On October 14, 2025, we hosted a live webinar titled “Navigating SaaS Exits in the AI era”. The webinar was presented by Marcin Majewski, Managing Director and Filip Drazdou, M&A Director.
You can now watch the full webinar replay below. If you would like to download the presentation material used during the session, you can easily do so by clicking the download report button on the left (if you are using a computer) or by scrolling at the very end (if you’re using a phone).
Below is the full transcript of the discussion, edited or paraphrased for clarity, flow, and briefness.
Marcin kicked off the session by highlighting how quickly the SaaS landscape is shifting, especially with the rise of AI.
The session’s agenda focused on:
- The impact of AI on SaaS valuations
- Which SaaS segments are most at risk
- How to position your company as AI-ready
- What acquirers are looking for in today’s market
Filip followed by putting things in context: since the release of ChatGPT in late 2022, AI infrastructure stocks (semiconductors, cloud platforms) have surged, up over 360%. Meanwhile, SaaS companies have remained flat, underperforming both AI leaders and even the NASDAQ.
This divergence in performance, Filip explained, sets the stage for the rest of the discussion why SaaS has lagged, what founders need to know now, and how they can adapt or exit strategically in this new era.
Marcin Majewski:
Today, when it comes to AI valuations, I’m probably more bullish than bearish. It’s difficult to value businesses that have the potential to grow asymptotically, but that’s exactly the opportunity we’re seeing with AI.
Marcin Majewski:
By the way, we have a quick poll—we want to make this webinar a bit more interactive. So feel free to vote: Are you bullish on SaaS going forward?
Marcin Majewski:
My view on SaaS varies, honestly. Even within my own portfolio, I see a range of outcomes.
Marcin Majewski:
Most SaaS businesses today don’t yet have a clear answer for how to adapt to this new AI-driven environment. Many are still clinging to what worked in the past—but those strategies aren’t working anymore, and they won’t work going forward.
Marcin Majewski:
Transforming a business that’s been successful for a decade is a daunting task. Most companies either don’t do it or wait until it’s too late. And we’re seeing this reality reflected in their share prices.
Marcin Majewski:
So I don’t expect much to change in the short term. At best, I think stock prices will remain stagnant.
Marcin Majewski:
What about you, Filip?
Filip Drazdou:
I’m bullish today, and I want to present the case for that. Looking at the poll results—it’s close, almost a 50-50 split—but bullish is ahead by a small margin: 53% to 47%.
If I were to make a bullish case for SaaS especially when looking at the chart showing how AI infrastructure winners are dramatically outperforming I’d actually start by making a bearish case for AI.
Right now, there’s a lot of excitement around how cheap and easy it has become to build software. With tools like VibeCode, Loveable, and GitHub Copilot, creating new SaaS products is faster than ever. That’s part of why some people think SaaS companies are becoming less valuable.
But I think we’re still in an experimental phase. Many are trying to build with AI, but within a year or two, those experiments will largely end, and the bubble could burst. In fact, many professionals and economists now agree we’re in a bit of an AI bubble.
When that happens when companies realize many of these AI projects aren’t delivering—some will return to tried-and-tested SaaS solutions. A report from a few months ago said 80% of AI experiments fail to bring value. So I see potential for a SaaS resurgence once the dust settles.
Marcin Majewski
That’s a very interesting take. I like it, and I think it resonates with what we’ve seen as well. Doing something quickly with AI is easy. But doing something that actually delivers value—something robust and resilient that stands the test of time—is much harder.
Filip Drazdou
My argument is probably more long-term. I do agree that right now, all the attention is on AI. SaaS is not just losing attention—it’s likely losing capital too. We can dive deeper into why SaaS companies have been underperforming.
Marcin Majewski
And by the way, if you have any questions as we go through the deck, please feel free to type them into the Q&A box. We’ll address them after the main presentation.
Let’s talk about how we got here. Things were going so well. Back in 2011, Marc Andreessen said, “Software is eating the world,” and for a long time, it really was—right up until the launch of ChatGPT.
But what we’ve come to realize is that someone saw this shift coming much earlier. In 2017, Jensen Huang said, “AI is going to eat software.” That never would have occurred to me back then. I don’t know about you, Filip, but I didn’t anticipate that at all. Clearly, Huang was much closer to the action and had a better read on the direction things were heading. It’s pretty impressive, looking back.
So the question is—will AI eat software? Our hope is that not all software is on the menu. While much of it is at risk, we believe some categories are more resilient.
There are a few things that can help software withstand this wave of disruption. First, competitive moats—especially when tied to proprietary data. If your business depends on data that no one else has access to, or operates under strong regulatory protection, it’s in a much safer position. Think about industries like healthcare or energy infrastructure—those kinds of SaaS businesses are relatively insulated for now.
Marcin Majewski
Looking ahead five to ten years, we can already see certain characteristics that make some SaaS businesses more resilient in the face of AI disruption.
One of them is customer stickiness. If a business has low churn and high retention, it’s less likely that AI will significantly affect it. In fact, adding AI features to an already sticky product can make it even more difficult for customers to switch away.
Another key trait is mission-critical and technically complex software. AI tools like pipe-coding can easily disrupt simple or lightweight applications, which is already happening to smaller SaaS tools. For example, take Calendly. We noticed recently that Google added a very basic appointment scheduling feature directly into Google Calendar. It’s still immature, but it shows how incumbents can disrupt entire product categories with small incremental changes.
However, if your software is technically deep, serves a specific niche, and is not easily replicable, it’s in a more comfortable position. It’s still a relatively small group of companies, but there are quite a few that can operate with some confidence. That said, they can’t afford to be complacent. They still need to keep innovating and embrace AI. They’re just starting from a more defensible position compared to others.
Unfortunately, everyone else is at risk. But we’re going to talk more about how to handle that in the rest of the session.
Marcin Majewski
Let’s move on to ‘How We Got Here: Part 2’.
One thing we’ve observed—and which many SaaS companies either missed or didn’t pay attention to—was the rise of the AI funding wave. It likely began around 2015, with a peak in 2021.
Then there was a temporary slowdown as people tried to evaluate what was happening, but since then, things have accelerated to record highs in 2025. This marks a major shift in sentiment, not just among investors but among customers as well.
Over time, customers gradually began experimenting with AI. Initially, it wasn’t seen as a major change, and many companies benefitted from adding AI features to their offerings.
But that phase was an early signal—a clear sign that a broader shift was coming.
So what happened next?
Marcin Majewski
The next major shift we saw was the rise of AI agents, around 2023–2024. Many SaaS functionalities that users once relied on vendors for can now be handled by AI agents. This is pushing companies to rethink their buy vs. build decisions.
We’re already seeing examples—Klarna, for instance, is building more in-house using AI, cutting down their reliance on third-party tools. It’s cheaper and more efficient, and it reshapes the tech stack strategy for many companies.
Now, we’re entering what I’d call the third phase: SaaS consolidation. This trend will likely play out over the next five years. Today, even small companies might use 50 to 100 different SaaS tools, but that’s not sustainable.
Customers increasingly prefer a single, comprehensive platform over dozens of disconnected tools. And because it’s cheaper and easier to build and expand features, we’re heading toward a winner-takes-all market. Larger vendors are aggressively positioning to capture as much IT and cloud spend as possible.
This is great news for those involved in M&A. We’re already seeing a rise in large, strategic deals—Singapore, for example, has become active again in tech M&A.
While we’re bearish on many individual SaaS businesses, we’re bullish on the overall M&A activity. The companies that survive this phase will come out stronger and more resilient.
Filip Drazdou
I’d say we’re currently in the middle of Phase 2. The sentiment shift is already massive, with billion-dollar announcements happening weekly—whether from OpenAI, NVIDIA, Broadcom, Oracle, and others.
AI agents are already starting to replace traditional SaaS tools. Many companies are still experimenting, but over the next one to two years, we’ll likely see clear examples of tools being fully replaced—especially in smaller businesses and less mission-critical areas.
Filip Drazdou
We’re going to dive deeper into mission-critical use cases shortly.
Marcin Majewski
Looks like the poll results are in—60% bullish on SaaS. That’s encouraging. While some of it might be wishful thinking, I also hope it reflects a generally optimistic mindset among our audience. That’s a winning strategy, no matter the market conditions.
Filip Drazdou
And the lack of consensus is what makes a market. Some people will be buying, others will be selling—which aligns with the current wave of SaaS consolidation.
Marcin Majewski
When we talk to serial SaaS acquirers, one issue we’ve noticed is they often don’t have a clear exit strategy. They’ve made good returns and have strong operational processes, especially in turnarounds. But few are asking, “What’s next?” after acquiring multiple companies.
That said, there’s still plenty of opportunity. If you’re smart and strategic, there’s money to be made in SaaS—especially through M&A.
Filip Drazdou
Let’s move on to growth rates. What do you see there, Marcin?
Marcin Majewski
This is one reason I’m not very bullish. Year-on-year growth rates have declined significantly since 2015. SaaS is now a mature, or near-mature, industry—10% growth is still solid, but it’s no longer considered high-growth or particularly exciting.
Looking ahead, we expect growth to fall below 10%, which doesn’t help the investment case. And while valuations remain strong, companies are still paying a lot for each unit of growth in EV-to-revenue terms.
We anticipate a market correction, especially if the AI bubble bursts. The current AI hype and circular deals—like those involving OpenAI—aren’t sustainable. When that optimism fades, it could drag down the broader tech sector, including SaaS. Until then, we expect SaaS to remain flat.
Filip Drazdou
For companies growing under 10%, today’s valuations are still fairly generous. But in private markets, you wouldn’t see 6x revenue for that growth profile. We’re increasingly seeing deals close at lower multiples.
Importantly, both buyers and sellers are adjusting expectations. Profitability and cash flow now matter more, and sellers are starting to manage their businesses accordingly.
Marcin Majewski
Which essentially means low growth. And we might return to using good old EV-to-EBITDA multiples again. For smaller SaaS businesses, that metric is already more relevant. But for larger players, revenue multiples still dominate.
Filip Drazdou
Now let’s look at which SaaS segments are most at risk of AI disruption.
One useful framework comes from Leonis Capital. They analyzed which types of AI-native or traditional SaaS applications are most likely to be replaced by LLMs like ChatGPT or Gemini.
Their findings: the most vulnerable are technically simple, horizontal tools. These often have large addressable markets and user bases—and that makes them prime targets for disruption.
Filip Drazdou
The areas most at risk from AI disruption are horizontal, technically simple tools—especially where big tech is likely to step in. Take text generation or marketing copy tools, for example. Jasper was a pioneer here, but with ChatGPT’s release, it quickly lost relevance. The same applies to AI-based translation tools—LLMs now handle that just as well.
In contrast, highly specialized, vertical applications—like healthcare analytics or industrial software—are far less likely to be targeted by OpenAI or Google. These markets are too niche and technically complex to justify the investment. So if you’re building in that space, the risk of disruption is lower.
Although that framework was originally designed for AI-native tools, it applies just as well to traditional SaaS. Anything technically simple and broad is at high risk. Not only will it face competition from AI, but also pricing pressure from customers expecting cheaper solutions.
One other important filter is whether the process needs to be 100% accurate. In areas like marketing or customer service, a 95% success rate is acceptable. Errors don’t matter much if you’re writing emails or chatbot responses—and the cost savings are huge. That makes those segments ripe for disruption.
But in areas where accuracy is critical—like payroll, accounting, or safety—there’s much less tolerance for error. You can’t have 95% of employees paid correctly or only most of a balance sheet accurate. These segments will likely resist AI integration for a long time.
Marcin Majewski
We’re also seeing much more detailed risk assessments in SaaS acquisitions. The due diligence process has shifted. Buyers are more focused on downside risk than they were a few years ago.
I hope this gives you a helpful framework to position your business, and decide where to invest your time and energy.
We also wanted to focus on something practical: how founders can actively benefit from AI right now. Embracing AI can boost your productivity, help your team ship faster, and improve customer satisfaction.
Speaking from experience, we’ve recently started experimenting with building our own tech stack. As longtime consumers of SaaS and AI tools, this shift felt natural. And honestly, it’s something most companies will end up doing.
To stay relevant, it’s essential to understand what’s moving the needle in SaaS right now. The most impactful use case we’re seeing for AI—especially LLMs—is product development. Specifically, writing code.
There are several standout tools in this space: Cursor, GitHub Copilot, Avable, and Replit. They’re extremely powerful. Some companies are still hesitant to adopt them, but I think that’s a mistake.
My view is simple—if AI can’t quite do something today, it probably will within 6 to 12 months. The sooner you integrate it into your development workflow, the better positioned you’ll be.
Marcin Majewski
To get the most out of AI, companies need to retrain their teams and help them build new habits. One area where we’re already seeing big productivity gains is in operations and finance. We use tools like NA10, which automate many manual, Excel-based processes—often saving significant time and cost.
Sales and marketing is another area being transformed. With so much available data and AI’s capabilities, companies can now deeply understand customer behavior and micro-target their messaging. What was once impossible for a human team to manage is now fully scalable with AI.
AI is also proving valuable in leadership and people management. It helps us stay informed, improve internal communication, and understand team dynamics more effectively. It’s becoming essential for both day-to-day management and strategic thinking.
Filip Drazdou
Of course, there are downsides. In product development, for instance, senior engineers will caution against blindly using AI-generated code—it still needs review. But as a tool to boost development speed and team productivity, it’s absolutely worthwhile.
And there’s a risk in not adopting it. If your competitors are moving faster, building more efficiently, and selling smarter with AI, you’ll fall behind. The greater risk is staying still.
Marcin Majewski
There’s also the long-term risk of becoming too dependent on AI. That’s inevitable to some degree—but it needs to be balanced. As more companies go fully remote and rely on automation, human connection becomes even more critical.
We’re big believers in reconnecting—with teams, with customers, and through real conversations. People are starting to see through generic, AI-generated messages. Personally, I appreciate it when someone takes the time to write a genuine email. So yes, embrace AI—but don’t forget the human side.
Filip Drazdou
So, what should founders be doing right now?
Marcin Majewski
We’ve broken it down into offensive and defensive moves.
Marcin Majewski
Founders need to make both defensive and offensive moves. Defensively, it’s about recognizing that the world is changing whether you act or not. Search traffic is declining, and interfaces like ChatGPT are becoming the primary way people interact with digital tools. You need to optimize your product and positioning for this new reality—where traditional websites may eventually be replaced by AI agents handling tasks like ticket booking or software purchasing.
Data is more valuable than ever. Companies must protect and leverage the user data they have—it’s the key differentiator against competitors who may have the same tools but not the same insights. And every team should be educated in AI. At minimum, your staff must be AI-literate.
Offensively, it’s about moving beyond just adapting—you want to become an AI-native company. No one has a perfect playbook yet, but it starts with rethinking everything. If you could rebuild your business from scratch, without legacy code or outdated customer habits, how would it look?
Think about how you’d structure your data flows, what problem you’re solving, and how you’d solve it using modern tools. That might mean disrupting your current business—but the few companies that do this well won’t just survive, they’ll thrive.
Filip Drazdou
This shift is also changing how buyers evaluate SaaS companies. In the past, metrics like recurring revenue, customer acquisition cost, and LTV made SaaS easy to forecast. But in the AI era, the first question buyers ask is whether your company can still scale under these new dynamics.
Many AI-native tools use usage-based or value-based pricing, rather than traditional seat-based models. This could become a problem for SaaS companies relying on user-based pricing—especially if AI reduces headcount in areas like customer support. A declining user base means a shrinking revenue base unless your model shifts toward value.
Business size still matters—larger companies typically get higher valuations—but we’re also seeing billion-dollar acquisitions of tiny teams. Ten- or twenty-person AI startups are getting picked up by major players. That’s both exciting and threatening: a small, fast-moving team could become your biggest competitor overnight.
Marcin Majewski
We’re also seeing tiny AI-native startups trying to sell too early. That’s not always a winning move either. Timing is critical—there’s a balance to be found between scaling too slowly and exiting too soon.
Filip Drazdou
There’s been a lot of hype around small AI-native startups being acquired, but those are exceptions, not the norm. Stories like an eight-person Israeli company getting bought by Wix are outliers, yet they heavily influence how SaaS founders think. In reality, only a few small businesses have landed those big exits.
Another factor investors have always loved about SaaS is the moat—high stickiness, deep integrations, and high switching costs. In the AI era, though, the question is: can this product be replicated by AI at a fraction of the cost? Due diligence now includes testing how close AI tools like Lavable or Replit can get to recreating the product. The further away they are, the better the business’s defensibility.
We’re also seeing big changes in how unit economics and revenue predictability are viewed. Traditional KPIs like churn, LTV, and CAC used to make future revenue easy to forecast. But now, with AI-native competitors rapidly emerging, those metrics are more volatile. A product may be fine today but become obsolete if a better, cheaper AI alternative launches in two years.
The customer acquisition landscape is also shifting. Instead of relying on Google, buyers now consult LLMs to make software decisions. Directories like Capterra and G2 are less influential. Founders must now ensure their products appear in AI-driven interfaces and adapt their go-to-market strategy accordingly.
All of this adds up to more risk. SaaS used to be a stable, predictable business model. Now, disruptive changes from AI can happen quickly, which will likely put downward pressure on valuations.
Filip Drazdou
That’s why we created our SaaS Exit Readiness Dashboard. It evaluates your business across traditional dimensions—but now also includes an AI readiness assessment. We’ll share the deck after this webinar so you can download the tool and see how you score.
Marcin Majewski
It’s hard to sum up something evolving so quickly. SaaS companies have taken a hit from AI, and the entire market is trying to find its new equilibrium. Investors aren’t treating SaaS as the safe bet it once was.
But we don’t think SaaS is done. The stronger companies—the ones with vision—will adapt and emerge more resilient.
While it may feel like a revolution, it’s more like an evolution. We’ve moved from on-prem to cloud, to SaaS, and now we’re transitioning into the AI era. Eventually, AI and SaaS will merge into one unified category. And honestly, we don’t think AI can thrive without SaaS.
Marcin Majewski
We don’t think SaaS will be viable without AI, just as we don’t think AI can thrive without SaaS. That said, there’s no clear roadmap—the pace of change is too fast. Founders and investors need to stay alert, read the signals, and decide whether to adapt or exit.
Selling a business can be the easiest way to navigate this uncertainty. It’s similar to the shift from on-premise to SaaS: the winners were often those who sold before the transition hit full force. Those who held on too long saw valuations drop. Today, we’re in that in-between phase again—valuations are depressed, and they may stay that way for a while.
Eventually, we’ll see a new generation of AI-infused SaaS companies emerge stronger. But in the short term, sellers and investors need a clear strategy for managing AI-related risks.
Filip Drazdou
That’s a great analogy. On-premise software didn’t vanish overnight, and it still exists. But these transitions—whether it’s the internet, SaaS, or now AI—always take time. This one will likely unfold even faster than previous waves.
Marcin Majewski
Exactly. This tech shift is moving at a breakneck pace, and we hope today’s session has helped you better navigate it. We’re happy to answer any follow-up questions, both during the Q&A and afterward.
If anyone would like a free business valuation and AI risk assessment, feel free to reach out—we’re happy to help.
Now, let’s jump into the Q&A.
Q: How do you define an AI infrastructure winner?
Filip Drazdou
We define AI infrastructure winners as companies building the core hardware and cloud infrastructure for AI. This includes semiconductor players like NVIDIA, and hyperscale cloud providers like AWS (Amazon), Google Cloud, and Microsoft Azure. We also consider Meta and Oracle, both of which are growing their AI capabilities.
Q: Have SaaS multiples been trending down?
Marcin Majewski
They’ve been stable over the last few quarters, though obviously down from the COVID-era peaks. Still, many businesses are now growing slower, so while the multiples may seem stable, you’re effectively paying more for less growth—making it a higher multiple in real terms.
Q: What’s the current market for small but fast-growing SaaS companies under $5 million in revenue?
Marcin Majewski
It depends. If the company is AI-native, there’s strong interest. For traditional SaaS, buyers are digging deeper into fundamentals: where growth is coming from, sustainability, and whether there’s a defensible moat.
Q: How do you achieve product visibility in LLMs?
Filip Drazdou
It’s still early days. There’s a lot written about SEO, but not much about generative engine optimization. Key steps include having clear, structured content about your product—ideally text that’s machine-readable, even if it’s too dense for humans. Also, getting cited by reputable sources like Forbes or NYT helps. LLMs lean on trusted references to avoid hallucinations.
Marcin Majewski
Social media platforms like X can also play a role, especially if you’re targeting growth-oriented users. And traditional SEO still matters—Google’s AI Overviews often pull from the top-ranked pages.
Q: Will you explore other sectors not yet impacted by AI?
Marcin Majewski
We might look at IT services soon, but it’s still early to predict which sectors will benefit most. Eventually, most will. For now, we’re focused on those already seeing change.
Filip Drazdou
We’ve talked about the AI value chain—companies involved in infrastructure, construction, or HVAC that are benefiting indirectly. But it’s speculative to talk about untouched sectors just yet.
Q: Is the Rule of 40 still important to buyers?
Marcin Majewski
Honestly, it doesn’t come up much anymore.
Filip Drazdou
It was once a proxy to balance growth and profitability, but now investors prefer to look at unit economics in more detail. High-level shortcuts like Rule of 40 are less relevant in today’s cautious environment.
Marcin Majewski
Exactly. The word we hear more often is “trend”—as in, are your KPIs trending in the right direction?
Q: What about AI-empowered SaaS?
Marcin Majewski
That term can mean a lot of things. It could be a thin wrapper around ChatGPT, which likely has limited value. But if you’re talking about a company deeply integrating AI into its operations and decision-making, that’s much more meaningful—and potentially very valuable.
Q: What defines the value of AI-empowered SaaS?
Marcin Majewski
It depends on the depth of AI integration and the defensibility of your data. If you’re just layering prompts on GPT with no proprietary data or lock-in, that’s not sustainable. But if you’re truly leveraging unique data and embedding AI across your stack, that’s valuable.
Filip Drazdou
Exactly. Tools that just repackage GPT with a basic UI and high margins won’t survive long-term.
Q: What revenue multiple is appropriate for Rule of 40 SaaS companies?
Marcin Majewski
It varies. For a 40% EBITDA company with no growth, think 5–7× EBITDA, or 2–3.5× revenue. For a high-growth, no-profit business, we look closer at metrics like churn and customer retention. If it’s sticky and has high net retention, the revenue multiple might be even higher. If not, much lower.
Q: If the tech bubble bursts, will SaaS valuations improve?
Filip Drazdou
Not in the short term. When big tech drops, smaller companies usually suffer more. In the medium term, reduced infrastructure and labor costs might help. But overall, valuations would likely fall before they recover.
Final Thoughts
Filip Drazdou
Good luck, everyone. We’re in a period of rapid transformation, and adaptability will be key.
Marcin Majewski
Thanks for joining. Stay alert, stay flexible—and let us know if you need help assessing your readiness for AI or planning your next move.
