CompanyGraph
How to tell whether a competitive advantage strengthens through use or erodes as its original conditions change.
A moat has a direction
“Moat” is investor shorthand for an advantage that makes imitation, entry, or substitution difficult. It is not a physical measurement, an accounting line, or a guarantee of future excess returns. Two businesses can look equally protected today while moving in opposite directions: one may be accumulating participants, data, or know-how; the other may be spending more each year to preserve a position that technology or regulation is undermining.
The useful question is therefore not how wide the moat appears in one period. It is what happens when the business operates. Does success add something that makes the next customer easier to win or retain? Or does each year consume a finite protection that must be replaced? The answer requires a mechanism and observations over time, not a label.
Compounding requires a feedback loop
A compounding advantage has an output that becomes an input to the next round. A marketplace can become more useful as buyers attract sellers and sellers attract buyers. The effect is not automatic: liquidity can remain thin, participants can multi-home, and a rival can subsidize entry. The FTC describes these as indirect network effects, where participation on one side changes value on the other. That definition identifies a mechanism; it does not prove that a particular platform has durable market power.
Data can also reinforce a product when additional use improves a model, search result, fraud screen, or workflow in ways customers notice. Learning by doing can lower defect rates or shorten implementation time. Scale can spread fixed engineering, distribution, or compliance work across more output. Each case still needs a causal link: the new data must be usable, the learning must be retained, and the cost reduction must not be competed away.
Protection can be real and still expire
A patent or regulatory license can block direct imitation for a period without creating a self-reinforcing loop. The protection may be valuable while it lasts, but the business must replace it with new products, renewals, or another advantage. FDA explains that patent terms and exclusivity periods balance innovation incentives with later generic competition. The legal clock is observable; the commercial outcome still depends on manufacturing, clinical substitution, supply, and prescriber behaviour.
Legacy infrastructure can produce a cost advantage when it is already paid for and well located. It can decay when a new process removes the need for that infrastructure, when maintenance grows faster than output, or when a rival can build a more efficient system. A scarce license can weaken after rules change. A brand can lose force when service, quality, or distribution no longer support the promise. These are decay processes, not proof that the original advantage was imaginary.
One business can contain both trajectories
A software company may benefit from accumulated integrations and customer data while its original technical architecture becomes harder to maintain. A pharmaceutical company can have a durable manufacturing and regulatory capability while individual product exclusivities move toward expiry. A retailer can gain purchasing scale while a physical store network becomes less suitable for where customers shop. The overall moat is not a single score; it is a portfolio of mechanisms with different clocks.
Management can add a new compounding mechanism, but spending more on a decaying one does not automatically make it compound. A maintenance budget may preserve availability without increasing switching costs. A marketing campaign may protect awareness without improving the product. An acquisition may add a network or capability, or may simply buy revenue that leaves when the price or service changes.
What evidence can distinguish the two?
Look for changes that would be difficult to explain with spending alone. Are retention, conversion, defect rates, or unit costs improving as the installed base grows? Do customers bring additional workflows, suppliers, or users because the existing system is more useful? Does the company retain the benefit in prices and cash generation, or do competitors pass it through?
For a potentially decaying moat, identify the outside clock. When does a patent expire? What new process, standard, supplier, or regulation changes the entry condition? Is maintenance spending rising faster than the revenue or capacity it protects? Are customers accepting more migration risk because alternatives have become compatible enough?
Reported metrics only partially answer these questions. A higher gross margin can reflect price, mix, temporary input costs, or accounting classification. A larger user count can conceal low activity, subsidies, or multi-homing. A stable market share can coexist with a weakening product if the whole market is shrinking. Compare operational evidence with claims in filings, customer behaviour, and the cash required to sustain the position.
Why a wide moat can still be a bad investment
Trajectory is not the only issue. A compounding advantage can be overpaid for, regulated, or attached to a market that stops growing. A decaying advantage can be cheap enough to produce a return if its cash flows arrive before the protection disappears. The concept helps describe the direction of an operating mechanism; it does not determine valuation by itself.
Nor does “compounding” mean permanent. Network effects can reverse when users leave, data can become stale, and learning can be copied. The relevant test is whether the reinforcing process is still operating and whether its gains exceed the forces eroding it.
Tests for moat trajectory
- Name the loop: specify what each additional customer, unit, data point, or production cycle changes for the next one.
- Name the clock: identify expiry, substitution, maintenance, regulation, or technology that can reduce the advantage.
- Separate stock from flow: distinguish accumulated assets such as installed integrations from current spending needed to keep them usable.
- Check who captures the gain: determine whether lower cost, better service, or higher willingness to pay remains with the company or is competed away.
- Test alternatives: observe whether customers can multi-home, migrate, qualify a rival, or perform the task internally within their actual time and budget.
A moat compounds when operations create more of the conditions that sustain the advantage. It decays when the original protection is consumed faster than the business replaces it. That distinction is a hypothesis to test against time, alternatives, and cash—not a certificate attached to a business.