Who is this actually for?

• icp
• ideal customer
• who is it for
• segment
• target
• persona
• churn

• Which five {unit}s are the most profitable, and what do they have in common?
• Which five caused the most work, and what do they have in common?
• Which came through which channel?
• What did the good ones already have in place before buying?

• margin or effort per {unit} for the last twenty
• acquisition channel per {unit}
• retention by customer profile
• the trigger event that made each one buy

• Serving two incompatible customer profiles with one offer
• A channel that reliably delivers the wrong profile

• If
  The profitable customers share a trigger event
  Then
  Target the trigger, not the demographic.
• If
  The painful customers all came from one channel
  Then
  Fix the channel before fixing the offer.
• If
  There is no visible pattern
  Then
  There is not enough data yet. Twenty {unit}s is the minimum before drawing a profile.

Write the profile of the five best {unit}s as a one-page definition, then qualify the next ten enquiries against it and record the difference in {conversion}.

Read after 30 days.

• {conversion} is materially higher among enquiries that match the profile
• Effort per {unit} falls for matching customers

• Matching and non-matching enquiries convert the same — the profile is describing a coincidence
• Applying the profile leaves too few enquiries to run the business

• The profile predicts outcomes → Rewrite acquisition and qualification around it.
• It does not → Keep collecting until twenty {unit}s exist, and do not narrow acquisition on a guess.

Aury can carry out: read company, prepare document, create task, file memory. Specialists involved: research, sales, strategy.

Requires a real customer base — at minimum ten, ideally twenty completed relationships.

Early customers are biased by how they were found. A profile drawn from one channel describes that channel as much as the market.

• Sourced operator experience: Customer-development and account-review practice, widely used across services and software
• Aurygine synthesis: Aurygine synthesis across founder journeys

Confidence: high. Last reviewed 2026-08-25. Your company's own recorded results override this playbook whenever the two disagree.