"The M&A negotiation table is a battlefield. I'll share the hard-won lessons I learned from selling my company, including the common traps, negotiation tactics, and how to avoid getting screwed by experienced corporate development teams."
Why 90% of Founders Get AI Competitive Moats Completely Wrong
I’ve sat on both sides of the table. As a founder who navigated the treacherous waters of a successful exit, selling my company RemoteTeam to Gusto. And as an angel investor in over 200 startups, placing bets on the next generation of world-changers, including giants like Anthropic, OpenAI, Scale AI, and Hugging Face. I’ve seen the inside of more pitch decks, boardrooms, and late-night deal negotiations than I can count. And I’m telling you, the story most founders tell themselves about their "competitive moat" is a dangerous fantasy.
They come to me, eyes wide with the fire of their conviction, talking about their "proprietary algorithm" or their "all-star team of PhDs from Google." I listen patiently, nod, and try to gently guide them back to reality. But what I’m really thinking is: that’s not a moat, that’s a puddle. A shallow, temporary puddle that’s going to evaporate the second a real competitor turns up the heat.
This isn’t just a theoretical exercise for me. This is scar tissue. When we were in the trenches building RemoteTeam, we were a small, scrappy team going up against giants with budgets bigger than our entire seed round. We couldn’t out-spend them. We couldn’t out-hire them. We had to be smarter. We had to build a defense that wasn’t based on things that could be easily copied or bought. We had to build a real moat, brick by painful brick. And that’s a lesson I’ve carried with me into every investment I’ve made since.
The Great Illusion: Chasing Ghosts in the Machine
Let’s get one thing straight right now: a cool feature is not a moat. A slick UI is not a moat. A model that scores a few percentage points higher on a benchmark is not a moat. These are, at best, temporary advantages. They’re the kind of thing that gets you a fleeting moment of glory on Product Hunt or a nice little write-up in a tech blog. But they won’t protect you when a well-funded, determined competitor decides to eat your lunch.
I see founders fall into this trap constantly. They become obsessed, spending months, sometimes years, in a hermetically sealed lab, perfecting their algorithm. They believe its mathematical elegance and computational efficiency will be their salvation. But here’s the harsh, cold reality of the AI landscape in 2026: the pace of innovation is not just fast, it's brutally, relentlessly fast. The open-source community is a global force of nature, and the big platform players (Google, Amazon, Microsoft) are in a perpetual arms race to commoditize the cutting edge. Your clever little algorithm? It’s a speed bump, not a fortress wall. It will be replicated, open-sourced, or offered as a cheap API call faster than you can say "paradigm shift."
And what about the "all-star team"? Look, I love brilliant engineers as much as the next guy. I’ve hired dozens of them. But a team is not a moat. It’s a valuable, critical asset, for sure, but it’s a mobile asset. In the hyper-competitive AI talent war, loyalty is a rare and expensive commodity. Your star engineer, the one who built your core model, can and will be poached by a bigger company with deeper pockets and more stock options than you can imagine. I’ve seen it happen more times than I can count. Relying on a handful of key people is not a strategy for defensibility; it’s a recipe for a very stressful, and likely short, founder experience.
I remember one startup I advised a few years back. They had a team of absolute rockstars, the kind of engineers who could make a model sing. They built an incredible piece of technology, truly groundbreaking stuff. But they were so mesmerized by the tech that they neglected everything else. They had no real data advantage, no distribution channel, no ecosystem to speak of. When a big, slow-moving incumbent finally woke up and released a "good enough" version of their product as a free add-on, the startup’s customers, who had no real reason to be loyal, fled. The all-star team couldn’t save them. The company was acqui-hired for pennies on the dollar. A tragic, and entirely avoidable, outcome.
The Real Moats: Building a Fortress That Lasts
So, if it’s not about the algorithm or the team, what is it about? Where does true, durable defensibility lie? The real moats in the age of AI are built on structural advantages. They are harder to build, they take more time, and they are far less glamorous than a breakthrough algorithm. But they are the only things that will actually protect your business in the long run.
1. Proprietary Data: The Unfair Advantage You Create
This is the big one. The alpha and the omega of AI moats. If you have a unique, proprietary dataset that your competitors cannot easily acquire or replicate, you have the foundation of a powerful, lasting advantage. This isn’t just about having a lot of data; it’s about having the right data. Data that is clean, well-labeled, and, most importantly, generated as a unique byproduct of your business operations.
Think about it. An AI model is a voracious learning machine, but it’s only as good as the data it’s fed. If you’re training your model on the same public datasets as everyone else, you’re going to get the same commodity results as everyone else. But if you can create a data flywheel—a closed loop where your product generates unique data, which you then use to improve your model, which in turn makes your product better and attracts more users, who then generate even more unique data—you have a virtuous cycle that becomes a powerful competitive barrier. Each new user makes the product better for all other users. This is a classic network effect, but supercharged with AI.
This is why I’m so bullish on companies that are tackling "dirty data" problems in unsexy, complex industries like logistics, manufacturing, or healthcare. The data is hard to get, it’s messy, it’s fragmented, and it requires a ton of domain expertise to understand and label correctly. It’s a slog. But if you can crack that nut, if you can be the one to aggregate and structure that data, you’ve got something truly special. You’ve built a data asset that no one can just buy off the shelf.
2. Distribution & Deep Workflow Integration: The Sticky Web
This is the moat that we built at RemoteTeam, and it’s the one that I think is most consistently underestimated by technically-minded founders. You can have the best technology in the world, but if you can’t get it into the hands of your customers and make it an indispensable part of their daily lives, it doesn’t matter. And once you’re in, you need to make it as painful as humanly possible for them to even think about leaving.
How do you do that? By becoming part of their daily workflow. By weaving your product into the very fabric of their business operations. At RemoteTeam, we didn’t just offer a standalone HR tool. We plugged directly into the payroll and compliance systems that our customers were already using. We became part of the plumbing. We were the system of record for international hiring. That made us incredibly sticky. The thought of ripping us out and migrating all that data and all those processes to a new system was a nightmare. The switching costs were immense.
When Gusto came knocking, that’s what they saw. They didn’t just see a product; they saw a distribution channel into a rapidly growing market segment. They saw a customer base that was locked in, not by iron-clad contracts, but by the sheer, unadulterated inconvenience of switching. That’s a powerful, powerful position to be in during an acquisition negotiation.
3. Community & Brand: The Army of True Believers
This is the most intangible of the moats, but in many ways, it can be the most powerful and enduring. If you can build a genuine community of passionate users who love your product and believe in your mission, you have an army of evangelists who will fight for you. They will be your marketing team, your support team, and your product development team, all rolled into one.
Hugging Face is a masterclass in this. They didn’t just build a library of models; they built a movement. They created a central gathering place, a digital town square for the entire AI community. A place where developers could come together to share, collaborate, and learn. That community is their moat. It’s what makes them the default choice for millions of developers, and it’s what will protect them from the giants who are desperately trying to build their own competing platforms. You can’t just buy that kind of organic, grassroots loyalty.
Building a community is not easy. It’s not a marketing tactic you can just spin up with a few blog posts. It takes time, authenticity, and a genuine, unwavering commitment to your users. You have to show up, you have to listen, and you have to add value long before you ask for anything in return. But if you can pull it off, you’ll have a moat that money, no matter how much of it, simply cannot buy.
The Exit Perspective: What Acquirers Really, Truly Look For
When you’re sitting at the negotiation table, sleeves rolled up, coffee going cold, the acquirer isn’t just buying your technology. They’re buying your business. They’re buying your customers, your revenue, and your position in the market. And most importantly, they’re buying your moat.
They are running a cold, hard calculation. They’re asking themselves: "If we buy this company, how hard will it be for our competitors to replicate what they’ve done? How sticky are their customers? How defensible is their position in the market? Can we build this ourselves for less than the acquisition price?"
I can tell you from my own experience selling RemoteTeam that the conversation was 90% about our integration into the Gusto ecosystem and our sticky customer base. It was about how our product would make their platform more valuable and harder to leave. Our algorithm was an afterthought, a checkbox. They knew they could build the tech. They couldn’t build our position in the market.
So, my advice to every founder who will listen is this: stop obsessing over your algorithm and start obsessing over your moat. From day one, from the very first line of code you write, be thinking about your data, your distribution, and your community. Because in the brutal, unforgiving world of AI startups, that’s the only thing that will save you. That’s the only thing that will let you build a business that lasts.
Frequently Asked Questions
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
How has this view evolved over time?
My thinking on most topics has changed significantly over the years. Early in my career, I held many conventional views that experience proved wrong. I try to update my beliefs when the evidence changes.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.