Any conversation with more than one mind in it.
The machinery does not know whether it is running a focus group or a sales call. It knows there is a cast, a context for each of them, a running order, and whether you are in the room.
Research
You watch-
Focus groups and concept tests
Put a concept or a price in front of two to eight participants. Private first reactions, then a moderated discussion, then a private rapid-fire round you can compare against the first.
- Cast
- Buyers across the segments you care about, each with their own history with the category.
- The part that matters
- Private first. You see where each participant started before anyone else in the room had spoken.
-
Positioning tests
Three ways of saying the same thing, put to a room that has never heard any of them, with the order randomized per participant.
- Cast
- One buyer who already shops for this, one who has never considered it, and the range in between.
- The part that matters
- First reaction. Positioning is judged in seconds, and only the private round catches that.
-
Design critique
Practitioners with different tastes and different axes to grind, each reacting to the same screen without seeing the others’ notes first.
- Cast
- Five designers with incompatible convictions about density, motion and color.
- The part that matters
- Separate context. It separates taste from usability, which a live critique rarely does.
Rehearse
You take part-
Interview prep
Practice with the panel you will actually face. Each interviewer listens from their own seat and asks what they would ask, then tells you afterwards how your answers landed with them.
- Cast
- The hiring manager, a future peer who wanted someone more senior, the skip-level who asks about scope.
- The part that matters
- The debrief. Interviewers rarely agree on how an answer landed, and you get each view separately.
-
Sales call simulation
Run the call against a buying committee before you run it for real. Objections come from the participant who would raise them, in the order they would raise them.
- Cast
- The CFO, the security lead, the daily user, procurement with a competitor’s quote.
- The part that matters
- Separate knowledge. The security lead has read the SOC 2 report and the CFO has not, so each objection comes from what that participant knows.
-
Meeting rehearsal
Run the meeting before the meeting. Present, take the questions, and see who you won over and who you lost.
- Cast
- The CFO who asks about margin, the lawyer who asks about consent, the head of sales who asks what changes on Monday.
- The part that matters
- Private reactions. Each stakeholder’s note on your answer, written before they reply to it.
Ideate
Take part or watch-
Brainstorm
Private ideation first, so the first good idea does not anchor the room, then a group round that builds on what was generated independently. Add your own ideas or stay out and read.
- Cast
- Deliberately mixed: two participants close to the problem, three who have never seen it.
- The part that matters
- The order. Group-first brainstorms converge on whoever spoke first.
-
Pre-mortem
Run the failure before it happens. Cast participants in the roles that would be blamed, the roles that warned you and the one that signed it off, then ask each of them privately what went wrong.
- Cast
- The sponsor, the engineer who flagged the risk, the support lead who takes the calls, the skeptic.
- The part that matters
- Private first. In a real pre-mortem the most senior voice sets the failure mode and everyone else elaborates on it.
-
Red-teaming a plan
Participants briefed to find the failure rather than to be helpful. Each attacks independently, then they see each other’s attacks.
- Cast
- A regulator, a competitor, a journalist, a disgruntled former employee.
- The part that matters
- Isolation. One loud skeptic otherwise sets the tone and the rest pile on to the same weakness.
What a concept test costs you today
- $6k–15krecruitment, incentives, facility, moderator
- 3 weeksscreening, scheduling, fieldwork, write-up
- 1 runchange the stimulus and you start over
- Per studyno recruitment, no incentives, no facility, no moderator day rate
- ~25 minsetup to readable report
- Re-runsame cast, changed stimulus, comparable row by row
The comparison is about speed and cost, not equivalence. A simulated session is not a substitute for talking to your actual customers. It is what you do before you have earned their time, and how you decide what to ask them when you do.