What Makes a Moat: Durable vs Temporary Competitive Advantage

The difference between a business that doubles every decade and one that gets cut in half is often invisible until it's too late.

I'm here with Edge, who spent fifteen years analyzing competitive positioning for institutional capital. Edge, let's start with the foundation. When you're evaluating whether a company has a durable moat versus something that just looks good in a pitch deck, what's the first layer of distinction you draw?

EDGE: The first thing I look at is whether the advantage compounds or decays. A real moat gets stronger as the business operates. Network effects are the clearest example. Each new user makes the product more valuable to every other user. Facebook in 2008 was decent. Facebook in 2012 with 800 million users was nearly impossible to displace because everyone you wanted to reach was already there. Compare that to something like brand preference for a restaurant chain. That can erode quickly if execution slips or tastes change. It doesn't build on itself.

Right, and that compounding quality is what separates the durable from the temporary. Let's dig into network effects since you opened with that. There are different types, and not all of them are equally powerful. Walk me through how you categorize them when you're assessing a business.

EDGE: Direct network effects are the strongest. That's where my usage directly increases value for you. Think telephone networks, or more recently, payment platforms like Venmo. If I can't send you money because you're not on the platform, the platform is useless to me. Each person who joins makes it more useful for everyone else in a very literal sense. Then you have indirect network effects, which are more common but often weaker. That's marketplace dynamics. Uber gets more drivers because there are more riders, gets more riders because there are more drivers. Sounds powerful, but it's vulnerable because the network is local. Uber dominating San Francisco doesn't help them in Singapore. Someone can build a competitor market by market. We saw that with Grab and Gojek in Southeast Asia. Finally, there are data network effects, which people overestimate. The idea is that more usage generates more data which improves the product which attracts more users. Google Search is the real example. But most companies that claim data network effects are just doing basic machine learning improvements that plateau quickly.

That's a critical distinction because I see data network effects claimed constantly, especially in anything touching AI. It sounds sophisticated but the practical advantage disappears faster than people expect. Let's shift to switching costs because that's another moat that gets misunderstood. What makes switching costs actually prohibitive versus just annoying?

EDGE: Real switching costs have to be structural, not just friction. Friction is having to reset your password or re-enter your credit card. That's annoying but it takes five minutes. Structural switching costs are when moving to a competitor requires rebuilding workflows, retraining staff, risking compliance failures, or losing proprietary data configurations. I'll give you a concrete example. Bloomberg Terminal has extraordinary switching costs. It's not the $24,000 annual fee that locks people in. It's that traders have muscle memory on the keyboard shortcuts, the data feeds integrate with their execution systems, their compliance is built around Bloomberg's audit trails, and everyone they need to message is on Bloomberg chat. Switching to a competitor means operational risk during the transition, retraining costs, and losing connectivity to counterparties. That's a real moat. Compare that to switching from Spotify to Apple Music. You lose your playlists, which is annoying, but you can rebuild them in a weekend. The music is the same. That's not a moat.

And the Bloomberg example highlights something important, which is that switching costs often tie into installed base and ecosystem. The value isn't just the product, it's everything built around it. But I want to push you on something. We've seen technology disrupt businesses with apparent switching costs before. What's the difference between switching costs that hold and ones that evaporate when a better product comes along?

EDGE: The question is whether the switching cost is tied to value creation or value extraction. If your switching costs are high because you've deeply integrated into a customer's workflow and you're genuinely making them more productive, that's durable. If your switching costs are high because you've deliberately made it painful to leave while not innovating on the core product, you're vulnerable to disruption. Oracle databases in the 2000s looked unassailable. Enterprises had entire systems built on Oracle, consultants specialized in it, certifications, the whole ecosystem. But the switching costs were tied to lock-in, not continuous value improvement. When cloud databases offered genuinely better economics and capabilities, companies started the painful migration because the math worked. It took years, but it happened. Contrast that with something like TSMC. The switching costs for Apple or Nvidia to move their chip production away from TSMC are enormous, but those costs exist because TSMC keeps pushing the frontier of process technology. They're not resting on lock-in. They're maintaining the lead.

That's the key. The moat has to be defended with continuous improvement, not just contractual complexity. Let's turn to cost advantages because this is the one that sounds most straightforward but I think has the most subtlety. What actually creates a sustainable cost advantage?

EDGE: Scale economies in the right part of the cost structure. Not all scale advantages persist. If your cost advantage comes from buying ads more efficiently because you have a bigger budget, that's not durable. Competitors can match that. The question is whether you have a fixed cost that can be amortized over a larger base in a way competitors cannot replicate. Costco is textbook. They make almost all their profit from membership fees, which means they can price goods at essentially cost plus a tiny margin. That pulls in more members, which gives them more volume, which gives them better supplier terms, which lets them lower prices further. A competitor starting from zero cannot offer those prices because they don't have the volume to get the supplier terms and they don't have the membership base to fund operations. The scale advantage is structural. Netflix in the streaming era had this for a while. They could spend three billion on content and spread it over 200 million subscribers. A new entrant spending the same three billion over ten million subscribers has wildly different unit economics. But that moat eroded when other scaled players like Disney entered with their own large subscriber bases.

And that brings us full circle. Moats are not permanent. They're conditional on the competitive landscape and the pace of technological change. When you're looking at a business today and trying to assess whether the moat is durable or temporary, is there a single question you find most clarifying?

EDGE: I ask whether the advantage gets stronger or weaker as the market grows. A real moat benefits from market growth because it reinforces the gap between the leader and followers. Network effects get stronger with market growth. Switching costs deepen as customers build more on top of the platform. Scale advantages widen as the leader captures disproportionate share. If the advantage weakens as the market grows, because new entrants can come in with fresh technology or because customers have more options, then it's temporary. That's the test. Does success breed more success, or does it invite competition that erodes the edge?

See you Monday. The best moats are the ones that turn your growth into your competitors' problem.