---
title: "Making better decisions as a group | Optimize All Academy"
description: "Why group decisions go wrong Groups can make better decisions than individuals because they combine knowledge and perspectives. But they often fall into…"
url: https://optimizeall.com/learn/leadership-and-communication/group-decisions
updated: 2026-10-05
---

Leadership & Communication · Meetings, writing and decisions · lesson 9 of 21 · 13 min

# Making better decisions as a group

## Why group decisions go wrong

Groups can make better decisions than individuals because they combine knowledge and perspectives. But they often fall into predictable traps:

- **Groupthink:** desire for harmony suppresses dissent.
- **Anchoring:** the first number or idea mentioned dominates.
- **Authority bias:** everyone defers to the most senior person.
- **Confirmation bias:** people seek information that supports their existing view.
- **Unclear decision rights:** nobody knows who decides, so decisions drift.

## Step 1: clarify how the decision will be made

Before discussion, state **who decides** and **how**:

| Method | When to use | Watch out for |
|---|---|---|
| Leader decides alone | Urgent, low-impact or confidential | Low buy-in |
| Leader decides after consultation | Most important decisions | Consultation must be genuine |
| Majority vote | Many equal stakeholders, moderate stakes | Losing minority disengages |
| Consensus (all can live with it) | High need for commitment | Slow; can water down decisions |
| Delegate to the group or an individual | Expertise sits with them | Must respect their decision |

Frameworks such as RACI or RAPID (Recommend, Agree, Perform, Input, Decide) help make decision rights explicit.

## Step 2: frame the decision well

- What exactly are we deciding? (One sentence.)
- What criteria matter? (Cost, time, risk, customer impact, strategic fit.)
- What options exist? Always generate at least three, including "do nothing".
- What information do we need, and what is good enough given the stakes and time?

## Step 3: improve the discussion

Techniques that reduce bias:

- **Independent input first:** ask people to write their view or estimate before discussion (reduces anchoring and authority bias).
- **Senior people speak last.**
- **Devil's advocate or red team:** someone argues the strongest case against the preferred option.
- **Pre-mortem:** "Imagine it is a year from now and this decision failed. Why?"
- **Separate facts from opinions:** label clearly.

## Step 4: decide and commit

State the decision clearly, including the reasoning and the key trade-offs accepted. Ask for commitment: people may disagree but should support implementation ("disagree and commit"). Record it in a decision log.

```
Decision log entry
Date: 12 March
Decision: Enter Saudi market first; UAE in phase 2
Decided by: CEO after consultation with leadership team
Options considered: KSA first; UAE first; both simultaneously
Key reasons: partner readiness, market size, regulatory timeline
Trade-offs accepted: slower UAE revenue; higher initial travel costs
Dissent noted: CFO preferred UAE first due to lower setup cost
Review point: after 6 months, compare against KPIs
```

## Reversible and irreversible decisions

Distinguish decisions that are easy to reverse (try something, learn, change course) from those that are hard to reverse (major investments, public commitments). Move quickly on reversible decisions; invest more time and analysis in irreversible ones.

## Worked example

*Illustrative.* A fictional e-commerce startup in Lahore had to choose a payment provider. In the first meeting, the founder stated a preference, and everyone agreed. A board adviser suggested restarting: each team member independently scored three providers against agreed criteria (fees, reliability, settlement time, support, integration effort). The independent scores revealed that the founder's preferred provider scored poorly on settlement time, which mattered greatly for cash flow. The team chose a different provider, and the founder later described the process as the most useful change to how they made decisions.

## Hands-on: a decision record and a pre-mortem you can run in 20 minutes

**1. One-page decision record** (keep them in a shared folder or wiki so decisions are findable):

```text
DECISION RECORD — [title] — [date]
Decision needed:   (framed as a question with options, not yes/no only)
Decider:           (one named person or body)     Rule: consult / consent / vote / consensus
Input from:        (names/roles)                   Informed: (who hears afterwards)
Type:              one-way door (hard to reverse) | two-way door (easy to reverse)
Options considered: A / B / C with main pros, cons and risks
Evidence:          (data, sources; flag assumptions)
Decision:          
Why:               
Dissent recorded:  (who disagreed and why; "disagree and commit" agreed?)
Review date / signal that would make us revisit:
```

**2. Pre-mortem (20 minutes, from Gary Klein's technique):**

```text
1. "Imagine it's 12 months from now and this decision has failed badly." (1 min)
2. Everyone writes reasons it failed, silently and independently. (5 min)
3. Round-robin: one reason each until all are shared. (7 min)
4. Pick the top 3 risks; agree a mitigation or an early-warning signal for each. (7 min)
```

**3. AI as a devil's advocate** (anonymise sensitive data):

```text
We are about to decide [decision] for [context]. Our preferred option is [A] because [reasons].
Argue the strongest case against A and for the best alternative. List assumptions we
are making that could be wrong and what evidence would test each. Do not decide for us.
```

Use the output to widen the discussion. It is not evidence, and it cannot know your organisation's context.

## Common mistakes

- Starting discussion before clarifying who decides.
- Only one option considered.
- Leader states preference first.
- Endless consensus seeking on low-stakes decisions.
- Decisions not recorded, so they are reopened.

## Quick self-check

Think of your team's last significant decision. Could everyone say who made it, what options were considered and why the chosen option won? If not, use the decision log template next time.

## Decision-making under time pressure

When time is short, decide who decides, set a deadline for input, and accept "good enough" information for reversible choices. Communicate the decision and the reasoning quickly, and schedule a review point. Delaying a decision is itself a decision, often with hidden costs.

## Learning from decisions

Judge decisions by the quality of the process as well as the outcome. A good process can still produce a bad result through bad luck, and a poor process can get lucky. Reviewing decision logs periodically helps teams learn which processes lead to better results over time.

## Video lecture: Making better decisions as a group

Lecture coming soon · 13 chapters · about 8 minutes. Read the full transcript below.

1. Group decisions
2. Why it matters
3. Clarify the process first
4. Jury, referee or family holiday?
5. Frame and discuss
6. Decide and commit
7. Worked example 1: Grace's agency (illustrative)
8. Worked example 2: Karachi retail board (illustrative)
9. The pre-mortem (Gary Klein)
10. Watch me: decision record
11. Measuring decision quality
12. Common mistakes
13. Recap and try this now

## Lecture transcript

### Group decisions

Have you ever sat in a meeting where everyone nodded along to a decision, and then, in the corridor afterwards, half of them said they thought it was a mistake? That's not a people problem. It's a process problem. Groups can make much better decisions than individuals, because they bring more information and perspectives. But only if you design the process. In this lecture, you'll learn why group decisions go wrong, how to clarify who decides and how, how to frame and discuss a decision well, and how to decide and commit, including using a pre-mortem.

### Why it matters

Why does it matter? Because the quality of a team's decisions compounds. A slightly better choice on priorities, hiring, pricing or suppliers, made again and again, adds up to a very different business over a year. And groups have well-known traps. Groupthink, where the desire for harmony suppresses disagreement. Anchoring on the first idea mentioned. The most senior or loudest person steering the room. And endless debate with no clear decider. Each of these has a practical fix.

### Clarify the process first

Here's the first concept. Before discussing the content, clarify the process. Who decides? One person, a small group, or everyone? And what's the rule? Consult, where the decider hears input and then decides. Consent, where the proposal goes ahead unless someone has a reasoned objection. Vote. Or full consensus, which is slow and best kept for decisions everyone must live with deeply. Frameworks such as RAPID, from Bain, or DACI, used in many product teams, name the roles explicitly. Here's the key idea: most decision conflict is really confusion about who decides.

### Jury, referee or family holiday?

Here's an analogy. Think of a jury, a referee and a family choosing a holiday. A jury must reach a verdict by a defined rule. A referee decides alone, instantly, after taking in what they can see. A family discusses, but in the end one or two people book the flights. Each is a legitimate way to decide. Chaos comes when people think they're on a jury, but actually there's a referee. Tell people which game they're playing before the discussion starts.

### Frame and discuss

Next, frame the decision well. Instead of a yes or no question, like should we launch in Saudi Arabia, frame it with options: which market should we enter next, Saudi Arabia, Qatar, or deepen the UAE? That invites real comparison. Then improve the discussion. Ask people to write their views independently before anyone speaks, so the first opinion doesn't anchor everyone. Have the most senior person speak last. Assign someone to argue the other side. And separate facts from assumptions on a whiteboard as you go.

### Decide and commit

Then decide and commit. Distinguish between one-way doors and two-way doors, a phrase popularised by Amazon's shareholder letters. A two-way door is a decision that's easy to reverse: decide fast, learn, adjust. A one-way door is hard to reverse: slow down and do more work. Once the decision is made, ask for disagree and commit. People can disagree openly during the discussion, but once it's decided, everyone supports it fully. Record dissent, and set a review date or a signal that would make you revisit.

### Worked example 1: Grace's agency (illustrative)

A simple worked example. A four-person social media agency in Manchester must choose a new project management tool. Last time, the debate dragged on for a month. This time the founder, Grace, states the process: she'll decide, after input, by Friday. It's a two-way door, since they can switch in a few months. Everyone writes their top criteria independently. They compare two options against those criteria in thirty minutes. Grace decides on Friday. One designer disagrees, says so, and commits. They review in three months.

### Worked example 2: Karachi retail board (illustrative)

Now a realistic business scenario, illustrative. The board of a family-owned retail chain in Karachi must decide whether to open a large new store in a new mall. It's a one-way door: a long lease and significant capital. The chief executive frames three options: open the large store, open a smaller format, or invest the same capital in e-commerce. Before discussion, each director writes their view. The finance director argues against the preferred option. Then they run a pre-mortem, which surfaces a risk nobody had raised: parking and access during mall construction. They choose the smaller format, with a clause to expand.

### The pre-mortem (Gary Klein)

Let's zoom in on the pre-mortem, a technique developed by the psychologist Gary Klein. Before committing, say: imagine it's a year from now, and this decision has failed badly. Why did it fail? Everyone writes reasons silently, then shares them in turn. It works because it gives people permission to voice doubts without seeming negative. The group is no longer criticising the plan; it's explaining a hypothetical failure. You then pick the top three risks and agree a mitigation or an early warning signal for each. Twenty minutes can save a year.

### Watch me: decision record

Watch me do it. I'm filling in a decision record for an illustrative choice: whether our team adopts an AI assistant for first drafts of client reports. Question with options: adopt tool A, adopt tool B, or wait six months. Decider: head of client services, after consult. Type: two-way door, but with a one-way element, data handling, so our data protection lead is an input. I ask an AI assistant to argue the strongest case against our preferred option. It raises client confidentiality and review time. I add both to the risks, and we set a review date after a sixty-day pilot.

### Measuring decision quality

How do you measure decision quality? Not only by outcomes, because good decisions can have bad luck. Also judge the process. Was the decider clear? Were options compared? Were assumptions tested? Was dissent heard? Keep decision records, and at the review date, compare what you expected with what happened. Over time, you'll see patterns, such as always underestimating time, or overestimating demand. That's how teams learn to decide better.

### Common mistakes

Common mistakes. Starting the debate before agreeing who decides. Yes or no framing. Letting the boss speak first. Mistaking silence for agreement. Treating every decision as a one-way door, so everything is slow, or every decision as a two-way door, so big risks are rushed. Re-opening decisions endlessly. And using an AI's answer as if it were evidence. It's a useful sparring partner, but it doesn't know your context.

### Recap and try this now

Let's recap. Clarify who decides and by which rule before discussing content. Frame decisions with options. Collect views independently, let senior people speak last, and name a devil's advocate. Match your speed to one-way or two-way doors, run a pre-mortem on big calls, then disagree and commit. Record decisions and review them. Your try this now: pick one decision your team faces this month and complete the decision record in the lesson before the meeting. Next module, we move on to stakeholders and influence.

## Key takeaways

- Group decisions suffer from groupthink, anchoring, authority bias and unclear decision rights.
- Clarify who decides and how before discussion; frame the decision with criteria and options.
- Collect independent input first, have senior people speak last, and use pre-mortems or devil's advocates.
- Decide clearly, record reasoning and dissent, and move faster on reversible decisions.

## Try it

For your next team decision, use independent written input, have the most senior person speak last, and record the result in a decision log.

- [Previous: Writing that gets read and acted on](https://optimizeall.com/learn/leadership-and-communication/clear-writing)
- [Next: Stakeholder management](https://optimizeall.com/learn/leadership-and-communication/stakeholder-management)
- [All lessons of Leadership & Communication](https://optimizeall.com/learn/leadership-and-communication)