Ask a founder what their moat is and you'll usually get a features list wearing a costume: "our UI is nicer," "we move faster," "our onboarding is white-glove." None of that is a moat. A moat is something a well-funded, competent competitor cannot simply build in a year if they decided to try. Most companies have never actually run that test against their own claims. It takes about an hour and one page.
List the candidates
Start with the usual six, defined concretely instead of abstractly:
- Switching costs — what it actually costs a customer to leave: data export friction, retraining, a multi-year contract term, integrations that would need rebuilding elsewhere.
- Accumulated customer data — not just storage, but data that compounds into value a competitor can't replicate on day one: benchmarks, personalization, or predictions that get measurably better the longer someone stays.
- Workflow embedment — whether you're the system of record people build daily habits around, or a tab that closes without anyone noticing for a week.
- Integrations surface — how many other systems depend on the connection, such that removing you breaks something else the customer relies on.
- Brand or category ownership — whether buyers use your name as shorthand for the category itself. Rare, and usually only true for whoever defined the category first.
- Distribution advantages — an audience, channel, or placement a competitor can't simply buy: a default install, a marketplace ranking, a partner network built over years.
Stress-test each one with a single question
For every candidate: what would it cost a well-funded competitor to neutralize this in twelve months? Not "is it hard" — what would it actually cost, in money and time, for a team with real funding and none of your legacy constraints?
Run the usual suspects through that question and most of them collapse fast. A nicer interface is a design sprint. More features is a roadmap and a quarter, and often a liability rather than an asset once support has to cover all of it. White-glove onboarding is a hiring plan, not a structural advantage — it costs headcount, and headcount is exactly what a funded competitor has.
What survives usually has one of two properties: it took years to accumulate and can't be compressed by spending more money faster (multi-year data that meaningfully changes output quality, a distribution channel built relationship by relationship), or it's a structural asymmetry that money doesn't fix on its own (contractual exclusivity, embedment so deep across a customer's other systems that removal breaks unrelated things, switching costs high enough that even a clearly better product loses to inertia).
Test each claim from the other side
The self-assessment is useful but biased — everyone rates their own moats generously. Check each claim against what competitors are actually doing, using the same research you'd run on any other competitive question. If you're claiming switching costs as a moat, look for a competitor's own migration guide aimed at your customers; if one exists and reads as straightforward, the switching cost is lower than you think. Scan review sites for people who describe actually leaving, and how painful it was in practice rather than in theory. A structured comparison template keeps this consistent across competitors instead of a fresh, ad hoc judgment call each time.
If a competitor is actively building tools or content aimed at moving your customers through a specific door, that door is weaker than assumed. That's a live test result, not a hypothetical one.
Keep the output to one page
The discipline is in the constraint. Three real moats, maximum — if the list runs to eight, none of them survived the stress test, and what you're holding is a feature list again. For each of the three, write one concrete action that deepens it: for workflow embedment, the specific integration that would make removal break something else too; for accumulated data, making the compounding value visible to the customer instead of just quietly felt, so they know exactly what they'd lose by leaving. Revisit the page when a competitor announces something that touches one of your three — that's the only trigger it needs. Canopy's competitor profiles pull together published pricing and reviews in one place per competitor, useful raw material for the "test it from their side" step without opening a dozen tabs.