Time
30–50 min
People
4–8 people

Overview

A persona the team builds from what it currently believes, rather than from finished research. The Kairos deck lists it as a framing methodology and defines it as creating preliminary personas grounded in assumptions, to frame user motivations, needs and behaviours.

The “proto” is the important part: it is an explicit set of assumptions about who you are designing for, written down so they can be tested rather than absorbed.

  • When to use it.

    During framing, once there is some research signal but before it is fully synthesised. Also useful when a team has been designing for a user nobody has ever described out loud.

  • Who to involve.

    Four to eight, deliberately mixed — design, product, and at least one person who talks to customers. Support and implementation colleagues are usually the most accurate source in the room.

  • Effort.

    Low prepMedium engagement

    Little preparation beyond gathering whatever signal already exists.

  • Outcomes.

    One to three protopersonas, with every claim marked as known or assumed. The assumed ones become the research questions — which is the whole reason for doing this before research rather than after. It also feeds Scenario mapping, which pins a persona above the scenario.

  • Recommendations.

    Do not add a stock photo: it makes an assumption look like a finding. Cap it at three personas — if the team cannot say why two segments behave differently, they are one segment. And revisit it after research; a protopersona that never got corrected was probably never used.

How to run it

  1. Agree the segment first

    As a group, agree how many meaningfully different users you are designing for — one protopersona each, and no more than three.

  2. Fill in behaviour before demographics

    Before anything else, fill in what they are responsible for, what a bad day looks like, what they use today and what they work around — age and job title tell you almost nothing about an industrial software user.

  3. Mark each claim as known or assumed

    Go back through every claim on the board and colour-code it: known in one colour, assumed in another.

  4. Name the pressure they are under

    Who is chasing them, and for what number. Most ERP behaviour that looks irrational is rational once you know this.

  5. Book the check

    Agree when the assumed claims get tested, and who does it.