Conversational research
Research people
can actually experience.
Most research asks participants to imagine a situation and report how they would feel inside it. The recall is shaky and the distance is real. A scenario interview closes that distance: the payment declines in front of them, the support agent is typing, the choice is theirs. You collect the decision while it is being made, with the follow-up attached to the exact moment that produced it.
The interactive builder ships on the Studio plan. The specimen needs no account.
A payment fails.a
The classic interview question "how would you feel if your payment failed" asks for imagination. This one asks for a decision. Answer it the way a participant would.
Meridian Bank
Your payment of $420 to Cloudline Hosting was declined by your bank.
18:41The participant decides
What would you do?
What just happened matters more than what they said. The participant did not rate a hypothetical. They chose one of four actions while the decline was on screen. The choice routed the interview: the follow-up they received exists because of the decision they made. A traditional questionnaire would have asked all follow-ups to everyone and measured nothing.
The typed answer is the richest field. Open text in a form gets skimmed and coded later. Open text here arrives attached to a specific moment in a specific scenario, which makes it dramatically easier to theme, because half the context is already structured by the branch.
a The scenario, the stakes and the branch points are authored by you. The engine supplies the conversation, the timing and the logging.
Show the thing.
Then ask about it.
The old instrument
Would you trust this marketing message?
The participant scores a sentence in isolation, guessing at a context neither of you can see. Attention is on the wording of the question. The answer lands in a spreadsheet stripped of the moment.
The scenario instrument
The notification slides down. The message is right there.
What do you think this company is offering?
What makes you say that?
[shows a second variation]
Which would you be more likely to tap?
The message is the stimulus and the medium at once. Comprehension gets tested before preference, which then gets asked against a real second option instead of a memory. Message testing stops being a survey about ads and becomes exposure with questions attached.
Studies people
are running this way.
- UX research
- See how people react to a proposed experience before any of it is built.
- Screenshots of the flow → tasks attempted in place → friction points identified by branch
- Concept testing
- Put the new product or feature in front of participants and ask what they would do.
- The concept is shown in context → comprehension question → preference between variations
- Message testing
- Find out what people understand from an ad, an email or a notification.
- Exposure → free-text comprehension → second variation → preference with a reason
- Customer journey research
- Meet customers at a stage and ask only about that stage.
- Stage selector → the questions for that stage → a follow-up tuned to the pain they name
- Behavioral research
- Give people a situation and observe the decisions they make inside it.
- Scenario → decision → consequence → the next decision, scored or logged per choice
- Academic studies
- Turn experimental manipulations into interactive experiences participants complete.
- Between-subject conditions as branches → consistent logging → verbatim capture per cell
- Interview pre-screening
- Qualify respondents before you spend an analyst hour on them.
- Screener branches into study lanes → the long path only for the people who fit
- Diary-style check-ins
- Catch the in-the-moment answers a recall study misses.
- Short links sent in waves → each wave follows up on the previous answer
From hypothesis
to field in one sitting.
- i
Author the situation
Write the scene as a chat conversation or describe it in one sentence and edit the AI draft. The scenario is the stimulus: get the stakes, the sender and the pressure right and the rest of the study writes itself.
- ii
Place the interrupt points
Drop questions where the decisions happen. Each answer carries its own follow-up, so a participant who tries the payment again is asked a different question than one who contacts the bank.
- iii
Field it to your panel
Publish and send the link. Participants play the scenario in a browser, their paths and timings are logged per person and the CSV export lands in the analysis tool you already use.
Questions from
careful researchers.
Put the next study
inside the situation.
Author the scenario, wire the branches and send the link to your panel. The interview plays itself, the answers arrive structured and the transcripts write themselves. Ships on the Studio plan with the answer log and CSV export included.