World Model Readiness
Engraved foundations instrument

For your company

Foundations

Module · What do you know that rivals do not?

The Proprietary Moat Audit

The durable value in AI comes from teaching it something your competitors cannot buy off the shelf. That means a real knowledge advantage, captured as data, deepening over time, and instrumented so it keeps growing. This module tests whether your moat is real and compounding, or just a story you tell.

Question 1 of 5 · The advantage is real

Can you name something your company knows that your competitors do not?

Vague claims about experience or relationships do not count. A real advantage is specific and articulable: a failure mode you have seen ten thousand times, a process others get wrong, a dataset only you have accumulated.

Question 2 of 5 · Captured as data

Is that knowledge captured as data, or does it live only in people's heads?

Knowledge trapped in experts cannot train a model and walks out the door at retirement. Captured knowledge, decisions and outcomes and edge cases recorded over time, is the only kind AI can learn from and competitors cannot poach.

Question 3 of 5 · Deepening, not static

Is your knowledge advantage widening over time, or slowly being caught?

A static advantage is a depreciating one: competitors and vendors close gaps every year. A compounding advantage gets deeper the more you operate, because every case you handle adds to what only you know.

Question 4 of 5 · The source is instrumented

Do you instrument the process that generates your advantage, so it keeps producing data?

The advantage is not the data you have; it is the process that keeps producing it. If that process is not instrumented, the moat stops filling the moment your current data ages out.

Question 5 of 5 · Competitors cannot buy it

Could a competitor replicate your advantage by buying a tool or a dataset?

The test of a moat is whether money closes it. If a vendor sells the same model to your rivals, or the data is purchasable, what you have is table stakes, not a moat.

For the statistics · one click each

Three questions for the public picture

These do not affect your score. They feed the anonymised, aggregated statistics; groups under 8 respondents are never shown.

Could a competitor replicate your main advantage within a year?

Yes, easily
Yes, with effort
Only with difficulty
No
We are not sure

Where does your most valuable know-how live today?

In people's heads
In documents
In systems and data
Nowhere durable
We do not know

Is your competitive knowledge advantage widening or narrowing?

Widening
Holding steady
Narrowing
Already gone
We are not sure

Your context

Used to calibrate the report. Company size and sector remain in the anonymized dataset; your email does not.

What the five levels look like

Every dimension in this assessment is scored 1 to 5. This is what the levels mean, dimension by dimension. The graded report diagnoses where your own answers land and what to do about it.

The advantage is real

  1. 1Nothing specific
  2. 2Only slogans
  3. 3Vaguely sensed
  4. 4Clearly named
  5. 5Named and tested

At the low end: If you cannot name your edge specifically, you may not have one, or you have one nobody has noticed. Spend a session naming exactly what you know that a new competitor could not. What good looks like: A clearly named, tested advantage is the raw material of a real moat. Now the question is whether it lives anywhere but in your people's heads.

Captured as data

  1. 1In heads only
  2. 2Scattered notes
  3. 3Partly captured
  4. 4Mostly captured
  5. 5Captured by default

At the low end: Knowledge that lives only in people is a moat with a retirement date. Start recording the decisions and outcomes of your best experts now, in any form. What good looks like: Knowledge captured by default is what turns expertise into an asset AI can compound. Guard the capture habit; it decays the moment it feels optional.

Deepening, not static

  1. 1Eroding
  2. 2Standing still
  3. 3Holding steady
  4. 4Slowly widening
  5. 5Compounding fast

At the low end: An eroding advantage becomes a commodity on someone else's schedule. Find the one thing you could measure and improve every month, and start compounding it. What good looks like: A compounding advantage is the rarest and most valuable kind. Protect the mechanism that deepens it; that mechanism, not the data itself, is the real moat.

The source is instrumented

  1. 1Not instrumented
  2. 2Manual capture
  3. 3Some instrumentation
  4. 4Well instrumented
  5. 5Instrumented and improving

At the low end: An uninstrumented process produces a moat that drains as fast as it filled. Put measurement on the step that generates your edge, before optimising anything else. What good looks like: An instrumented, improving process is a moat that fills itself. Keep the instrumentation ahead of your growth, or volume will outrun your capture.

Competitors cannot buy it

  1. 1Anyone can buy it
  2. 2Mostly purchasable
  3. 3Partly defensible
  4. 4Hard to replicate
  5. 5Effectively unique

At the low end: If money buys your advantage, it is not yours for long. Find the part of what you know that no vendor sells, and concentrate your capture there. What good looks like: An advantage money cannot buy is the definition of a moat. Keep feeding the defensible core; that is where AI turns your knowledge into distance.