Mind your moats

“In business, I look for economic castles protected by unbreachable ‘moats’.”
-Warren Buffett

In the press and in VC jargon, the two most common ways to taxonomize startups startups are by sector (e.g. EdTech, FinTech, AdTech, BioTech ), and by business model (e.g. marketplace, SaaS, on demand).

However, I think there is another categorization that we don’t think or talk enough about. That is economic moats.

By a “moat”, I mean a significant barrier to entering a market. Below, I’ve taken a shot at listing the major moats out there, in descending order of strength:

1. Network effects. This is the condition under which adding one user to your business improves the value for all users. Examples: telephones, eBay, Facebook, the Apple App Store, AirBnB. This is the mother of all moats. [1]
2. Technology. The invention of proprietary hardware, software, processes or mechanical products. Once developed, patents can make this moat even deeper and more expensive to cross. Examples: SpaceX, Google’s search algo.
3. Cost advantage. The ability to offer the same product as your competitors at lower cost. There are many ways of attaining cost advantage, including economies of scale and vertical integration. (The “full stack startup” movement capitalizes on this moat.) Examples: Walmart, Amazon, Warby Parker, Netflix.
4. Regulation. Navigating or erecting legal barriers, either by finding workarounds, or by getting the law changed. Many established industries have built colossal legal moats (see: agriculture, pharma, taxis, etc.) New business models can find ways around these; but with the caveat that “a rising tide lifts all boats”, and whatever technique is used to surmount a law can often be used by fast followers. Examples: Oscar Health, Simple.
5. Proprietary deal flow + high switching costs. A way of getting new customers your competition can’t, and making it unpleasant for them to leave you. (The easiest way to get proprietary deal flow is by being first to market.) Either one of these things by itself is not a moat; together they are, even for more-or-less commodity products. Examples: AWS, Quickbooks. [2]
6. Capital. Sometimes you just can’t start a business without a huge amount of upfront investment in infrastructure. Examples: Tesla, Jet, The Minerva Project.
7. User experience/brand. You can provide a similar product or service as your competition, but use a more pleasant interface or aspirational branding. This is often not a very deep moat, and is a last resort to differentiate a product that is otherwise a commodity. Example: Coca-Cola, Virgin America.

There are four main reasons I think moats are important to include in any discussion or analysis of startups:

The companies with the most—and deepest—moats win the biggest. The best companies, like Apple, enjoy literally all of the moats listed above. You can have a great technology moat, but if your competitor gets a cost advantage on you, you can still lose. As companies grow, they tend to need more moats to keep them safe.

Moats have two sides—and which side of the moat a company is on matters a lot in assessing its riskiness. Companies outside of a deep moat are in mortal peril; companies inside of them are relatively safe. If you want to assess a startup’s chances of survival, ask yourself what moats they face and how far they are from crossing them.

Companies with the same moats are similar — often more so than even those in the same sector. Even if a startup is “unprecedented” in its vertical, the history of other companies facing the same moats may be instructive. If AWS keeps gaining market share but Box is losing it, what is Amazon doing right that Box could learn from? IBM’s technology moat eroded over time, so how might SpaceX’s avoid the same plight?

Adding moats is a great way to build a business—especially startups. There are plenty of businesses out there with no moats at all; they’re just highly fragmented. Hence, you get industries like construction contracting, laundromats, real estate brokerage and plain ol’ commodities. “Rolling up” an industry by adding moats is a phenomenal way to make money. For example, SolarCity rolled up the solar installation and leasing industry by inventing an awesome cost advantage.

The concept of economic moats is well understood and has been around for a long time. Still, it’s an extremely powerful—and, in my opinion, under-used—conceptual framework.

[1] This explains why a wretched business like Craigslist is still around.

[2] An important outcome of this moat is that it results in international land grabs. Companies like Rocket Internet can clone existing services and build them out in new territories before the incumbent gets there.