Example data. Fieldnote is a fictional company used to fill every board, so the examples tell one consistent story.
Fieldnote, session run with four customers, including Sophie. Date set at the Tuesday after the spring trade show, ten months out.
What the team took away
Nobody mentioned the features the roadmap led with. Every outcome pointed at parts and at the first week, which reordered the quarter.
How to run it
Preparation
Pick the date out loud before the session, and make it specific: 'the Tuesday after the trade show', not 'next year'. Vague futures produce vague answers.
Invite customers if you can. The team playing customers is a fallback, and it shows.
Prepare the tense rule and repeat it: everything is said in the past. 'We will be able to' is the sentence to catch.
Have the current metrics on hand so claims about the future can be compared to today.
Agenda
10 min
Set the date Name the future date and describe the world around it briefly. Everyone writes the date at the top of their sheet.
25 min
What happened Each person writes, in the past tense, what the product did for them between now and that date. Correct anyone who slips into the future tense, every time.
20 min
Tell a colleague Second pass: what did you tell a colleague about it? This forces the answer into a sentence someone would actually repeat, which strips out the features nobody cares about.
25 min
Work backwards For each outcome, walk back to what had to be true for it to happen. This is where the roadmap items appear, and they appear as consequences rather than wishes.
10 min
Keep the outcomes Rank the outcomes by how much the group wants them, not by feasibility. Feasibility is the next session's job.
Pitfalls
Letting the future tense back in. The whole mechanism rests on the past tense: it turns a wish list into a story, and stories expose what was missing.
Accepting feature names as outcomes. 'We had the mobile app' is a feature. 'I stopped doing the paperwork in the van at 7pm' is an outcome. Ask what it let them stop doing.
Setting the date too far out. Beyond a year, people describe a fantasy. Six to twelve months keeps it close enough to be accountable.
Afterwards
The outcomes feed Impact mapping or the Product Vision Board directly. If the walk-back produced a lot of assumptions, run Assumption mapping before committing to any of it.