Dream, Debate, Decide & Do

"I thought we already decided that."

Picture a leadership team forty minutes into a conversation about next year's budget. The head of product floats an idea for a new service line. The CFO starts pressure-testing it: what's the margin, who staffs it, what do we stop doing to fund it. Three seats down, the COO looks up from her notes and says the sentence that derails a thousand meetings a day:

"Wait. I thought we already decided that."

Nobody in that room is doing anything wrong. The head of product is dreaming. The CFO is debating. The COO believed she walked into a planning conversation about a decision made last quarter. Three smart people, one room, three different meetings.

That's the diagnosis most meeting advice misses. The problem usually isn't a bad agenda, a weak facilitator, or the wrong people in the room. The problem is that the people in the room think they're in different meetings. And because nobody named the meeting out loud, nobody knows they disagree about what kind of conversation they're having. They just know it feels like wading through wet cement.

The Four Meetings

Every working conversation is doing one of four jobs. There are only four meetings: Dream, Debate, Decide, and Do.

Each one asks something different of you before you walk in, and each one has a different definition of success. Get those right and the meeting mostly runs itself. Mix them without saying so and even a great agenda can't save you.

Dream

The job: Explore possibilities. Generate options you don't have yet.

What you bring: Ideas. Including the ones you'd normally keep to yourself. The pre-work for a Dream meeting is simple: get people thinking about the question before they gather, so the meeting starts with material instead of blank stares.

What good looks like: More options at the end than the beginning, and nobody punished for the strange ones. A Dream meeting makes no commitments and closes no doors. That's not a bug. Deciding too early is how organizations end up choosing from a menu of two mediocre options because nobody made space to find the third.

A tool that helps: Easy / Hard / Safe / Crazy. One big sheet, a "How might we..." question at the top, four quadrants below. Easy: what are some easy things we can do right away? Hard: what are some hard things we can begin to explore? Safe: what are the safe things everyone is expecting us to do? Crazy: what are a few things so crazy they just might work? Groups as small as two or as large as seven work one sheet together. Don't censor on the way in. Crazy is where the good ones hide.

Debate

The job: Build understanding. Test ideas against each other and against reality.

What you bring: A point of view, and the willingness to have it changed.

What good looks like: The group understands the real disagreement, which is frequently not the one they walked in with. Debate is not combat and it is not consensus-building. It's how a group finds out what it actually thinks before it has to commit.

A tool that helps:The Confidence Compass. Each person states their position, their confidence in it as a percentage, and what would change their mind. That last question does the heavy lifting. We've watched two executives argue for twenty minutes before the Compass revealed they were at 55% and 60% confidence, worried about the same thing. What looked like disagreement was shared uncertainty. That's a completely different conversation, and a much shorter one.

Decide

The job: Evaluate options and commit to one.

What you bring: Knowledge of who the decider is and how the call gets made, before anybody walks in.

What good looks like: A decision, obviously. But also this: everyone leaves knowing what was decided, who decided it, and why. Most Decide meetings fail before anyone speaks, because people think they're arguing about the answer when half of them haven't agreed on the question.

A tool that helps:The Decide Guide. It's a preparation document, not an agenda, and it gets filled out before the meeting is scheduled. Five phases: Frame the question. Lay out the options. Agree on the criteria before applying them to the options, because criteria chosen after the fact are just justifications. Make the call. Assign what happens next. The Guide also carries a prerequisite check: has this group Dreamed? Has it Debated? If not, don't pretend the calendar invite makes it a Decide meeting. You can't decide what you haven't explored.

Do

The job: Turn the decision into motion. Plan, assign, coordinate.

What you bring: Your capacity and your calendar.

What good looks like: You leave knowing who's doing what, by when. That's the whole test. A Do meeting that ends without names and dates attached to work is a Debate meeting wearing a disguise.

A tool that helps: The commitment check. Before anyone leaves, go around the room: each person says out loud what they own and when they'll deliver it. It takes four minutes. It converts "we should" into "I will," and it surfaces the overloaded person now instead of at the deadline.

The expensive part: mixed meetings, unnamed

Here's what happens without the vocabulary. A team schedules a "weekly sync." Inside that hour, someone dreams, someone debates the dream, someone tries to decide, and someone else just wants to know what to do by Friday. Every one of them leaves frustrated, and every one of them blames the meeting.

One meeting trying to do two jobs does both badly. Dreaming gets strangled by premature evaluation. Debating gets cut off by someone pushing for a decision. Decisions get relitigated because half the room never got to debate. And the person who came ready to Do sits through ninety minutes of exploration wondering why they were invited.

Sometimes you genuinely need two of these conversations in one room on one day. Fine. Have both. Just name the seam: "For the next twenty minutes we're dreaming. No evaluation. Then we'll switch and pressure-test." Two meetings, two purposes, both clear. The failure isn't mixing. The failure is mixing without telling anyone.

The fix is three sentences

Before you send the next invite, finish these three sentences from our Meeting Guide:

We're gathering to dream / debate / decide / do (about) ______.

It is important for us to do this together, and at the same time, because ______.

And by the end of this meeting we'll have ______.

The first sentence names the meeting, so everyone arrives in the same one. The second is the honesty test: if you can't finish it, you don't have a meeting, you have an email. And the third defines done, so the meeting ends when the work does instead of when the calendar says so.

Put all three in the invite. Everyone who accepts knows what kind of conversation they're walking into, why they're in the room, and what they're building together.

And when the stakes are high, or you suspect people are arriving in different meetings anyway, open with a two-minute expectations check we call I Wish / I Wonder / I Worry. Everyone writes against three prompts: I wish we would... I wonder if we could... I worry that we won't... Then everyone shares. It takes five minutes, and it surfaces the mismatches before they cost you forty. The person who wishes you'd revisit last quarter's decision gets to say so at minute two instead of derailing minute forty.

We'll be honest about what we learned rolling this out ourselves: it was the hardest habit to build and the highest-payoff habit we've kept. Hard because "let's discuss" is comfortable and the three sentences force the organizer to know what they actually want. If you can't finish them, you've learned something important before the meeting ever happens. Maybe it's two meetings. Maybe it's an email. Maybe it's not ready to be a meeting at all.

Bigger than meetings

The framework earns its keep on the calendar, but the real compounding happens in conversations nobody scheduled.

A CEO stops someone in the hallway and asks, "What do you think about moving the launch?" What kind of question is that? If it's a Dream question, the honest answer is speculative and half-formed, and that's fine. If it's a Decide question, that same half-formed answer just became input to a call the CEO makes at 4 PM.

Without shared vocabulary, people guess. And when the stakes feel high, they guess conservative: they hedge, they say what's safe, they save the interesting thought for never. With the vocabulary, something better becomes possible. The person can ask: "Is this a dreaming conversation, or are you deciding?"

That question is permission. Permission to think out loud without it being held against you. Permission to disagree without it being read as obstruction. Permission to ask the CEO to clarify, which in most organizations feels vaguely insubordinate right up until everyone has words for it. Teams that adopt the Four Meetings don't just run better meetings. They have better conversations, because everyone can finally tell each other what kind of conversation they're in.

Try it this week

Open your calendar. Look at the next five meetings and label each with one word: Dream, Debate, Decide, or Do.

The ones you can label will get sharper the moment you share the label. The ones you can't label are the ones costing you the most, and now you know where to look.

Last Updated: July 8, 2026