A Moat Taxonomy for Private-Market Diligence

Jagdeep Singh · Jagdeep Ventures · Last revised July 2026

"Defensibility" appears in every deck and means something in perhaps one in five. The problem is not dishonesty; it is that moat claims are rarely stated in a form evidence can test. This framework restates the five moats that matter in private-market diligence as falsifiable claims, with the evidence that supports each and the indicators that show one eroding.

1. Network effects

The claim: each additional user makes the product more valuable to other users, so scale compounds into advantage.

The test: cohort-level engagement or retention should improve as network density grows, measurably, in the data, not in the narrative. Ask for retention by cohort against network size at join time. Distinguish true network effects from mere scale: a large customer list is not a network unless customers derive value from each other. Watch for the ceiling: many networks saturate locally (by city, category, or workflow), and the marginal-user value curve flattens far earlier than the deck implies.

Erosion indicators: multi-homing rising in the user base; the network's dense clusters churning together; an entrant achieving critical mass in one segment.

2. Switching costs

The claim: leaving is expensive, in money, risk, data, or retraining, so customers stay and pricing power accrues.

The test: arithmetic, not adjectives. What does a real migration cost, in hours and dollars, for a typical customer? Evidence: renewal rates at price increases (the cleanest test, pricing power is switching cost made visible), depth of integration (workflows and data embedded, not just logins), and expansion behavior. Beware switching costs that bind the champion but not the budget holder.

Erosion indicators: migration tooling appearing in competitor marketing; standardization of data formats in the category; renewal negotiations lengthening.

3. Scale economies

The claim: unit costs fall with volume in a way sub-scale competitors cannot match.

The test: the cost curve itself, unit economics by volume tier over time. In software, claimed scale economies are often just fixed-cost amortization every competitor also enjoys; the durable version requires a cost driver rivals cannot access at equal terms (proprietary supply, data that improves the product, distribution density). Ask what specifically gets cheaper, and why a well-funded entrant cannot buy the same curve.

Erosion indicators: the cost advantage narrowing in successive pricing rounds; input suppliers consolidating; a platform shift resetting the fixed-cost base.

4. Brand

The claim: the name itself changes buyer behavior, commanding a premium or shortcutting trust in high-stakes choices.

The test: price-realization data against functionally comparable alternatives, win rates in competitive processes the brand should not win on features alone, and unaided recall in the buying population. Brand moats are real but rarer than marketing spend suggests; most "brand" in B2B is a current performance reputation that depreciates quickly when service slips.

Erosion indicators: discounting frequency rising; win rates normalizing in head-to-heads; the premium surviving only in legacy segments.

5. Regulatory position

The claim: licenses, compliance infrastructure, or approved status create a barrier competitors must spend years and capital to cross.

The test: how long and how expensive is the barrier for a determined, funded entrant, actually priced out, not gestured at. Regulatory moats are strong while they hold and binary when they break; diligence should include the docket, not just the license: pending rule changes can erase the moat or, occasionally, hand it to someone else.

Erosion indicators: deregulatory proposals in consultation; incumbents lobbying defensively; compliance-as-a-service vendors packaging the barrier.

Using the taxonomy

Three habits make this operational. Demand the mechanism: every moat claim should name which of the five it is and what evidence would falsify it; claims that resist classification are usually execution speed wearing a costume, valuable, but a different underwriting. Weight by durability, not count: one deep moat with clean evidence beats four shallow ones listed on a slide. Monitor the erosion indicators, not the moat: moats fail gradually in the indicators before they fail suddenly in the P&L, which is why our diligence reports ship with a watchlist rather than a verdict alone.

None of these claims is provable in advance, defensibility is a forecast, and forecasts earn confidence only through structure. The taxonomy does not remove the judgment; it gives the judgment something to grip.