Design Thinking Workshops Without the Sticky-Note Theatre

A design thinking workshop earns its day when the group leaves with a small number of ideas they have actually chosen between, and someone's name against each one. It fails when it produces a wall of four hundred sticky notes, a photograph in a Slack channel, and no change to what anyone does on Monday. The difference is almost never the quality of the ideas. It is how much of the day you spent converging.

Sticky-note theatre is the recognisable version of this failure: lots of movement, coloured dots, standing up, a facilitator with a marker pen, and a final half hour in which the group agrees that all of it was very interesting. Everyone had a nice time. Nothing was decided, because deciding was never scheduled.

This article walks through a one-day structure that keeps the useful parts of design thinking — customer evidence, wide divergence, cheap prototypes — and removes the parts that exist mainly to look like a workshop. Real timings, real group sizes, and the specific moments where these sessions come apart.

What sticky-note theatre actually looks like

You can diagnose it before lunch. The first sign is that nobody has said a number out loud. The group is discussing user needs, opportunity areas and how-might-we statements, and not one person has mentioned how many customers this affects, what it costs today, or how long a fix would take. Abstraction is comfortable and it is where a workshop goes to die.

The second sign is that ideas are still being added at 3pm. Divergence is enjoyable, low-risk and socially rewarded — nobody has ever been criticised in a workshop for suggesting one more idea. Convergence is the opposite: it requires someone to say that a colleague's idea is not the one, in front of the colleague. Groups avoid it by generating more.

The third sign is the language of the closing round. If the last activity is people saying what they found interesting, you have run a seminar. If the last activity is people committing to a next step with a date, you have run a workshop.

Start from a decision, not from a stage model

Design thinking is usually taught as a sequence of modes — empathise, define, ideate, prototype, test — and that framing is useful for explaining the discipline. It is a poor way to plan a specific day, because it tells you what to do and nothing about when to stop.

Plan backwards instead. Write the sentence you want to be able to say at the end: "We are going to test three onboarding changes with new sign-ups in February, and Priya owns the first one." Then work out the minimum set of activities that makes that sentence honest. Usually it is a shared view of the problem, one wide divergence, one hard convergence, and something concrete enough to show a real person.

This also tells you who needs to be in the room. If the closing sentence requires a budget decision, the person who controls the budget attends the last ninety minutes. Workshops that produce recommendations for an absent decision-maker convert at a dismal rate, and everyone in the room knows it by mid-afternoon.

Ground the group in evidence in the first hour

The empathise phase collapses into anecdote unless you bring raw material. Two or three verbatim customer quotes, a short recording, a support-ticket export, a funnel chart: something the group did not write themselves. Fifteen minutes of reading in silence beats an hour of someone presenting a summary of the same material.

Then check what the room actually believes. A quick anonymous poll on the two or three assumptions the day rests on — "Do you think new users understand what this product does in their first session?" — takes ninety seconds and frequently reveals that the leadership team and the support team are working from different pictures. Running that in Session Flo rather than by show of hands matters here, because the visible-hands version anchors on whoever raises first.

Close the hour by writing the problem statement on the wall in one sentence and leaving it there all day. Every later argument about scope gets resolved by pointing at it.

Diverge in silence before you diverge out loud

The standard image of ideation — a group calling out ideas while someone writes them up — is the weakest available format. Only one person can speak at a time, so everyone else is holding their idea in memory instead of developing it, and the first few suggestions set the frame that later ones vary around. The research literature on brainstorming has documented this production-blocking problem for decades.

The fix costs nothing. Give people six to eight minutes of silence to write their own ideas, one per note or one per line in a shared capture tool, before anybody speaks. Then round-robin: each person reads out one idea at a time until the group runs dry. Nominal group technique formalises exactly this — silent generation followed by structured sharing and ranking — and it consistently gets you more ideas and a wider spread of them.

Two practical details. Set a target number rather than a time only: "eight ideas each, and the last two are meant to be bad" gets people past their first, obvious answers. And run at least two divergence rounds separated by a stimulus — a constraint like "now assume we have no engineering time" — because the second round is where the non-obvious material appears.

Keep ideation groups at four to six

Per-person effort drops as a group grows. In an ideation table of ten, two people will do most of the work, three will contribute occasionally, and the rest will nod. In a table of five, everyone has to speak because the silences are too obvious to sit in.

If you have thirty people, that means five or six tables, not three. Yes, it makes the share-back longer — so do not share everything back. Have each table bring their two strongest ideas to the wall and leave the rest on the table for the digest.

Converge on purpose, in three passes

Convergence is where facilitation earns its money, and it works best as separate passes rather than one long argument. Each pass narrows the field by a defined amount and each has its own rule.

Do not let the group jump straight to discussing merits. Discussion favours the articulate and the senior; silent scoring first gives you a starting distribution that reflects the whole room, and then you discuss only the disagreements.

Prototype with what is in the room

Teams skip prototyping because they think it requires design skills and a tool nobody has open. It does not. A prototype in a workshop is anything concrete enough that a person outside the room could react to it, and a 45-minute slot is enough.

Useful low-effort formats: write the announcement email as if the thing already shipped; storyboard the six screens on paper; script and act out the two-minute support call; draw the one chart the change is supposed to move. Each of these forces specificity, and specificity is where the disagreements you have been avoiding all day finally surface.

Give every group the same fixed output format and the same deadline. Different formats make the comparison at the end a matter of taste rather than substance, and a soft deadline means the polished group beats the better idea.

Test it before the day ends

The most valuable half hour in a design thinking workshop is the one where the prototypes meet someone who was not in the room. Invite two colleagues from support, sales or operations for a 30-minute slot at 4pm. They see each prototype for four minutes and say what would go wrong.

If you genuinely cannot get an outsider, the fallback is a structured pre-mortem: assume this shipped and failed badly six months from now, write down why, individually and in silence, then read them out. It is not as good as a real reaction, but it beats a room congratulating itself.

A one-day agenda that ends in a decision

This is a nine-to-five shape for twelve to twenty-five people with two facilitators. Compress it to a half day by cutting the second divergence round and the outsider test, not by cutting convergence.

Running it remotely or hybrid

Remote design thinking is harder in one specific way: video-mediated groups generate a narrower set of ideas than the same people working face to face, which researchers have attributed to the way a video window narrows visual attention. It is not that remote teams are worse at thinking; it is that the format constrains the wandering attention that idea generation depends on.

The counter-measure is more silence and less video. Turn cameras off for the generation phase and tell people to look away from the screen, out of a window if there is one. Extend silent generation from eight minutes to twelve. Capture into a shared surface so nobody has to hold their idea while waiting for a turn.

In hybrid rooms, the killer is the wall. If the sticky notes are physical, remote attendees are spectators for the whole clustering phase. Either go fully digital for capture and voting, or give remote participants a dedicated in-room partner whose only job is to place their notes and read their comments aloud. Do not attempt to run a physical wall and a digital wall in parallel — they diverge within the hour and reconciling them eats your convergence time.

Failure modes and the recovery for each

What to send afterwards, and when

Send the recap the same day, while people still remember which idea was which. It needs four things: the problem statement, the three chosen bets with owner and date, the ranked list of everything that did not make the cut, and the parked items. The ranked list matters more than facilitators think — it is what stops the same idea being re-proposed as new in six weeks.

Do not send the photograph of the wall as the artefact. Send it as an appendix if you must. A photograph of sticky notes is unreadable within a fortnight and it signals that the wall was the point.

Then put a fifteen-minute review in the calendar for three weeks out with the three owners. Most of the value of a design thinking workshop is destroyed in the gap between the workshop and the first piece of follow-through, and a short scheduled check is the cheapest insurance available.