Skip to content
RAY BEHARRY

Agent 4 of 71 · Market Intelligence

Buyer Persona Architect

Keep personas alive instead of static PDFs. Build each persona from VoC evidence, segment by motivation rather than demographics, and record which proof types and channels move each role, so every message variant knows who it is for.

Agent contract

Use to build or refresh evidence-based buyer personas and buying committees (roles, motivations, objections, proof preferences, channels) inside an approved ICP.

When it runs

  • An ICP was approved or changed.
  • A new VoC synthesis shows a persona's priorities shifted.
  • Message Multiplier needs persona definitions that do not exist yet.

What it needs

InputRequiredFrom
Approved ICPYesICP Library
VoC synthesis and phrase bankYesMarket Signals
Existing persona cardsOptionalBuyer Personas
Win/loss notes by roleOptionalPerformance Observations

How it works

  1. Confirm the parent ICP is approved; refuse to build personas for a draft ICP.
  2. Map the buying committee: economic buyer, champion, technical evaluator, blockers, users.
  3. For each role, extract motivations, fears, success metrics and objections from VoC evidence with counts.
  4. Record proof preferences (metric, logo, third-party validation, peer quote) and channel habits per role.
  5. Write one persona card per role; mark any attribute without evidence as hypothesis.
  6. Diff against the previous persona version and note what changed and why.
  7. Store as 'proposed' in buyer-personas and request approval.

What it delivers

Persona card per role: motivations, success metrics, objections, proof preferences, channels, verbatim quotes, change log.

Output contract: persona-card

Quality bar

  • Motivation-based, not demographic. No invented names, ages or stock-photo details.
  • Each attribute is tagged as evidenced (with reference) or hypothesis.
  • Every persona lists its top three objections.

Hands off

When a person steps in

  • If evidence exists for fewer than half of a persona's attributes, publish as hypothesis and recommend interviews.

Example

Request

“Build the buying committee for the Regulated Mid-Market Finance ICP.”

Result

4 persona cards (CFO, Controller, IT Security, AP Manager). 81% evidence coverage. Controller's top objection: 'migration during close'. Handed to Message Kernel Architect.

Illustrative example from the package documentation, not a client result.