Discover, Discuss, Decide
"People generally think the purpose of a meeting is for three things: discover, discuss, decide. The only thing that's ever supposed to happen in a meeting is the discussion part." - Annie Duke
What It Is
Discover, Discuss, Decide is a framework for restructuring how teams make decisions to avoid the pitfalls of group dynamics. Most teams conflate three distinct activities—discovering what people think, discussing those views, and making decisions—into a single meeting. This leads to cross-influence, loudest-voice-wins dynamics, and false consensus.
The solution: separate these activities. Discover opinions independently and asynchronously before the meeting. Use the meeting only for discussion. Make decisions after the meeting, also independently.
This approach comes from research showing that groups working as "nominal groups" (independently but in coordination) produce better outcomes than groups that interact in real-time from the start.
How It Works
Discover (Before the Meeting)
Gather opinions from each participant independently and asynchronously:
- Send prompts via email, Slack, or forms
- Ask for specific inputs: forecasts, priorities, concerns, ideas
- Request brief rationales (3-5 sentences) explaining their reasoning
- Use tools like Google Forms that hide others' responses
- Compile results before the meeting
Example prompts:
- "What's your estimate for how many sprints this will take? Give your lower bound, best guess, and upper bound."
- "Force-rank these five features from most to least important. Explain your top choice."
- "What are the biggest risks you see with this approach?"
Discuss (During the Meeting)
Use meeting time for genuine discussion:
- Share the compiled results so everyone sees the spread of opinions
- Focus discussion on areas of disagreement, not agreement
- The facilitator reflects back what each person says without adding their own opinion
- Ask clarifying questions: "What I heard you say is X—is that right?"
- No interrupting, no "I disagree," no coercion toward alignment
Key principle: The goal is conveying information, not convincing anyone. People should explain why they believe what they believe, not argue for their position.
Decide (After the Meeting)
Make the final decision outside the group setting:
- One decision-maker (DRI) is ideal, but voting can work if done independently
- If voting, use private forums or forms—no seeing others' votes
- Don't seek "alignment"—disagreement is natural and healthy
- Use "Nevertheless": acknowledge the input, make the decision, move forward
How to Apply It
For Product Roadmap Planning
Before:
- Send a list of potential features to each stakeholder
- Ask them to force-rank the list and provide brief rationale for their top 3
- Compile the responses into a single view showing each person's ranking
During:
- Review the compiled rankings together
- Identify features where rankings diverge significantly
- Have each person explain their reasoning for divergent items
- Facilitator reflects back without judgment: "So you're saying X because Y"
After:
- Product lead (DRI) makes the final prioritization decision
- Communicates the decision with acknowledgment: "I heard the team's input. The discussion surfaced important trade-offs. Nevertheless, here's what we're doing..."
For Sprint Estimation
Before:
- Share the feature specs with the team
- Ask each engineer independently: "How many sprints? Give lower bound, estimate, upper bound"
- Collect rationales: "What assumptions are you making?"
During:
- Display all estimates side by side
- When estimates differ significantly, ask the outliers to explain their reasoning
- Others can ask clarifying questions only
After:
- Tech lead decides on final estimate based on discussion
- Record the estimate and the rationales for later learning
On-the-Fly Application
You can use this even when something comes up unexpectedly in a meeting:
"Stop. Everyone take out a piece of paper. Write down your estimate for X. Don't share until everyone's done."
Then collect and discuss the spread of answers.
When to Use It
- Any meeting where decisions will be made - Planning, prioritization, strategy
- Forecasting and estimation - Sprint planning, budgeting, timelines
- Brainstorming - Generate ideas independently first, then combine
- Policy discussions - Where opinions vary and you need to surface real disagreement
- Hiring decisions - Interviewers share assessments independently before debriefing
Why "Alignment" is a Trap
Annie is direct about this: "The word alignment is stupid and it shouldn't be used."
Why? Because true alignment doesn't exist. Ten people with different experiences will naturally have different opinions. Pretending otherwise:
- Ignores reality
- Makes meetings coercive (pressure to agree)
- Silences dissent
- Reduces information quality
Instead: Accept that disagreement is natural. Hear everyone's view. Make the decision anyway. People feel more ownership when they're heard, even if the decision doesn't go their way.
Source
- Guest: Annie Duke
- Episode: "This will make you a better decision maker"
- Key Discussions:
- (00:25:18) The easiest-to-implement change: stop talking to each other so much
- (00:26:29) The problems with discovery in groups—cross-influence, loudest voice wins
- (00:30:14) Using forms and tools to gather independent input
- (00:31:39) Why "alignment" is a problematic concept
- (00:36:58) The "curious, not coercive" model
- (00:38:02) Reflecting back without judgment
- YouTube: Watch on YouTube
Related Frameworks
- Thinking in Bets - The underlying philosophy of making implicit explicit
- Three Questions to End a Meeting - Ensure meeting outcomes stick
- Curious Disagreement - How to discuss differences productively