a trend. For founders in SaaS companies, it is all about adapting to the market shift and assessing the signals in the trend.

This points to the fact that SaaS is adapting and not dying; and is far from being obsolete. The current landscape is driving SaaS companies to innovate and rethink their business models. The notion that enterprises can simply build their own AI solutions is proving to be overly optimistic.

While some companies are attempting to develop in-house AI capabilities, many are finding this approach to be costly and complex. As a result, there remains a substantial market for specialised SaaS applications that offer AI-enhanced functionalities such as Snowflake, Databricks, and many others.

AI-as-a-service (AI-aaS) is the next evolution

AI capabilities are embedded into SaaS platforms, providing advanced analytics, automation, and intelligent decision-making tools. “Legacy and new SaaS companies will truly become AI-first, abstracting away the complexity of deploying LLMs,” said Matt Turck in his latest blog post.

The question that remains is what would happen to the current SaaS unicorns while this transition happens. Obviously there are a lot of extreme views about the industry, but none grounded in anything except hype.

“When I hear about the hard tech renaissance because SaaS is tough, I pause to reflect how absurd these statements are,” said Alex Iskold, managing partner at 2048 Ventures. “The undercurrent of these conversations is moats and long-term defensibility. SaaS in a crowded space is way less defensible these days – disruption is hard!” he added.

Most companies are investing in AI without any meaningful revenue out of it. Even though it is becoming easier for companies to build software using AI, this would only be true for some companies, not all. Hemant Mohapatra from Lightspeed India explains that we are now moving into the AI-cycle of SaaS, shifting away from the internet-driven 1st cycle.

SaaS Moats Might be Dead

Matteo Franceschetti raises a pertinent question: “Given the speed at which you can build SaaS products and services, will current SaaS companies keep expanding and dominate because they already have distribution or is there space for new SaaS companies starting from zero?”

The dynamic nature of the SaaS market presents opportunities for both established companies and new entrants. Turck replies saying that there is room for “intelligence first” SaaS startups with AI at the centre.

“Merely having data isn’t enough,” said Aaron Erickson. “SaaS with mediocre UI also dies if 90% of its use cases can be replaced via chatbot,” he added, citing the example of HR apps. “The world has a lot of very bad SaaS!”

This is perfectly illustrated by Will Gendron, who is currently bootstrapping a SaaS product without any coding experience. He said that the company will be profitable this year because “I’m the only developer” and that it was impossible without AI.

While another user on X said that switching to AI is extremely cost intensive and still would take years to be on par with AI when it comes to providing features.

“Software is not dead. But it certainly isn’t as easy as it once was,” said one of the industry leaders, noting that it is the worst sentiment ever. The only thing that SaaS now needs to adopt is AI as a service and move towards a transaction-based model instead of subscription-based, or a combination of both, as software with LLM has marginal costs.

The era of SaaS 3.0 is here and AI is at the centre of it.

The post SaaS is Far From Dead appeared first on AIM.