AI made code cheap.
We price what it can’t copy.

Revenue multiples can’t price an asset with no revenue. Rebuild cost used to be the answer — until AI started deflating it every quarter. The ADR measures the part of an asset a competitor cannot reach, then shows you how much of today’s value is still standing in five years.

Score your asset — free

Two minutes, twelve questions, no account.

The 90-Day Test

Could a competent competitor, with unlimited AI tooling and $50,000, reach this exact position within 90 days?

If yes, the moat is close to zero — however sophisticated the codebase. If no, something specific is blocking them, and that blocker is what the report prices. Note the test is not “could this be built.” Almost anything can be built. The question is whether the position can be reached.

Only four blockers survive that test

Each scores 0–25. Together they produce a Moat Score out of 100.

Time Moat
0–25 pts
Calendar months that cannot be compressed

Domain age, backlink profile, review history, months genuinely live with users. The purest moat there is, because it is non-negotiable — no amount of capital or compute ages a domain.

Domain ageReferring domainsReview and rating historyOperating history
Permission Moat
0–25 pts
A third party has to say yes

The integration takes an afternoon. The approval takes nine months and may never arrive. Regulatory licences, restricted API tiers, compliance certifications, gated app-store categories.

Regulatory licencesRestricted partner tiersSOC 2 / HIPAA / PCI / ISOGated marketplace approval
Data Moat
0–25 pts
Records that only exist because it ran

Volume, uniqueness, and whether the data measurably improves the product. With one qualifier most methodologies ignore entirely: data that cannot legally transfer to a buyer scores negative, because they inherit a liability rather than an asset.

Proprietary record volumeUniqueness vs scrapableFlywheel effectLegal transferability
Embedment Moat
0–25 pts
Users and integrations already depend on it

Active usage with genuine switching cost, third parties who have built against the API, contracted commitments. Registrations are not embedment — usage is. Distribution is the hardest thing on earth to rebuild.

Active usersThird-party integrationsContract commitmentSwitching cost

Score your asset against those four axes

Free and self-assessed. You’ll get the score, the per-axis breakdown, and the three changes that would move it most.

Twelve questions, about two minutes. Your full score and improvement plan appear on this page the moment you submit — we ask for an email so you have a copy to keep.
What's it called? · optional
Only used to label your result and your email copy.
Category · optional
Lets us compare you against verified assets in the same space.
Time moat
Things that take calendar months and can't be rushed
Domain age (years)
How long the web address has been registered — not the company's age. Check with a WHOIS lookup.
Months live with users
Since real users started using it, not since you started building.
Total reviews
How many, across all platforms — not the star rating.
Permission moat
Things a third party had to approve — the code takes a day, the approval takes months
Regulatory licence
A licence a regulator granted you — money transmitter, broker-dealer, lending, medical. A business registration or your own terms of service doesn't count.
Compliance certification
Audited certifications: SOC 2, ISO 27001, HIPAA, PCI DSS. Self-attestation isn't one — it needs an auditor's report.
Partner / API tier
Access a platform had to approve you for. “Public tier” is anything anyone gets with an API key; “restricted” means you applied and were accepted.
Data moat
Records that exist only because your product actually ran
Proprietary records
Rows your product generated by actually running. Seed data, scraped data, and anything you bought don't count.
How unique is it?
Could a competitor assemble something equivalent from public sources or by buying a dataset? If so it's public or scrapable, however much of it you hold.
Does your data legally transfer to a buyer?
Check the change-of-control clause in your privacy policy. Most people haven't — “not sure” is the honest answer.
Embedment moat
How stuck your users and integrations already are
Active users
People who used it in the last 30 days. Registered accounts that never came back aren't users.
Third parties integrated with you
Others who built against your API or connected their systems to you — not services you consume. Your Stripe integration is you depending on Stripe, which is the opposite of a moat.
Customer commitment
What your customers are contractually on today. Free users and trials are “no contracts.”
Where should we send your copy?

One email with your results, and at most one follow-up. No sequence, no sharing.

Free, no account. Self-assessed — verification comes with listing.

The payoff

Same score today. Very different assets.

The DX Score tells you whether an asset is good today. The Moat Score determines how much of that survives. Combined, they produce a five-year durability projection.

Licensed fintech
6yr domain · 2.4M proprietary records · 7,200 actives
Moat 93/100
78 → 71
Fortified
Vertical SaaS
3yr domain · annual contracts · 4 integrations
Moat 54/100
78 → 50
Softening
Thin API wrapper
New domain · public data · 12 users
Moat 9/100
78 → 35
Eroding

All three score DX 78 today. On a scoreboard that reports one number they are indistinguishable — and a buyer would price them identically. They should not be priced identically.

Commodity Risk

The number sellers don’t enjoy

Every ADR publishes the inverse of the moat score: what share of the asset is simply reproducible code. A mainstream stack with no novel algorithm scores high. A trained model or genuinely proprietary algorithm scores low.

We publish it anyway, because a valuation you can only trust when it flatters the seller isn’t a valuation. It also gives sellers a concrete reason to build real moats before listing — which improves what buyers see on the board.

The range is not a price

The ADR range answers one question: how much of the asking price is backed by something a competitor cannot cheaply reproduce?

That is not the same as what the asset is worth. An asset can be worth considerably more than its range — most good ones are. The range is the part a buyer is acquiring rather than betting on.

A worked example

A SaaS product built largely with AI tooling. Three live surfaces, five capabilities, a month of build time, roughly a thousand dollars of actual spend. No proprietary data, no integrations, no clearances, no revenue yet. The seller asks $100,000 on the strength of their revenue projections.

The ADR comes back around $3,000 – $15,000, with a Moat Score near zero.

That is not us saying the asset isn’t worth $100,000. It may well be. It is us saying that roughly $14,000 of that price is backed by position, and the remaining $86,000 is a bet on execution the buyer is being asked to take. A competitor with the same tooling can reach that position next quarter for the same thousand dollars. Nothing is stopping them — so nothing is protecting the price.

Sellers with a real moat get the opposite result, and it argues in their favour. When years of accumulated data, a certified partner tier or a regulatory registration sit underneath the price, the range rises to meet it and the buyer can see exactly what they are paying for. That is the whole point of separating the floor from the bet: it rewards the assets that earned it.

We don’t price your projections

Forward revenue, expected users, pipeline, the model in the deck — none of it enters the range. That is deliberate, and it is worth being direct about why.

Projections are the most inflatable input in any marketplace. Rebuild cost was already too easy to move, and it at least referenced money that actually left a bank account. A projection references nothing. The moment a number moves the valuation, it stops describing the asset and starts describing the seller’s optimism — and a number every seller can move is a number no buyer can use.

There is a second reason, and it matters more. The ADR is a durability instrument. It measures what survives contact with a competitor who has the same AI tooling and a year to catch up. Forward revenue is a different question, answered by a different method, and folding the two together produces a figure that means neither thing.

So the report states the range and what it is anchored on, and leaves the asking price where it belongs — as your claim, argued on its own terms, next to a number that was not. We would rather hand a buyer two honest figures than one blended one.

What’s in the report

Eleven sections. Delivered as a shareable page and a PDF you can hand to a buyer, a board, or an accountant.

Moat Score & four-axis breakdown
Per-axis evidence and confidence level
Five-year durability projection
Where the asset lands, and why
Commodity Risk
How much is reproducible code
DX Score breakdown
Time-to-Position, IP, market, traction
Capability breakdown
Build, research, permission & traction capital
ADR valuation range
Capability-cost anchor, moat-weighted, floor / mid / ceiling
Moat improvement plan
Ranked actions to raise the score before sale
SWOT analysis
Written against the actual findings
Deal structure fit
Which of the five structures suits this asset
Buyer guidance
Open, target, and walk-away numbers
Moat EstimateVerified Moat Score
CostFree$199 – $699
Based onYour answersEvidence, reviewed by a Deal Manager
Confidence levelAttestedDocumented / verified
Repository scanBuild Integrity Report
Shows on your listingNoYes
Buyers can price against itNoYes
Verification comes with listing

We check the licences, read the change-of-control clause, scan the repository, and a Deal Manager reviews the evidence. Full ADR report fee: $199 (under $50K asset), $349 ($50K–$250K), or $699 ($250K+).

List your assetRead the methodology
What the ADR is not

It is analytical guidance, not a legal valuation, and the durability projection is a model output rather than a forecast. Our multiplier bands are reasoned rather than empirically calibrated. We tune them against assets we have scored in full and, as deals close, against what they actually sold for — and we will say publicly when each of those moves. Every report states its own confidence level, and an axis resting on seller attestation is labelled as such rather than presented as verified.

A Moat Estimate is weaker still — it is your own answers scored against the same rubric, with nothing checked. We publish no valuation range against an estimate for that reason: a dollar figure built on unverified inputs is the one output that could genuinely mislead you about your own asset.