Ranked by time-to-neutralise
| Claimed moat | Time to neutralise | Verify by |
|---|---|---|
| Feature set | Weeks | Competitor changelogs |
| UX quality | Months | Head-to-head task timings |
| Proprietary data | 1-3 years, if genuinely exclusive | Contractual exclusivity terms |
| Regulatory licence | 1-3 years | Licence register entry and scope |
| Switching cost | Years, if embedded in workflow | Contract length and integration depth |
| Network effects | Years, if cross-side | Retention by cohort density |
Data moats are usually not moats
A data advantage defends only when the data is exclusive, replenishing and directly improves the product for the buyer. Logs that any competitor would accumulate at the same rate on their own customer base are not exclusive; they are a by-product.
Ask what specifically breaks for a competitor who starts today with zero data. If the honest answer is "our recommendations get slightly better sooner", the moat is a head start, and head starts expire.
Distribution is the underrated one
An owned channel — an existing customer base, an exclusive integration slot, a regulated intermediary relationship — is often more durable than any product property, and is far easier to verify. Contracts either exist or they do not.
- Named partners with signed agreements and terms
- Integration placement that is capacity-limited by the platform
- Customer acquisition cost falling as the channel matures
How to score it in a memo
Score defensibility on evidence type rather than on claim strength: contractual, structural, behavioural, or asserted. Only the first three belong in a weighted score; asserted defensibility belongs in the risk register as a dependency to re-check at the next round.
Part of a guide
Market sizing: TAM, SAM and SOM done properly
How to build market size from the bottom up: what belongs in TAM, SAM and SOM, how to calculate serviceable obtainable market from channel capacity, and how sizing interacts with defensibility.