Competitive advantage and moats: network effects, switching costs, scale, brand

Plenty of businesses make good profits for a few years. Far fewer keep doing it once competitors notice. The difference is usually a competitive advantage that rivals find hard to copy or attack, which investors call a “moat”, after the water that protects a castle.

This guide explains the main sources of moats (network effects, switching costs, scale and brand along with other intangible assets), how to test whether your business has one, and the mistakes people make when they claim one.

What a moat is

The term is commonly credited to Warren Buffett, who used it repeatedly in Berkshire Hathaway’s shareholder letters, according to Wikipedia. The investment research firm Morningstar rates companies on it and defines the idea simply: “A company with an economic moat can fend off competition and earn high returns on capital for many years to come.”

Morningstar uses three ratings. A wide moat should hold for 20 years or more. A narrow moat should keep the company competitive for at least 10 years. No moat means no lasting advantage. It groups moat sources into five types: cost advantage, intangible assets, network effects, switching costs and efficient scale. A company can have more than one, which is generally better, because individual sources can weaken over time.

Network effects

A network effect exists when a product becomes more valuable to each user as more people use it, as described on Wikipedia. A telephone was almost useless when few people had one. Direct network effects come from users of the same product, as with messaging apps. Indirect, or two-sided, effects come from linked groups, such as buyers and sellers on a marketplace, or hardware owners and software developers.

Network effects can be very strong, but they have conditions. A market only tips towards one winner when network benefits outweigh differences between products, and when using several competing services at once (multi-homing) and switching are costly. If those conditions fail, several providers can coexist. Networks can also suffer from congestion, where each extra user makes the service worse for others.

Andrei Hagiu and Simon Rothman make a related point in Harvard Business Review: marketplaces are “highly defensible once established, owing to network effects”, but reaching critical mass is only the first hurdle. Their pitfalls include growing too fast too early, failing to build trust and safety, and relying on penalties rather than incentives to stop users dealing directly with each other off the platform.

Switching costs

Switching costs exist when leaving is expensive, risky or inconvenient for the customer. Morningstar’s example is business software, where retraining staff and reconnecting systems make a change costly. Banking is another common case.

Switching costs come in several forms: money (exit fees, new equipment), time (setup, learning), risk (disruption, lost data) and relationships (a trusted account manager). The Harvard Institute for Strategy and Competitiveness notes that buyers have most power when switching costs are low, so raising them, fairly and by adding value, reduces buyer power. Lock-in that customers resent can invite complaints, regulation and rivals who promise easy exits.

Scale and cost advantage

A cost advantage lets a company sell at prices rivals cannot match profitably, or earn higher margins at the same price. Scale is a common source: bigger buyers get better purchase terms and spread fixed costs such as advertising and distribution over more units. Morningstar gives McDonald’s as an example of using size to buy food and packaging more cheaply than smaller rivals. Cost advantage can also come from a better process, location or access to cheaper inputs.

Efficient scale is a related idea. Some markets can only support a few profitable firms, so new entrants are deterred because they would push everyone’s returns down. Morningstar cites railways and utilities. In this case the moat is the size of the market relative to the cost of serving it, not the company’s own brilliance.

Brand and other intangible assets

Intangible assets include brands, patents, proprietary technology and licences. Morningstar cites Starbucks as a brand that supports premium pricing, and notes that patents can give a healthcare company sole-producer status for up to about 20 years.

Brand is the most over-claimed moat. Morningstar states that a well-known brand or long history does not on its own create a moat, and gives United Airlines and Ford as examples of famous names it does not rate as moat companies. The test is whether the brand lets you charge more or sell more than a rival with an equal product. Recognition alone does not count.

Testing your own moat step by step

1. Look at results. Has the business earned returns above its cost of capital, or better margins than peers, for several years? No sustained excess returns usually means no moat yet.
2. Name the source. Which of the five sources explains those returns? Be specific about the mechanism.
3. Imagine a well-funded rival. If a competitor with plenty of money copied your product tomorrow, what would stop your customers leaving?
4. Check the evidence. Customer retention, price premiums, unit costs against peers and share of repeat business all show whether the moat is real.
5. Estimate how long it lasts. What could erode it: new technology, regulation, a change in customer habits?
6. Decide how to widen it. Pick one source and invest in it deliberately.

A worked illustration

An example (illustration only). A UK software firm sells booking and payment tools to independent physiotherapy clinics. Its margins have been above peers’ for four years. The owners test the possible sources:

• Switching costs: strong. Clinics store years of patient notes in the system and train reception staff on it.
• Network effects: weak. Patients book with one clinic and gain little from other clinics using the same system.
• Scale: modest. A larger rival has lower costs per customer.
• Brand: known within the profession, but clinics choose mainly on features and price.

The conclusion is a narrow moat built on switching costs. To widen it fairly, the firm adds features clinics value, such as an easy data export (which builds trust) and integrations with insurers (which add value and depth). It does not rely on trapping customers.

Common mistakes

• Confusing a good product with a moat. Good products get copied. A moat is what stops the copying working.
• Calling recognition a brand moat. Without pricing power, it is just awareness.
• Assuming network effects are permanent. Multi-homing, trust problems and congestion can all weaken them.
• Relying on one source. Sources can erode; two that reinforce each other are safer.
• Building lock-in customers hate. It can invite regulatory attention and churn as soon as an alternative appears.

The moat lens is useful for investors comparing companies, for founders deciding what to build beyond the first product, and for managers deciding where to invest. Pair it with a five forces analysis, since barriers to entry and moats are two views of the same question.