The Pigeon Test
Structure · 60 min01 / 05

Same Plan, Ten Times

Find the two kinds of work your organisation plans as one.

What is made possible

A group puts its own work on a wall and sorts it by one question, whether the same plan would give the same result twice. Then it marks which pieces of work were planned as though it would. If the marks pile up on work that wouldn’t give the same result twice, the group sees that for itself, in its own handwriting, and nobody has to persuade anyone of it.

Five structural elements (min specs)

Structuring invitation.

Invite the group to think of all the actual work the organisation does, and to sort it twice to see what the wall says. The wording is in the Intro step, below.

Space and materials.

A wall with two zones marked in tape, left and right, unlabelled. Sticky notes, pens, dot stickers, and room for everyone to stand at the wall together.

Participation distribution.

Everyone writes, sorts and dots, and nobody’s stickies carry more weight than anybody else’s. The facilitator asks the questions and otherwise stays out of it. Best with eight people or more; below that the wall is too quiet to argue.

Group configuration.

Alone, then pairs, then foursomes, then everyone, then foursomes again.

Steps and time allocation

  1. 2 min

    Intro. Give the invitation. “Think of the actual work your organisation does. All of it, from fixing servers to merging teams. We’re going to sort it, twice, and see what the wall says.” Point at the two taped zones without explaining them; the sort question comes later. Say once, plainly, that some of the work going on this wall was planned by people standing at it, and that nobody is here to be blamed for it.

  2. 4 min

    Write alone. Each person writes real pieces of work, one per sticky. E.g. Fix a server, deliver a service, launch a product, restructure a department, merge two teams.

  3. 5 min

    Pairs. Combine your stickies and keep the eight most specific.

  4. 15 min

    The sort. Pairs join into foursomes and take their stickies to the wall. Ask the question this one is named after. “If you did this ten times with the same plan, would you get the same result every time?” Left zone, yes. Right zone, no. Two minutes of argument for the borderline ones, then they go where the foursome lands. Draw a line under any sticky that used the full two minutes; the wall should remember what was fought over.

  5. 8 min

    Second pass. Still in foursomes, nothing moves. Ask the foursomes to put one new question to every sticky in both zones. “When your organisation last did this, was there a plan that promised a particular result by a particular date?” If yes, it gets a dot.

  6. 12 min

    Stand back. Everyone. Ask, “What do you notice about where the dots are?” Then set a timer for two minutes of silence for everyone to think and write one thing they notice on a sticky. Then go round the room, each person reading theirs out, with no commentary.

  7. 9 min

    Live inside it. Back into the same foursomes. Each foursome picks one dotted sticky from the right zone, a piece of work its members are involved in. Ask, “What does its plan pretend that everyone at this wall knows isn’t certain?” Then look at the right zone again and ask, “Is anything there working without a plan?”

Call it an hour once people have moved about.

Practice insights

  • Tips

    The sort question has to stay that concrete. “Is this predictable?” lets people answer with their self-image instead of their experience. The argument is half the point; the two-minute timebox stops it eating the whole session. Dots in the left zone are fine, they belong there; watch whether the right zone collects just as many. Let the timer hold the silence after the stand-back question. If the right zone comes back clean, let it stand. Either the organisation really doesn’t plan its unpredictable work as though it were predictable, or the room believes more of its work is predictable than it is, and either is worth knowing.

    Someone may ask why there is no middle zone for work that’s hard but knowable. Because the question doesn’t need one. A rocket launch is about as hard as work gets and it still goes in the left zone. Simple and complicated both give you the same result twice, which is all the sort is asking. The other reason is that a middle zone is where the argument goes to die, since hard-but-knowable is the flattering answer and given the option the wall fills up with it.

  • Riffs and variations

    For a group that has read the book, name the zones at the end. Left is the catapult and the ball. Right is the pigeon.

  • Practical applications

    Before annual planning, or at the kick-off of anything with transformation in the title. Retrospectives that keep concluding the plan was fine and the people weren’t.

  • Optional Liberating Structures string

    Carry each foursome’s chosen sticky into 15% Solutions (opens in a new tab), or into Creative Destruction (TRIZ) (opens in a new tab).

Where it came from

The official repertoire already sorts work this way in Agreement-and-Certainty Matrix (opens in a new tab), which maps challenges by how much people agree and how much is known. This one asks a blunter question and puts the answer on a wall.

The book behind these

These come out of The Pigeon Test: Why Change Fails, which makes the argument for them. Buy it on Amazon (opens in a new tab). If you’d rather I ran this one with your group, see work with Mark.