收购即落地：以并购战略重塑传统服务业的 AI 商业化新范式

[music] >> Hey everyone. My name is Pratik Naidu. I'm a founding partner at Long Lake and today I'm excited to talk a little bit more about why acquiring businesses is a unique strategy to deploy AI into the real world. So to start out very familiar experience to many of you. Let's say you're building an AI product and now you're trying to sell it into the enterprise. Well, where do you start? You found a pain. You built an agent for it and it works, but for one very narrow use case. It's a great starting point. But now you have to actually get it into production. And that might involve dealing with security review, internal coordination, integrations with 20-year-old systems. It's a lot of hard work and this is typically the first wall in order to deploy AI into an enterprise. But once you get past this, now you have to get the team to use it. So you typically will iterate on the product interface, go through trainings, get feedback from users. And there's a typical question that comes up, which is whose job is it to actually drive up that adoption? Is it yours as an AI vendor or is it the management team's job? And you know, you can typically find a few friendly champions, but getting true adoption across the entire company is really hard. And that's usually the second wall in order to deploy into the enterprise. And so, what does this actually lead to? You'll typically work through a few different questions. You know, how do you handle security permissions? How do you actually get data that is stuck in 20-year-old systems out of it? And the bigger picture question here is as an AI startup deploying into the enterprise, how do you cross this chasm of selling a point solution to actually driving true enterprise transformation? And so, this leads to three different observations of why deploying AI into businesses is challenging. The first is that organizations are incredibly complex. On paper, your org chart might look something on the left. You know, you have your owner, you have teams, managers, and this is a standard team structure. But, the reality is that businesses are complex living organisms. There's unofficial leaders. There's sub-teams that operate on their own rhythms. And these dynamic teams drive how work is actually done. And so, understanding that is critical to getting adoption. The second observation is that workflows are messy. You might start with a very clean, straightforward, documented standard operating procedure. But the reality is that the real work is messy. It's the exceptions, the edge cases, the tacit knowledge, the, you know, quick questions that get asked over Slack or in the office. These are where the true work is done and it's very hard to capture that tacit knowledge from the outside. And the third observation is that, you know, value accrual is deeply misaligned. So, let's say you do all the hard work. You get past that first wall of deploying into production. You cross the second wall of getting adoption. And now, you've saved hundreds of hours for the team. But from a value accrual standpoint, your pricing is probably tied to software seats or usage. And who captures all that value? The business, right? So, let's take a simple example. A services business doing, say, $100 million in revenue with 10% EBITDA margins. So, they're generating $10 million in EBITDA. If you can come in and automate even just a small percentage of their workflows, say, 50% of the back-office or core work, you might be able to double their EBITDA margins to 20%, which means they're now generating $20 million in EBITDA. That's a $10 million increase in recurring earnings to a business with value that you can already reason about. And so, you know, using a, you know, roughly conservative 10x multiple, you can actually generate for your client $50 million in enterprise value, which is really interesting. And so, there's this question here, which is, you know, as a vendor, you did all the work. You built the technology. You had to go through all the work of deploying into the company and getting people to use it. But you captured just a tiny fraction of that value. And why are we giving up all of that value when software vendors created all of that upside? A lot of big questions to answer. And the market reflects this. What we're seeing is that services businesses are now trading at incredible discounts, right? And there's a reason for that, which is they're being priced as though AI is going to disrupt them. But the reality is that AI might actually make them more valuable, not less. And so, there's this question here, which is, you know, why are we capturing so little value and why is the market pricing these businesses so cheaply? And this led us to our core thesis at Long Lake. So instead of being outside in and selling an AI agent to a company or trying to build from scratch, what if you could partner far more closely with incredible businesses that already exist through acquisition?

And I want to talk a little bit more about why ownership is such an interesting strategy here. Starting with on the engineering side, which is access. The ability to have your team sit down with the people who do work, observe their workflows, see the edge cases in real time, and understand the tacit knowledge that lives inside their heads. That is an unfair engineering advantage that you simply cannot get as an external vendor. The second is data. When you own the company, you own the data. You have complete access to the proprietary training data that has been built over decades of customer interactions, workflows, and edge cases. And that data flywheel can be fed directly back into your models to create something truly differentiated. And the third is alignment. When you're an owner, your incentives are completely aligned with the business. You're not trying to optimize for software seats or usage metrics. You're optimizing for enterprise value and real bottom-line profitability. And that means you can make the right long-term product and architectural decisions that actually move the needle for the business. And this flywheel is really interesting, and it comes with all the elements of ownership to uh you know, power this flywheel from from day one. And, you know, importantly, rather than coming outside-in or building from scratch, ownership has the ability to accelerate your go-to-market by 5 to 10 years, right? You immediately acquire a brand that customers trust, existing distribution, real cash flows, and proprietary data that you can use to build your technology on top of. And so, what's interesting here is that ownership uh can compound rapidly over time.

So, we talked about why ownership is is really interesting. But what markets are ripe for acquisition, especially in the AI era? I want to start with a broad market observation. This is a global software market. It's about $600 billion to $1 trillion in size. It's a huge market. It's created a lot of incredible software companies over the last couple of decades. But let's zoom out. If you zoom out, this is the global services market. It's tens of trillions of dollars in size. It's massive. It represents a huge portion of the global economy. And the reason it's so large is because most of the work in the world is not done by software. It's done by people. And it's done through services businesses. Everything from accounting, to legal, to compliance, to logistics, to healthcare services. These are all services businesses that are powered by human labor. And the fundamental economic model of a services business is broken. Because revenue scales linearly with headcount. If you want to double your revenue, you have to double your headcount. And that means your gross margins are capped, your operating margins are capped, and your ability to scale is fundamentally constrained by your ability to hire, train, and retain talent. And so, what you end up with is a business that looks like this, where revenue scales proportionally with labor costs. And that's why services businesses trade at 1 to 2x revenue or 5 to 10x EBITDA, whereas software businesses trade at 10 to 20x revenue. But what happens when you introduce AI to this equation? What's interesting here is that AI actually reduces the cost of service while also improving the quality of your service. So, look back at our chart here where revenue scales proportionally to labor costs, but when you deploy AI you can see here that you get actually operating leverage. So, your labor cost no longer scales linearly with revenue. In fact, it flattens out, while your revenue can continue to scale exponentially. And what that means is that your margins expand dramatically. You go from a 10 or 15% EBITDA margin business to a 30, 40, or even 50% EBITDA margin business. And when that happens, the valuation multiple of that business re-rates. It goes from trading like a services business to trading like a software business. And that is where the real value creation happens. But what types of businesses are right for this? Not every services business is created equal. We've spent a lot of time thinking about this at Long Lake, and we've developed a framework for the types of businesses that are most ripe for this kind of transformation. And it starts with a simple observation, which is that intelligence is getting cheaper, right? With every new frontier model, the cost of intelligence is dropping by orders of magnitude. And that trend is only going to continue. So, if intelligence is getting cheap, what remains expensive? We really focus on mission-critical businesses, and it starts with a few key questions that we typically evaluate these companies on. The first is is there a customer base that trust these businesses with real end-to-end outcomes? The second is is there human judgment in the service that's actually being provided? And the third is that are there, you know, rich proprietary workflows and data that exist within that business. So, looking at these three different criteria, we think about businesses that are mission-critical. If the business goes down or if the service fails, it's a catastrophic outcome for the customer. These are things like regulatory compliance, specialized tax, legal services, complex supply chain logistics, back-office operations for healthcare. These are not nice-to-have services. They are must-have services. And customers are willing to pay for certainty and outcomes, not for software. And so, that's what we actually focus on and who we partner with.

Awesome. So, once you've found your business and you've partnered with a company, what does P&L transformation actually look like? And I think a very simple way to explain this is actually through the visual of a J-curve. So, on day one, you make the acquisition, you start deploying your technology. And the first part of this curve is actually really hard. You have to do all that engineering heavy lifting, right? You have to build the integrations, clean up the data, map the workflows, sit down with the teams, and overcome the resistance to change. And that takes time and effort. There's a real upfront investment that has to happen. And this is why most external software vendors fail, because they don't have the patience, the capital, or the access to get through this initial trough of the curve. You have to overcome the messy workflows and data that exist within that business. But once you make it up this hill, you'll start to see folks actually using your product, which is incredibly exciting, and you start to actually unlock a lot of capacity. Time savings, usage is going through the roof. They're producing the same output, fewer hours, and this is where most AI startups stop. They celebrate the time savings. They say, "Look, we saved our customer 50% of their time, or we automated 80% of this task." But time savings alone does not transform a P&L. If you save someone 10 hours a week, but they just spend that 10 hours browsing the web or doing low-value tasks, you haven't actually created enterprise value. And so, you actually have to put that capacity to work. Think about incentive alignment, working very closely with our management teams and trusted operators, to ensure that the capacity you unlock is directly channeled into revenue-generating activities. Can you take on more clients without hiring more people? Can you offer new premium services that you couldn't offer before? Can you deliver a turnaround time that is 10x faster than anyone else in the industry, and therefore win market share and charge a premium? That is how you turn unlocked capacity into real revenue growth and margin expansion. And that is what drives the exponential ascent of that J-curve. And I think what's cool about this is that this isn't just a bespoke playbook that you run once on a single company. If you have the platform and the playbook to do both the AI deployment and the business transformation, you can repeat this across an entire sector. You can acquire a platform company, build out the AI-native operational infrastructure, and then execute add-on acquisitions where you immediately migrate the acquired customer base and volume onto your highly efficient, AI-powered platform. And when you do that, the synergies are not just 10 or 20% cost cuts like in traditional private equity. The synergies are structural, margin-expanding, and transformative. So, we talked a lot about the opportunity, the thesis, and how we actually think about executing on this. But I want to leave you with one final thought, which is that we are at the very beginning of a generational shift in how services are delivered. And if you are an AI builder or an operator, think about that at that J-curve, right? Don't just stop at capacity and time savings. How do you actually ensure that the work you're doing is driving real business impact for the companies that you work with, and delivering a far better client experience uh for the service that you're providing? So, if you're excited about owning outcomes and partnering with uh incredible services businesses, feel free to reach out. Uh we're excited about this thesis. We think there's a lot of opportunity to go continue to build in the services economy, and uh feel free to chat more if there's any questions. Thank you.