Startups: How to survive the SaaSpocalypse? Investors and founders have their say

When the market decides to fixate on something, the ripple effect can be overwhelming. This is what was observed in the first three months of 2026, when a suspicion that had already been circulating in recent years became the panic of the moment: will SaaS be destroyed by the rise of artificial intelligence and innovations such as agents and vibe coding?

Looking at the numbers, the excitement is more than justified. According to a recent survey by the SaaStrSaaS companies have not fallen in value. They plummeted. If in 2022 most of them still had price/earnings projection multiples of around 84x, in 2026 this number went to 22.7x, a value below traditional companies listed on the S&P 500.

At the beginning of the year, an update from Claude at Anthropicwas the trigger for an even greater frisson, resulting in the loss of more than US$300 billion in market value for companies like Salesforce, Microsoft, Workday, ServiceNowamong others. From this, a term was born: SaaSpocalypse, the final judgment in which AI will be the marker of who lives or who dies in the software market, and in which incumbents supposedly have their days numbered.

However, looking beyond the collective panic, what is really true and what is just speculation in this whole movement? In conversation with investors and founders, the report from Startups heard divergent views on how AI will still impact the SaaS market, although everyone has a certain consensus — that nothing will be the same as before.

“Software companies have always been loved by investors because they were predictable”, highlights William Cordeiro, managing partner at SaaSholicsoftware-focused fund. According to him, the sell-off of traditional companies shows that they are under pressure to demonstrate how they reinvent themselves in a rapidly changing market. “Many of them have not yet managed to show direction on how to reinvent themselves”, he adds.

Also read:

An example cited by William is that of Salesforcea CRM giant that has been moving towards AI for a few years now, especially with products like Agentforce. According to the latest results from Marc Benioff’s company, it seems that the gains are appearing, with the new product already representing US$500 million in ARR for the multinational, quadrupling in size every year.
William Cordeiro, managing partner at SaaSholic | Photo: publicity

Even so, there is a big difference compared to the traditional per-seat model, in which contracts were more predictable, unlike pricing based on usage or results, as is the case with AI. At the end of the day, according to the executive of SaaSholicthe change in billing is still a gamble, sacrificing revenue in the short term.

“A Salesforce He’s doing what he has to do to update himself. But if you have a pension fund, your fiduciary duty is to protect and multiply that wealth. Will you sell your position now or will you wait to see if the Salesforce Will you be able to update and grow like before? Or are you going to invest in those who are on the rise right now?”, he says.

Overreaction?

Asked about the so-called “death of SaaS”, the consensus from most sources is that the market reaction was somewhat exaggerated. “I don’t really buy the story that SaaS is dead nor that software is dead”, says investor Manoel Lemos, from HeyHo Ventures. “But I agree 100% that software as we know it is dead”, he adds.

For Sidney Chameh, founder of DGF Investimentosone of the most traditional Brazilian software investment funds, the market has already gone through several panics, since the time when software still ran on monolithic mainframes. “We’ve been through the mainframe, millennium bug, cloud, total software outsourcing. The need for applications has always existed, and it always adapts”, he points out.
Sidney Chameh, founding partner of DGF | Photo: publicity

And, speaking of adaptation, for founders, in a scenario governed by AI, the business renewal cycle becomes increasingly tighter. Rapha Avelar, founder of BrandLoversfelt it in your skin. Shortly after receiving a contribution of R$35 million from Kaszekmartech decided to “kill” its main product to create a new AI-based offering, Creator Ads, a decision that generated results in the medium and long term.

Asked about his vision regarding SaaSpocalypse, Rapha points out that AI has become a “stone in the side” of traditional technology businesses.

“Until recently, large software companies were used to trading at very high multiples, which implies that you are paying for growth and cash flow that is up front. So, what is happening is basically that AI is making us reevaluate: how long does a competitive advantage last in the modern world? Or is there, in fact, a competitive advantage that lasts?”, highlights Rapha.
Rapha Avellar, founder and CEO of BrandLovers | Photo: publicity

On the other hand, the founder of BrandLovers admits that even this reassessment of the market does not escape exaggerated reactions. “I think there is a bit of hysteria too. I believe that businesses that have established network effects and credibility, with responsibility for the transactions they execute, have a better chance of lasting”, he assesses.

Also read:

Innovation x credibility

In a possible new sea of ​​applications and agents that could emerge from artificial intelligence, the big ones still have something relevant in their favor: reputation and data, the so-called “system of record”.

“If you are your client’s system of record, everything happens there. You need to open the system to know what is happening with the company. There is enormous value in that. But if someone creates an agentic solution that, via APIs, reads the data in the system of record and executes the task, this guy captures much more value”, says William Cordeiro.

For Daniel Heise, from DGFthis new value capture through AI and agents represents a new order of magnitude for software. “I think the ability to scale is increasing. Companies have the opportunity to capture much more value with AI, selling not just the tool, but the complete work capacity, charging per result”, he says.

It is under this logic that the SaaS that suffered the most were those from companies like Workday, Atlassian e Asanawhich act in a transactional layer. According to sources heard by the Startupsthese applications that are less sophisticated in their use of data should be those that will suffer most from the increase in the use of agents.

“Think of a company with a high employee turnover, which opens 500 vacancies per month and has people to analyze CVs. With more AI, they will need fewer seats for this work. Furthermore, this is the type of activity that will be easily replaced by agents, so it no longer makes sense to charge leave per employee”, says Daniel Chalfon, partner at Trundle.

Daniel Chalfon | Photo: publicity

“The companies most threatened are those that do not generate a clear ROI and have easily replicable solutions. Mainly very simple, deterministic products or with a low technological barrier. These are the ones that AI can replace most easily”, adds Henrique Uehara, partner at DGF.

How to survive SaaSpocalypse?

Here we come to the initial question of this article. Amid so much uncertainty, can SaaS survive, and how? For William Cordeiro, the big question is not whether SaaS will survive, but how the players that made a name for themselves in the SaaS boom over the last 15 years will “cross the chasm”.

“The middle layer of software has become the big field of contention: the businesses that will be able to connect AI and the large ‘systems of record’. This is where it will be possible to connect multiple systems and allow agents to execute entire workflows”, points out Daniel Heise. “If the incumbents get confused, this system of record can be rebuilt from scratch by AI”, he adds.

For Guta Tolmasquim, founder and CEO of PurpleMetricsit will not be possible to “survive” the SaaS apocalypse if the software remains just a tool that opens and in which the employee still needs to perform all the work.
“If you understand that your application needs to have the person’s entire workflow and resolve their demands, leaving the human more of an editor’s role, it will be necessary to reevaluate the scope that each software will have in the customer’s life”, says Guta.

Guta Tolmasquim | Photo: Disclosure/Purple Metrics

Now, the question that remains to be answered is who will be able to capture this value. “Investors want to understand whether established companies will be able to catch up technologically and continue delivering to their customers or not. There are companies that have not been able to show this direction. And then the market punishes”, says Daniel Chalfon.

On the other hand, those that manage to “get through the surf” may see unprecedented growth. “In the long term, five to ten years, I think we are still underestimating the SaaS market. The software model that will exist in the future, and the size it will have, we may not even be able to imagine”, ponders Daniel Heise.

“People haven’t stopped listening to music because they no longer buy CDs and listen to it online. Spotify. In fact, they listen to more music. People won’t stop buying software because the way it is distributed has changed. They will buy more software”, adds William Cordeiro.

To close with the voice of experience, Sidney emphasizes that the big question is how the SaaS model can and should evolve from now on. “We don’t believe that software will be decimated. On the contrary, we continue to invest. What matters is to continue solving customers’ real pain points. This is what has sustained the sector for decades”, he concludes.

Content produced by .