Prep: the volunteered venture agreed at the end of W02; the journey-map prompt rehearsed on the model in use; tables seeded from the W02 register; primers printed; wall space and tape for about 18 maps.
First CM-owned session. It sits between W02 (agentic research) and W04 (customer discovery), both Ryan’s. The session produces two artifacts students will use in W04.
Keep this to assumptions, not pitches. The question is narrow: what on this canvas would sink the venture if it turned out to be false. Stop anyone who starts describing the product.
Write the assumptions down. Several will reappear in the pin-up at 5:45, and naming that back to a student shows the session has a through-line.
The bridge out of the standup. The VPC describes one customer and one set of jobs. Today’s two artifacts ask who else is in this and when exactly does it hurt. Say that once and move on.
Fifteen minutes. Instructor talk for the session totals thirty minutes across two blocks, well under the fifty-minute cap that protects the studio blocks.
Most of this room can make a polished journey map and has never used one to make a decision. Say so directly and without judgement. For most students it has been a training artifact, not a working one.
Naming the two methods that are out matters. A method can be real and still not be useful this term. Students who have made empathy maps before should hear why it is absent, or they will assume it was forgotten.
[assumption] in the cell. This is fully acceptable and costs you nothing.Repeat the second row. Students hide assumptions because they read them as weakness. A map that is 80% [assumption] and honest about it is better than one with hidden guesses.
This is the design-axis version of the Evidence Handling criterion, which appears on every instrument in the course.
The fifteen minutes that has to land. A direct port of 317’s fabrication demo, moved onto the design axis, where it bites a room full of designers.
Rehearse it beforehand on the model you will use. If it hedges or asks for research first, find that out in rehearsal, not in the room.
Produce a customer journey map for someone considering [the volunteered venture]. Five stages, with emotions and pain points at each.
Read the whole thing out. Stage names, emotional arc, pain points. All of it.
Read it slowly and straight. No editorialising. It will be immediate, complete and plausible: stage names that sound right, an arc that tracks, pain points that sound like things a person would say.
It will look like four hours of workshop output. The room nodding along is the setup.
Now the student whose venture it is reads out one thing they actually heard in W02. One sentence, from a real person. Put the two side by side and leave them there.
Do not narrate the comparison. Let the room look.
Whatever the model did, whether it invented the arc, hedged, or produced something generic, describe what actually happened rather than a pre-scripted punchline.
The student’s one real sentence will be smaller, messier and more specific than anything on screen. That contrast is the point of the fifteen minutes.
The full line: “It didn’t guess. It produced the shape of a journey map, and the shape is real. That is why it is convincing. What is missing is the only part that was ever yours: the evidence. Your job is the substance. With that in place the tool becomes genuinely useful, and you will see that in about an hour.”
Do not let this become an anti-AI moment. The studio block is the payoff: the same model, pointed at their real material, doing work they could not do as fast alone. Deliver both halves, or the term tips into cynicism.
Clustering twenty interview fragments into candidate stages, fast. Proposing stakeholders you had not considered. Second and third passes better than your first.
It fabricates the emotional arc, the most persuasive part of the map. It also collapses the payer into the user almost every time, because most of its training data does.
You know who actually pays. You know which stages you have evidence for. Ninety seconds of your attention, and it is the thing being graded.
Named in W01, in use all term. Say plainly that it applies to model output only. Evidence students gathered themselves needs no Lie column. An interview they ran is not a source to be audited this way.
Keep it separate from Ryan’s 4-Risks. 4-Risks is about your venture. Lift/Lie/Catch is about your sources. Neither nests under the other.
Draw a real example from the room rather than supplying one. Ask whose venture has a payer who is not the user. Two hands will go up and one of them will be surprised.
“Who loses” is the row that produces useful discomfort. Ventures in this room routinely displace the people they describe as beneficiaries, and the map is where that becomes visible.
Say the second chip out loud. The studio blocks run on the students’ own material, and the first five minutes are lost if eighteen people start hunting for files at 4:35.
Ten minutes of peer micro-teach, twenty-five of build. Tables are seating, not units of work. Every student makes their own map for their own venture.
Tables are seeded from the W02 register, with at least one experienced student per table per artifact. If a table has nobody, hand them the primer and check back after four minutes.
Do not teach the artifact from zero yourself. That is what the ten minutes and the primer are for. Teaching it from the front punishes everyone who already knows it.
[assumption]. No exceptions.This is the payoff to the set-piece: the same model, their real material, now useful. Walk the room and point that out to two or three people by name.
Watch for maps with one node. That student is designing for whoever they interviewed and has not noticed yet. This is why the artifact comes before W04.
The middle row is the one line of teaching in this block. Say it yourself if the micro-teach does not produce it. Locating the worst moment is what turns a canvas into a design brief.
[assumption].Third row, watch the wording. Students mark the stage with the most boxes on it. The worst moment is emotional, not operational, and it is often in a stage that looks quiet on paper.
Give a two-minute warning at 5:43. The pin-up needs the full thirty.
[assumption].” All three are good outcomes.Eighteen students, six groups of three, thirty minutes including the handoff. Keep the two-minute limit firm or the last two groups get nothing.
Listen for the riskiest assumptions from the 3:25 standup coming back. Naming that out loud makes the session read as one argument rather than two workshops.
“The worst moment you just marked is what you test next week. If you marked it from evidence, you have a question. If you marked it from a guess, you have a better question. Bring the map.”
W04 is customer discovery, and it is Ryan’s. This sentence is the seam between the design axis and the business axis. Say it in these words so the two halves read as one course.
6:10, not 6:14. This is the sentence the session exists to earn. Do not deliver it over the sound of eighteen people packing up.
Confirm the VPC is submitted on Moodle before students leave.
These slides are a fixed 16:9 canvas. In portrait they shrink to an unreadable strip; landscape gives you the whole slide.