39 | SaaSpocalypse? How AI Is Reshaping Software M&A

AI is reshaping what buyers and investors look for in software businesses faster than most founders realize, and many are still operating with an outdated playbook. In this episode, managing directors Mike Lyon and Jeff Koons break down what is driving the so-called SaaSpocalypse, why we think it is more of a market re-rating than a collapse, what criteria investors are now using to evaluate software companies, and what concrete steps founders can take right now to assess and help strengthen their position.

Vista Point Advisors is a boutique sell-side investment bank providing unconflicted M&A and capital raising advice to founder-led software, AI, and internet businesses. We partner with entrepreneurs of growing businesses to help them understand their options in the marketplace so they can maximize business value, leverage their options, and realize their ideal outcomes. 

Securities offered through Vista Point Advisors, member FINRA/SIPC. This has been provided for informational purposes only and should not be considered as investment advice or a recommendation. It is not intended to address all circumstances that might arise. The views expressed herein may change at any time subsequent to the date of issue. Opinions contained herein should not be interpreted as a guarantee of future results. Outcomes will vary depending on individual circumstances. Any examples used in this material are generic, hypothetical and for illustration purposes only. Testimonials from past clients may not be representative of the experience of other clients and there is no guarantee of future performance or success. Clients are not compensated for their comments. 

Transcript

[00:00:19] Mike Lyon: [00:00:00] Hello and welcome everyone. I'm Mike Lyon, founder and managing director at Vista Point Advisors, and this is the Path to Exit. This show is dedicated to helping founders of software, AI, and internet businesses understand what it takes to raise capital or sell their business, and how to do it well. My guest today is Jeff Koons. 

In this episode, we're going to talk about the SaaS apocalypse and how it's impacting private software and AI businesses, which companies are now in favor and how to think about your AI story. Please enjoy my discussion with Jeff. So a common question we get right now is what is this SaaS apocalypse and what does it mean for us and how bad is it going to be? 

I would just say for software and AI founders out there, I think everyone needs to take a deep breath. The market and the media loves drama, and obviously [00:01:00] a lot of this is coming from some of the public company commentary. This is really more what we think of as a re-rating or a refocus of what a quality company is. We've had this several times over the past 20 or so years. There was a big reset after the great financial crisis, there was this big transition from license to maintenance to SaaS. And then as recent as 2022, we had this really big reset where everyone was suddenly focused on profitable growth as interest rates went up. 

So this one is a little bit faster moving and more intense, but software businesses are still amazing businesses to own. But some things have changed, and that's what we're going to walk you through now. I would just say this too will pass. And it's similar to other uncertainty we've had with what happened with AI and content businesses. 

You'll remember a few years ago when the AI focus was mainly on generating content. There was a big reset there, but obviously lots of content businesses are still doing really well. So, Jeff, to get us started, just talk about what caused this turmoil starting the January February timeframe. What was the catalyst and where have we been moving to?  

[00:01:59] Jeff Koons: Yeah, [00:02:00] absolutely, Mike. Thanks for having me on. It is a dynamic and exciting time. And really, I feel the catalyzing event here in late January, early Feb was Claude's release on its 4.6 Sonnet extended, specifically towards the legal vertical. So before then, I think a ton of people used ChatGPT, Claude, whatever, and they found a lot of different personal job and personal tasks really efficient in it. 

But when you then tried to like scale that out enterprise wide, it just didn't work. It was klugy. You could connect anything to anything that well, and it just didn't feel like B2B ready. All of a sudden in February, Claude dropped this thing and it was really good. So what happened? Immediately that day, all the legal tech businesses traded down between like 20 to 45%. And the idea was, why would you need this purpose-built legal tech software if Claude can just do everything? And then they also released the financial services one for like banking, wealth management, things like that. 

And that, too, was really [00:03:00] good. And ultimately what I think it did is it really sharpened the focus that the LLM'S ability to replace or augment software in the B2B context, it's here and it's here right now. And so what that ultimately drove was very, Very different conversations for our founders and our clients, both in terms of the types of questions investors asked, and frankly, like what is important in today's market to get the kind of high valuation and good liquidity outcome that we Vista Point are focused on. 

All that's going on, and then the different market participants have reacted in different ways. The PE firms are in flight changing some of their plans for their existing portfolio companies. Strategics are asking very different questions around the technical capabilities of these systems. 

And I think a lot of founders were talking to are starting to feel it, that the tenor, the tone of the conversation is changing. So that's, we're at now in this period, I think you said it very well, Mike; this kind of refiguring out: what is a good company and where do we want to make good deals at this point? 

[00:03:59] Mike Lyon: [00:04:00] Yeah, great point Jeff. And I think what's interesting is for the past, I don't know, 10 or 15 years, what made a good software business, frankly hadn't changed that much. Right? If you have good growth, good retention rates, good margins, good TAM, good CAC to LTV; those were the characteristics that defined a really high quality software business. 

And obviously those businesses would price a little bit differently when they were sold based on those metrics and strategic value. If you asked a banker 10 years ago, what's a good SaaS business with what was a good SaaS business nine months ago, it would largely be the same thing. That is what is starting to change quite a bit. 

And I would just say that in January and February, we might hear wildly different things from investors about what they were looking for day to day as they were trying to figure it out. In general, the criteria we're getting ready to give you, this is, talking to hundreds of investors, hundreds of strategic buyers. 

This is the folks who we feel like have started to figure it out where they're coalescing. I would say there's some smaller funds or less active funds and buyers who are still kind of wait and see mode. But in general, the criteria we're going to [00:05:00] give you feels like the criteria that's roughly going to stick for a good company going forward. 

And when I say good company, I mean if you ran a process, it would be relatively competitive. So Jeff, maybe walk us through the list of what makes a good company right now and maybe highlight what are some of the different questions investors and buyers are asking now than maybe six months ago?  

[00:05:18] Jeff Koons: Yeah, absolutely. And one thing I really do want to hit up front here: to be a good company does not mean you have to be an agentic first AI business. I do want to clarify that. There is tremendous value in software, but there's more value in other areas of software than there used to be. And the first thing I want to hit on is horizontal versus vertical. 

I do think a lot of the horizontal players are in trouble. They are generally incredibly good at workflow optimization and everybody in HR has to do this thing, and it's a known workflow. In LLM, if you have that kind of very defined outcome that doesn't really change and it's not super nuanced, it is really good at that stuff. 

Other [00:06:00] side of what we're hearing, what we're seeing, what we're getting in bids: the more highly verticalized you are, you are in a much, much better position. And this makes sense when you think about the training models, how these out frontier models are ultimately put together. 

They're deep, but they're skimming across the surface of a range of things. If you're, let's make it up. Let's say you're an ERP business for HVAC or plumbers. So you're very niche, you're very narrow. If you have a purpose-built software that says, I don't think Claude understands that there's 14,000 different part numbers for this type of pipe fitting, and only these two are going to work because it's a, whatever, nickel finish or a pewter and something's going to interact there. 

That's the kind of depth that the AI is just not going to get to. And if you are a plumber in this example, that really is valuable to you and you're honestly probably not looking to vibe code that yourself anyway. But that verticalization is really important because your depth of workflows and your knowledge set is just [00:07:00] so much more applicable than the kind of median output that an LLM will provide. So vertical, better than horizontal right now. 

From a metric standpoint, gross retention and new bookings, net new bookings, are I think the two biggest ones with gross retention being the most important, and then new bookings being number two or one B. And the reason why is, if you're maintaining 90% plus gross margins, which we think of as the benchmark now, to go to an LLM or some kind of AI-enabled product. That's a very sticky customer base, which is just proof that your software is valuable to your customers. 

[00:07:36] Mike Lyon: And Jeff, just one thing I wanted to interrupt there on the gross retention. It's not like there's really anything magic around 90%. This is just investors taking the bar up a little bit and kind of using it as a crutch for AI risk. Obviously they're looking at a lot of other things, but if you're a 75% gross retention business, that is materially different than a 90% gross retention business. And this is a little bit of [00:08:00] a change. In the really growth on years, people were a little bit more focused on net revenue retention, but this is viewed as a little bit more of a AI protector, if you will. And again, it's more of a crutch right now and there's more substance being built behind this to analyze AI risk. 

But it's a crutch for investors and a lot of folks are drawing a soft or hard line around that number and like how investible a company is.  

[00:08:21] Jeff Koons: T hat's a really good point. Like gross retention is a heuristic for AI risk. I really like that concept, Mike and I do think it's fair. And then quickly on the other one, new bookings, all that does is signal there's still meaningful demand. And why new, not upsell, is because it means that potential client or customer is shopping. And that means they're exposed to all the new AI tools, all the AI agentic models. And if you're selling software to new logos in this world, that means you're winning on the value proposition at this point.  

And then the two other things I'll hit on quickly, Mike, in terms of: what questions are we getting asked? What companies are better positioned in this current environment? Ideally you're a system of [00:09:00] record. I think we'll have a podcast in the future about the different kind of strata of AI and how to think about their value. But suffice to say that being a system of record simply means you are the single source of truth for the data that lies within it. You can play very well with LLMs as a system of record. They can toggle your data if you have an MCP, but if you get a piece of data from that system of record, that is the answer and that is the source of truth. And it's valuable to have that. 

And then the final thing is on the defensive side. So what are things I would say that make you less susceptible to AI risk? Everybody here has heard this one: the defensible data moat is really important if you have it. I think people are a little over optimistic about how defensible their data moat actually is, but if you do have a unique, true proprietary data set, that is valuable because at the end of the day, the LLMs and AIs are designed to generate outcomes. If you have better data and you generate better outcomes, you could be a winner. And then the final thing we are seeing is anything that [00:10:00] has payments heavily integrated, the frontier models aren't great at that yet, but we're very hesitant when we're working with a client, when we pitching strategic investors to say these models can't do that. That's probably going to be wrong, you know, 12-24 months from now. 

[00:10:15] Mike Lyon: Absolutely. And the other criteria I would just add is, ideally a business would be Agent AI. That's a very small percentage of the companies out there, just given how new it is. But I think a key criteria would be if you're not agent AI, do you have a platform where you can distribute agents off the platform? Either as you retool the business for a more AI first business, or you're just able to distribute these agents and provide these outcomes to your clients? 

As we're shifting from a little bit more of a software; have your employees be more efficient, to more of; we're selling outcomes now for the companies that you're selling to. So just to reiterate all those, because I know we went through a lot. A good company in our view is now more vertical than horizontal. 

Frankly, we've always felt that way, but AI's really enhanced that. Gross retention is even more [00:11:00] important than ever. And we're talking about this 90% number. It doesn't have to be exactly that, but trending towards that. Ideally, you're a system of record and you have some moats, whether that be data or payments. 

And then we just talked about, either you're agentic AI or you have the ability to distribute these agents off your platform. That seems like where most of the market is coalescing right now for a really competitive process. And again, as the aperture shrinks, the companies that are in that aperture become highly valuable because you have a big investor in buyer universe competing for those companies. 

And so that's what gets us excited about this part of the market. I think, one of the things that's talked about a lot is obviously, coding costs are coming way down. So that should put pressure on software prices that, you know, you can sell to your customers. And obviously, there's TAM impacts and Tam's a big driver of valuations. 

I think that's all true, but I can't tell you the number of software businesses we've talked to in the past that had a really good product, no idea how to sell it or run an organization that could deliver those [00:12:00] products. So Koons, maybe talk a little bit about some of the other factors that go into building an awesome software business, even if the cost of coding goes way down. 

[00:12:08] Jeff Koons: Yeah it's super interesting right now. We touched about why verticalized software, we think, is poised to be a winner in this world. So I won't go back over that. But again if you have really good vertical expertise, you'll build a purpose-built product that solves those problems specifically. So that's one. 

Two, it it's funny for the past 10 years, as you said Mike, I don't know how many people we talked to that's like, "Look, we built the best mouse trap, so we won, right? It's over?" And we'll just be like, "But how do you think about sales and marketing?" And they'll be like, "Look, I'm a technologist, man, sales and marketing... who cares?" And we're always just like, "yeah, but you've got to sell the software, right?" Now it's becoming even more important to have really robust go-to market acumen, really good sales hygiene, and sales discipline because there is a world where technology is just pretty commodified. 

Now, again, I think if you have a verticalized offering, you can make that [00:13:00] less commodified, but there is going to be a lot more pressure on your go to market. Are you channel? Do you good enter salespeople? Do you have good customer support to enhance retention? Those are going to become more important because the technology differentiators are just going to slowly level out as coding becomes a commodity and tech becomes a commodity. Those more soft skill traditional aspects of the business are actually going to become big differentiators moving forward. 

[00:13:26] Mike Lyon: Absolutely. And one other point back to the TAM point I made where, you know, in theory the TAM for some of these markets could come down as just the amount you can charge for the software goes down. I think this misses a pretty key point. And not to date myself, but if you go back to the early days of software, remember there was a world where IT services was very prominent before software. 

And then founders got smart and said, rather than running a services business, I'd rather start a software business. So the services business is selling an outcome to their clients. The software business is, I'm going to sell you this really high margin software that allows you to make your [00:14:00] employees much more efficient. In agentic world, we're back to selling the outcomes, right? 

We'll just do all this for you. And no one really wanted to do that as a tech company in the past 10 or 15 years because that meant hiring a lot of employees to execute on those services. And those were not fun businesses to manage or grow or super capital efficient. In this world, you could see where the software TAM is shrinking, but now that Services TAM is getable for the software and AI business. So I think you're going to see quite a bit of expansion here as you're selling outcomes. And now with agents, you can go deliver on some of these procedures and outcomes that software companies did not want to deliver on in the past. 

And so I think that's what gets us really excited. I think that's a little bit of a ways off, right? As that's figured out and the AI improves enough. But I think that's what gets us really excited about what the next couple years looks like. 

[00:14:50] Jeff Koons: Really good point, Mike. It is interesting and we'll see where it goes, but we absolutely see software able to access a bigger service TAM than before. And then the question [00:15:00] becomes how much is that offset by the desire to just purchase the software going down? That's the push and pull there. 

[00:15:05] Mike Lyon: Absolutely. Well, before we wrap up, Jeff, maybe talk a little bit about some concrete steps founders could take right now as they're evaluating their business, how attractive that business is, and ways to mitigate some of the risks that AI brings to the market.  

[00:15:18] Jeff Koons: Yeah, absolutely. I think where I start is with the retention metrics. Being able to show really excellent gross retention is just critical. So you should be able to calculate it, you should be able to view it with different cohort slices. So that, to me, just absolutely know that. Number two is, is a little bit of soul searching we're recommending all our founders do right now, which is really do a proper product audit through an AI lens, and use AI to do it. 

We have had a bunch of conversations with founders who were like, "Yeah, you know, I didn't really believe it. I messed around with Claude Code Enterprise for about a weekend, and I was able to code 80% of the product. It took me eight years to build." And that's a real, sobering [00:16:00] moment in terms of, okay, the technology at the end of the day is not necessarily going to win this for me. I have to enhance these other aspects of my business, as we talked about earlier in the show here. But just engaging with a audit of your product. What are the risks? What are the opportunities? Through the AI lens. I just think is a very prudent thing for founders to do right now. 

If you are having conversations with strategics and investors, don't over claim. Document them specifically, how they work and the value add they generate. Don't stretch because this stuff, if it works, it works quickly and obviously. And you'll be asked to just pull it up on a demo. So be careful not to over claim.  

And then the final thing I would say is, if you are thinking about going to market for a deal, I would really advise you to speak to a banker or an advisor. Granted, somewhat self-serving, but the market as Mike's said up top has changed really fast in a really short period of time. And I know every founder who's listening to this has been called by 8 million private equity firms and has had a lot of conversations over the [00:17:00] past couple years with them. What they're looking for is just different than what they told you they were looking for a year before. So, we obviously are in market. We're talking to all these buyers, all these investors. We have a great sense of the pulse, but what you thought you knew from eight months ago, 12 months ago, is just not going to be how they're thinking today. 

[00:17:17] Mike Lyon: And back to the vein of, take a deep breath. I think one of the things that's always true, if you're not familiar with this, look at the Gartner Hype Cycle for every big innovation in technology. Go look at the one for internet, right? There's this massive period of panic at the early stages where you think it's going to take over and dominate very quickly. 

And then there's this huge trough of disillusionment. That's where we're headed, right? Where some of these things don't work out as well. But in the long term, you actually get more out of the the internet than we thought, if you go looked at the internet hype cycle. What that means is there's actually a lot of time to mold your business and think about this as an opportunity, not just a threat. And so, yes, coding will be cheaper, it will be easier, but that's a small part of building a [00:18:00] business and look for that as an advantage. Today we covered how AI is reshaping the types of questions buyers and investors are asking, what they're looking for, how they're evaluating businesses, and where founders may have some blind spots. This is going to evolve a lot so, this podcast might be outdated in two or three months, but I think the core concepts there are going to remain. Please feel free to reach out to us if you have any questions. And Jeff, thanks for joining us again. 

[00:18:23] Jeff Koons: Thanks so much for having me, Mike. Appreciate it.  

[00:19:00]