The Decision-Making Paradox

PART 1: THE TWO SYSTEMS OF THINKING

The Fast and Slow Mind

Daniel Kahneman, who won the Nobel Prize in Economics for his work on human judgment and decision-making, describes two distinct systems our brains use for thinking and deciding. He calls them System 1 and System 2. Understanding this distinction is foundational to everything else in this article.

System 1: Fast, Automatic, Intuitive

What It Is:

System 1 operates automatically and quickly, with little or no effort and no sense of voluntary control. It’s your brain’s default mode for processing information and making judgments.

System 1 is:

• Pattern recognition

• Gut instinct

• Immediate reactions

• Unconscious processing

• Emotional responses

• Heuristics (mental shortcuts)

When It’s Active:

• Walking into a meeting room and immediately sensing tension

• Glancing at a pitch deck and thinking “this feels off”

• Meeting someone and forming an instant impression

• Hearing a proposal and immediately knowing your reaction

• Driving a car without consciously thinking about every movement

• Reading faces and body language

• Answering simple questions without deliberation

Why It’s Valuable:

System 1 is incredibly useful. It’s how you process the vast majority of information you encounter daily without being paralysed by analysis. It’s how you function efficiently.

For experienced leaders, System 1 is where expertise lives. After years of pattern exposure, you can sense problems before they’re obvious. You recognise situations you’ve seen before. You make rapid, often accurate judgments because they’re based on deep experience.

A CEO with 15 years of experience can look at a financial report and immediately spot something anomalous. They can sit in on a sales call and sense that the deal won’t close. They can observe team dynamics and recognise dysfunction. This is System 1 working well, intuitive expertise developed through repeated experience.

The Critical Flaw

System 1 has predictable, systematic errors:

1. It jumps to conclusions from limited evidence

2. It’s overconfident about its judgments

3. It’s easily influenced by framing, context, and recent experiences

4. It mistakes correlation for causation without testing

5. It sees patterns in randomness where none exist

6. It’s biased by emotions, anchors, and cognitive shortcuts

7. It substitutes easier questions for harder ones without you realising it

Most critically: System 1 doesn’t know its limitations. It operates with high confidence even when it’s wrong. The feeling of certainty is not correlated with actual accuracy.

And here’s the most dangerous part: System 1 thinks it’s System 2. You feel like you’re analysing rationally when you’re actually reacting intuitively. The subjective experience of careful reasoning and that of rapid intuition feel very similar from the inside.

System 2: Slow, Deliberate, Analytical

What It Is

System 2 allocates attention to effortful mental activities that demand it, including complex computations, deliberate reasoning, and careful analysis. It’s your brain’s “slow mode.”

System 2 is:

• Conscious reasoning

• Deliberate analysis

• Step-by-step problem-solving

• Mathematical calculation

• Hypothesis testing

• Weighing options systematically

When It’s Active

• Building a financial model with multiple scenarios

• Working through a complex strategic decision with pros/cons

• Forcing yourself to consider alternatives you don’t instinctively like

• Analysing competitive positioning across multiple dimensions

• Calculating probabilistic outcomes

• Evaluating evidence that contradicts your initial impression

• Deliberately slowing down a decision to think it through

Why It’s More Accurate

For complex, novel, high-stakes decisions, System 2 produces better outcomes than System 1. It can:

• Consider multiple factors simultaneously

• Test hypotheses against evidence

• Catch logical errors

• Override intuitive but wrong answers

• Think probabilistically about uncertain outcomes

• Recognise when initial reactions might be biased

The Critical Limitation

System 2 is slow, effortful, and your brain resists using it. Engaging System 2 requires:

• Conscious attention and focus

• Significant cognitive energy

• Time to think deliberately

• Willingness to override intuition

Your brain is lazy (technically “efficiency-optimising”). System 2 is metabolically expensive. So your brain defaults to System 1 whenever possible and only grudgingly engages System 2 when forced.

Most of the time, you think you’re using System 2 when you’re actually using System 1 with post-hoc rationalisation. System 2 often just provides logical-sounding explanations for conclusions that System 1 already reached intuitively.

The Implication for Leaders

What This Means for Decision-Making

1. Most of your decisions are made by System 1. This includes decisions that feel deliberate and analytical. You think you’re reasoning carefully, but you’re often rationalising intuitive judgments.

2. System 1 is often right, but doesn’t know when it’s wrong. Experienced judgment is valuable. Intuition based on pattern recognition from thousands of similar situations is often accurate. But System 1 feels equally confident when it’s wrong.

3. Confidence is not correlated with accuracy. You can be very confident and very wrong at the same time. The times you feel most certain are often the times you most need to slow down and engage System 2.

4. You can’t operate entirely in System 2. You’d be paralysed. You need to be strategic about when to override System 1 versus when to trust it.

5. Your team is subject to the same dynamics. They’re not being careless or lazy when they make biased decisions. They’re being human. Their System 1 is just as overconfident as yours.

The Core Challenge

The art of good decision-making is knowing:

• When your intuition is likely to be right (familiar situations with clear feedback loops where you have experience)

• When your intuition is likely to be wrong (novel situations, decisions under uncertainty, emotionally charged contexts)

• How to engage System 2 strategically for high-stakes decisions

• How to build team practices that help everyone catch bias

This starts with understanding the specific ways System 1 systematically misleads us.

PART 2: THE COGNITIVE BIASES THAT DISTORT STRATEGIC DECISIONS

There are hundreds of documented cognitive biases. We’re going to focus on the ones that most consistently cause problems for leaders making strategic decisions under uncertainty about restarting growth.

Confirmation Bias: Seeking Evidence That Supports What You Already Believe

What It Is

Once you form a hypothesis or belief, you unconsciously seek information that confirms it and ignore, dismiss, or fail to seek information that would contradict it.

How It Shows Up

You believe your product is superior to competitors. Now every positive customer comment gets noticed and remembered. Every complaint gets explained away as an edge case or user error. You selectively attend to evidence that confirms “we’re better” and filter out evidence that suggests otherwise.

You believe a particular person should be promoted. You notice every instance of their good work. You miss or rationalise their mistakes. You interpret ambiguous situations in ways that support your belief. By the time you make the promotion decision, you’ve assembled a compelling case, made entirely of cherry-picked evidence.

You believe pricing is why you’re losing deals. Every loss gets attributed to pricing. When prospects mention other factors, you don’t update your belief; you explain them away. “They said it was features, but really it was just price negotiation.” You never seriously consider that pricing might not be the actual problem.

Why It’s So Dangerous

Confirmation bias feels like objective analysis. You’re gathering evidence! You’re doing research! But you’re gathering it selectively, without realising that you’re doing so.

And it’s self-reinforcing. The more evidence you accumulate (even if cherry-picked), the more confident you become. Your confidence goes up without your accuracy improving; in fact, while your accuracy may be declining because you’re moving further from reality.

For CEOs Diagnosing Growth Problems

You’ve already formed beliefs about why growth stalled based on internal listening and initial analysis. Now, as you gather more information and make decisions, confirmation bias will cause you to:

• Notice evidence that confirms your diagnosis

• Discount evidence that challenges it

• Interpret ambiguous evidence as supporting your view

• Stop seeking disconfirming evidence

Even if your diagnosis is partially or completely wrong, confirmation bias will make you increasingly confident you’re right.

How It Manifests in Teams

Once your team senses what you believe, they’ll exhibit confirmation bias toward your beliefs. They’ll bring you evidence that supports what they think you want to hear. They’ll filter or spin evidence that contradicts it.

This isn’t malicious. It’s partly social (people want to agree with authority) and partly cognitive (they’re subject to confirmation bias too, and now they’re influenced by your framing).

The Antidote

The only reliable antidote is actively seeking disconfirming evidence. Ask: “What would have to be true for my belief to be wrong? What evidence would convince me I’m mistaken?”

If you can’t articulate what would change your mind, you’re not holding a testable belief; you’re holding an ideology. And ideologies are dangerous in business.

Anchoring: Over-Weighting the First Information You Encounter

What It Is

The first number or idea you encounter disproportionately influences your judgment, even if it’s arbitrary, irrelevant, or wrong. This initial “anchor” shapes your thinking about what’s reasonable, even when you consciously try to adjust.

How It Shows Up

A competitor’s price (£99/month) becomes your mental anchor for pricing, even though your product isn’t directly comparable. Now every pricing discussion revolves around that number. “$149 feels expensive because it’s 50% more than the competitor.” The anchor set the reference point.

Your first VP hire cost £180K. That becomes your anchor for what VPs “should” cost. When recruiters tell you the market rate is now £240K, it feels inflated, not because it is, but because you’re anchored to $180K.

Someone suggests a revenue target of £15M for next year. That becomes the anchor. Now every discussion of growth targets references that number, even if it was somewhat arbitrary when first mentioned.

Why It’s So Dangerous

Anchoring happens unconsciously and persists even when you know about it. Research shows that even random numbers influence judgments. In one famous study, people shown higher numbers on a roulette wheel subsequently gave higher estimates for unrelated questions.

In negotiations, strategy discussions, and planning, whoever sets the initial anchor often wins by shaping the entire conversation, without anyone realising the anchor is influencing them.

For CEOs Making Strategic Decisions

Your first strategic plan becomes the anchor for your strategic thinking. Your first revenue projection becomes the anchor for expectations. Your first organisation design becomes the anchor for what’s “normal.”

This makes it harder to make necessary changes later because any significant deviation from the anchor feels extreme, even if the anchor was wrong or circumstances have changed.

How It Manifests in Teams

The first person to speak in a discussion often sets the anchor for everyone else. If your CFO opens a pricing discussion by saying, “I think we should be at £50/user,” everyone else’s suggestions will cluster around that number.

If someone presents a three-year plan with specific milestones in year one, those milestones become anchors that are difficult to move, even if they were guesses.

The Antidote

When possible, generate your own independent estimates before hearing others’ numbers. In negotiations, try to set the anchor first. In group discussions, delay stating your view to avoid anchoring others.

When you notice you’ve been anchored, explicitly acknowledge it: “I realise we’re all anchored to [number]. Let’s step back and think about this independently.”

Availability Bias: Judging Likelihood by What Easily Comes to Mind

What It Is

You assess the probability or importance of something based on how easily examples come to mind, not on actual frequency, data, or logic. Things that are recent, dramatic, or emotional are more mentally “available” and thus feel more common or likely than they really are.

How It Shows Up

A key customer just churned last week. The loss is fresh and painful. Now you overestimate churn risk across your customer base and make decisions based on inflated fear.

A new hire just failed badly in the first 90 days. This dramatic failure is highly available in memory. Now you’re overly cautious about hiring, implementing onerous vetting processes that slow down necessary hiring.

You read a news article about a company that failed after a failed product pivot. That story is vivid and memorable. Now you’re more risk-averse about necessary product changes than the data warrants.

A major customer gave you harsh feedback two days ago. Their concerns feel hugely important because they’re mentally available. Meanwhile, ten satisfied customers said nothing (because satisfied customers rarely reach out), so their experiences aren’t available in your mind and don’t factor into your assessment.

Why It’s So Dangerous

Availability bias causes you to make decisions based on what’s salient and recent, not on what’s actually probable or representative. You overweight:

• Recent events over past patterns

• Dramatic events over mundane patterns

• Your own direct experiences over data

• Negative experiences over positive ones (negative is more emotionally salient)

This leads to strategy-by-anecdote: one customer complaint outweighs ten silent satisfied customers, one failed initiative outweighs three successful ones, one frightening competitive move outweighs a year of market data.

For CEOs Assessing Reality

Your assessment of “what’s really happening” is biased toward whatever happened recently or memorably. The exec who just resigned feels like evidence of morale problems, even if retention data shows you’re at industry norms. The deal you just lost feels like evidence of competitive weakness, even if your win rate is actually improving.

This causes you to overreact to recent events rather than respond to actual trends.

How It Manifests in Teams

Teams are subject to collective availability bias. If there was just a major incident or crisis, that incident looms large in every discussion. “Remember what happened with [thing]” becomes the argument against any similar action, even if that thing was a low-probability outcome.

War stories become organisational wisdom, even when they’re not representative of typical outcomes.

The Antidote

When making decisions, explicitly ask: “Is this based on what happened recently/memorably, or on actual data about frequency and probability?”

Keep data visible. Dashboards, regular reports, and tracking systems help counteract availability bias by making patterns visible rather than relying on memory.

Overconfidence: Believing You Know More Than You Do

What It Is

Humans are systematically overconfident about their knowledge, predictions, and abilities. We think our beliefs are more accurate than they are. We think our plans are more likely to succeed than they are. We underestimate uncertainty and the role of luck.

The Research

When people say they’re 90% confident about something, they’re right about 70% of the time. When experts make predictions in their field, they’re barely more accurate than random chance, yet they remain confident in their predictions.

Studies of business plans show entrepreneurs systematically underestimate how long things will take and overestimate the probability of success. Even after being told about this bias, they insist their plan is different.

How It Shows Up

You think you understand why growth stalled and what will fix it. You’re confident in your diagnosis. But there’s actually significant uncertainty—maybe you’ve identified contributing factors but missed root causes, or maybe external factors you don’t control will matter more than internal changes.

You project that a new go-to-market strategy will increase close rates by 30%. You’re confident in this projection. But you’re really just guessing, influenced by optimism and the planning you’ve done. The actual outcome could easily be anywhere from +10% to +50%.

You believe a leadership hire will work out. You’re confident after a thorough interview process. But leadership hires fail 40-50% of the time in the first 18 months, regardless of how confident people were during hiring.

Why It’s So Dangerous

Overconfidence causes you to:

• Underestimate how hard things will be

• Commit resources based on overconfident projections

• Take risks you wouldn’t take if you accurately assessed probability

• Stop seeking information because you think you already know enough

• Miss warning signs because you’re confident things are on track

And it prevents learning. If you’re overconfident about predictions, you don’t track whether you were right. You assume you were, and move on. So you never calibrate your confidence.

For CEOs Under Board Pressure

Board pressure creates incentives for overconfidence. Boards want confidence. They respond positively to CEOs who seem certain about their strategy. This creates pressure to express certainty even when genuine uncertainty exists.

But overconfidence increases the risk of strategic failure by leading you to commit fully to plans that should be tested incrementally.

How It Manifests in Teams

Your team will mirror your confidence level. If you project certainty, they’ll project certainty to you. Everyone will be confident and collectively wrong.

The teams that say “we’ve got this” with high conviction are often the teams that fail, because they’re not seeing risks or adapting to early warning signals.

The Antidote

Practice confidence calibration. When you make predictions, write down your confidence level (e.g., “I’m 70% confident close rates will improve by at least 20%”). Then track whether you were right.

Over time, you’ll learn whether you’re well-calibrated (70% confident = right 70% of the time) or overconfident (70% confident = right 50% of the time). This teaches you to adjust.

Sunk Cost Fallacy: Continuing Because You’ve Already Invested

What It Is

You continue investing in something not because the future returns justify it, but because you’ve already invested so much that walking away feels like admitting failure and “wasting” past investment.

How It Shows Up

You’ve invested £2M in a product that’s showing weak market traction. Rationally, the question should be: “Will future investment generate sufficient return?” But emotionally, the question becomes: “Do we write off the £2M we already spent?”

So you keep funding it, even though the evidence suggests it won’t work, because stopping feels like admitting the £2M was wasted. You’re throwing good money after bad, but it feels like protecting past investment.

You hired an executive who’s struggling six months in. You invested heavily in recruiting, negotiating, onboarding, and integrating them. Walking away now means admitting that the investment didn’t work out. So you give them “more time” repeatedly, even as performance problems persist and costs accumulate.

You publicly committed to a strategic direction. Now evidence suggests it’s not working, but changing course means admitting the strategy was wrong. So you stick with it longer than you should because sunk costs aren’t just financial, they’re reputational and emotional.

Why It’s So Dangerous

Rationally, sunk costs should be irrelevant to decision-making. Money already spent is gone. The only question is: Does the marginal benefit of future investment exceed the marginal cost?

But humans hate loss. We hate admitting we made a mistake. We hate “wasting” past effort. So we irrationally continue bad investments to avoid the pain of recognising losses.

This causes you to waste future resources on things that should be killed, while starving things that deserve investment.

For CEOs Making Portfolio Decisions

Your product, initiative, and people portfolios all have sunk costs attached. The longer something has existed and the more you’ve invested, the harder it is to kill, even when evidence clearly shows it’s not working.

This creates organisational drag as you keep supporting legacy products, failed initiatives, and struggling people longer than you should.

How It Manifests in Teams

Teams exhibit strong sunk-cost bias toward projects they’ve invested in. The team that built the product will find reasons to keep supporting it long past when it should be killed. The team that championed the strategy will defend it past the point where evidence shows it’s not working.

This creates organisational inertia and makes it hard to admit mistakes and change course.

The Antidote

Separate past investment from future decisions. Explicitly frame decisions as: “Ignoring what we’ve already spent, does future investment make sense?”

Create clear criteria for continuing or killing initiatives: “We’ll continue if we see [specific evidence] by [specific date]. If we don’t, we stop.” This makes stopping automatic rather than a decision each time.

The Planning Fallacy: Underestimating Time and Difficulty

What It Is

People consistently underestimate how long projects will take and how much they’ll cost, even when they know about past projects that took longer and cost more than expected.

The Research

Studies show that most projects take 50-100% longer than initially estimated. This isn’t because people are bad at estimating—it’s a systematic bias. Even when people are told about the planning fallacy and shown data on past projects, they still produce optimistic estimates.

How It Shows Up

You estimate implementing a new CRM will take three months. It takes nine. You estimate hiring for a key role will take six weeks. It takes four months. You project that a product launch will generate £5M in first-year revenue. It generates £1M.

You’re not being unrealistic on purpose. You genuinely believe the estimate when you make it. You think: “Previous projects had problems we’ve now addressed. This time we have better people, clearer plans, and more commitment. This time will be different.”

But it’s not different. The same forces that delayed previous projects will delay this one, but you don’t account for them in your planning.

Why It’s So Dangerous

The planning fallacy causes:

• Strategic plans based on unrealistic timelines

• Board expectations you can’t meet

• Resource commitments that don’t account for reality

• Team burnout as people try to meet impossible deadlines

• Missed dependencies and cascading delays

And it’s self-reinforcing. When things take longer than expected, you don’t update your mental model of how long things take. You explain it as specific to that situation (“bad luck,” “unexpected issues”), then make the same optimistic projection next time.

For CEOs Setting Board Expectations

Your board plan outlines specific milestones and dates. Those dates are almost certainly too aggressive because of the planning fallacy. But once communicated, they become commitments.

Now you have three choices: miss the commitments (board loses confidence), burn out your team trying to meet impossible deadlines, or lower the quality to hit dates. All are bad outcomes driven by the initial planning fallacy.

How It Manifests in Teams

Everyone is subject to planning fallacy. Your product team underestimates development time. Your sales team overestimates how quickly they can ramp. Your marketing team underestimates how long it takes campaigns to show results.

When you roll these optimistic estimates into a company plan, the compounding effect makes the overall plan wildly unrealistic.

The Antidote

Use “reference class forecasting.” Instead of bottom-up estimation (“How long will this specific project take?”), ask: “How long have similar projects taken in the past?” Use the actual historical data as your baseline, then adjust only if you have specific, concrete reasons that this project is genuinely different.

Add explicit buffers to estimates. If something feels like it will take 3 months, plan for 4-5. You’ll rarely regret the buffer.

How These Biases Interact and Compound

Here’s what makes cognitive bias especially dangerous for strategic decision-making: These biases don’t operate in isolation. They interact and reinforce each other in pernicious ways.

The Compounding Cascade

You develop a strategic hypothesis (let’s say: “We need to move upmarket”). You anchor to this idea. Confirmation bias causes you to seek evidence supporting it and ignore evidence against it. Recent successes with larger customers (availability bias) feel like proof. You become overconfident the strategy will work. Your plan underestimates how long the transition will take (planning fallacy). When early results are mixed, you stick with it because you’ve already invested so much (sunk cost).

At every stage, bias reinforces bias. And it all feels like good judgment, experience, and commitment. It doesn’t feel like bias.

Team Dynamics Amplify Bias

When biases operate at the individual level, they’re dangerous. When they operate at the team level, they’re catastrophic because:

1. Social reinforcement – Team members mirror each other’s biases

2. Authority gradient – People defer to the CEO’s biases to avoid conflict

3. Shared narratives – Teams construct collective stories that explain away disconfirming evidence

4. Groupthink – Desire for consensus suppresses dissent and critical thinking

This is how smart, experienced leadership teams make terrible strategic decisions. Not because any individual is incompetent, but because collective cognitive biases go unchecked.

PART 3: THE BELIEF VS. FACT DISTINCTION

Why This Matters More Than Any Specific Bias

Before we get to solutions, we need to address something fundamental: the distinction between belief and fact, and why clarity about it is essential for good decision-making.

Most of What You “Know” Is Actually Belief

Right now, you hold many beliefs about your business. Examples:

• “Our ICP is mid-market companies in healthcare”

• “We lose deals primarily because of pricing”

• “This person will be successful in that role”

• “Our product is better than competitor X”

• “Moving upmarket is the right strategy”

• “Our churn problem is caused by poor onboarding”

Here’s what’s critical to understand: All of these are beliefs, not facts.

Some might be well-founded beliefs based on significant evidence. Some might be accurate beliefs that happen to match reality. But they’re still beliefs—interpretations of evidence, predictions about the future, theories about causation.

Facts are different:

• “We closed 23 deals last quarter” – This is a fact (assuming accurate data)

• “Our annual churn rate is 18%” – This is a fact

• “Competitor X charges $99 per seat” – This is a fact

• “This person has 10 years of experience” – This is a fact

Facts are directly observable or measurable. Beliefs are interpretations, predictions, or theories.

The problem: We treat beliefs as facts. We say “Our ICP is mid-market” with the same confidence as “We closed 23 deals last quarter.” We make decisions based on beliefs as if they were established truths. And we stop testing beliefs once we’re confident in them.

The Confidence Calibration Problem

Not only do we mistake beliefs for facts, but we’re also bad at assessing our confidence in our beliefs. We feel more certain than we should.

The research is clear:

When people say they’re 90% confident in a belief, they’re actually right about 70% of the time. When experts make predictions in their domain, they overestimate their accuracy significantly.

This is overconfidence bias, but it’s so fundamental it deserves special attention: Your subjective feeling of confidence is not a reliable indicator of objective accuracy.

You can feel very confident and be completely wrong. You can feel uncertain and be exactly right. There’s only a weak correlation between confidence and accuracy.

Why This Matters

When you operate based on unjustified confidence in beliefs you treat as facts, you:

1. Stop seeking disconfirming evidence – Why test something you “know” is true?

2. Miss warning signs – You explain away evidence that contradicts your belief

3. Make brittle strategies – Built on assumptions you never validate

4. Create defensive cultures – Challenging beliefs feels like challenging established facts

5. Fail to update – You don’t revisit beliefs as new evidence emerges

The Solution: Explicit Belief Statements with Confidence Levels

The practice of explicitly stating beliefs and confidence levels might feel awkward at first. But it’s transformational for decision-making.

Instead of saying: “Our ICP is mid-market companies”

Say: “I believe our ICP is mid-market companies, and I’m about 70% confident in that based on our best customers, but I’m less certain whether we can win in that segment at scale given our current product and GTM motion.”

Instead of saying: “This hire will work out”

Say: “I believe this person has the capabilities we need. I’m 60% confident they’ll succeed in this specific role given our culture, stage, and constraints. The main uncertainty is whether they can adapt to our level of ambiguity.”

Instead of saying: “We lose deals because of pricing”

Say: “I believe pricing is a factor in lost deals, maybe 65% confident it’s the primary factor based on sales feedback. But I’m less confident because win/loss data is often unreliable and we haven’t done systematic competitive analysis.”

What This Accomplishes

1. Forces you to think about your actual confidence level – Usually lower than the certainty you feel

2. Creates space for updating beliefs – Beliefs can change with new evidence; “facts” can’t

3. Invites testing and validation – Low confidence beliefs should be tested

4. Signals appropriate caution – Team understands uncertainty around decisions

5. Models intellectual humility – Shows it’s okay to be uncertain

6. Enables learning – You can track whether your confidence was justified

Thinking in Bets: The Annie Duke Framework

Annie Duke, a professional poker player who transitioned to studying decision-making, describes this as “thinking in bets.” Every belief is a bet with varying odds. Being a good decision-maker means accurately assessing the odds and being willing to update them as new information comes in.

In poker

• You never know if you’ll win the hand

• You only know the probability based on cards and opponent behavior

• Good players make decisions based on probability, not certainty

• Results don’t always validate decisions (you can make a good bet and lose)

• Good players update their probability assessments as new cards are revealed

In business

• You never know if your strategy will work

• You only have beliefs with varying confidence levels

• Good leaders make decisions based on their best probability assessment

• Results don’t always validate decisions (you can make good decisions that fail due to luck)

• Good leaders update their beliefs as new evidence emerges

The mindset shift

From: “This is the right strategy” (certainty about unknowable future) To: “This strategy has the highest probability of working given what we know now” (probabilistic thinking)

From: “I know why we’re losing deals” (certainty about complex causation) To: “I’m 65% confident pricing is the main factor, which means there’s 35% chance I’m wrong or missing something” (calibrated confidence)

This doesn’t mean being wishy-washy. It means being honest about uncertainty while still making decisions and committing to executing. You can be both uncertain and decisive at the same time.

Building the Practice: A Simple Framework

For Every Major Belief or Decision

1. State the belief explicitly “I believe [specific belief statement]”

2. State your confidence level “I’m [percentage]% confident in this”

3. Identify the basis for confidence “Based on [what evidence or reasoning]”

4. Identify what would increase confidence “I’d be more confident if I saw [what evidence]”

5. Identify what would change your mind “I’d decrease confidence or change this belief if [what evidence]”

Example:

“I believe our churn problem is primarily caused by poor onboarding experience. I’m 60% confident in this based on qualitative feedback from churned customers and the correlation between time-to-first-value and retention. I’d be more confident if we saw that improving onboarding metrics actually decreased churn in a cohort test. I’d change this belief if we improved onboarding significantly and churn didn’t decrease, or if deeper analysis showed churned customers actually completed onboarding successfully but churned for other reasons.”

This framework turns vague certainty into testable hypotheses.

PART 4: PERSONAL PRACTICES FOR BETTER DECISION-MAKING

Now that you understand how your brain works and the specific ways it misleads you, let’s talk about practical techniques for improving your own decision-making.

Practice 1: The Pre-Mortem for Major Decisions

What It Is

Before committing to a major decision, imagine it’s 12-18 months in the future and the decision was a complete disaster. Now work backwards: What went wrong?

This technique, developed by psychologist Gary Klein, is one of the most effective bias-busting practices for strategic decision-making.

How It Works

You’re about to commit to a significant strategic shift, say, moving upmarket from SMB to mid-market. Before you fully commit:

Gather your leadership team. Say: “It’s 18 months from now. Our upmarket strategy failed badly. We’re worse off than before. Revenue is down, team morale is shot, we’re burning cash, and the board has lost confidence. What happened?”

Then let people brainstorm failure scenarios. Don’t defend or dismiss. Just capture them.

Example Pre-Mortem Outputs

• “We couldn’t actually deliver enterprise features on the timeline we promised”

• “Our sales team couldn’t sell to larger companies—the motion is completely different”

• “Implementation time tripled, which killed our unit economics”

• “We lost our SMB customers faster than we acquired mid-market ones”

• “Mid-market buyers wanted integrations we couldn’t build”

• “We underestimated how long sales cycles would be and ran out of cash”

• “Our product positioning was wrong for the enterprise buyer persona”

Why It Works

Pre-mortems counteract overconfidence and confirmation bias by:

1. Legitimising doubt – Gives people permission to voice concerns without seeming negative

2. Surfacing suppressed information – People know risks they haven’t voiced; this gives them a safe way to surface them

3. Forcing you to engage System 2 – You have to think deliberately about failure modes

4. Identifying blind spots – The team collectively sees risks you individually missed

What To Do With Pre-Mortem Results

After brainstorming failure scenarios, evaluate:

• Which are actually likely? (Be honest—don’t dismiss things as “we’ll figure it out”)

• Which would be catastrophic if they happened?

• What early warning signs would indicate these risks are materialising?

• What can we do upfront to mitigate the most likely/dangerous risks?

• Does this change our decision or approach?

Often, you’ll still proceed with the decision. But you’ll have:

• Identified risks to mitigate

• Established early warning indicators to monitor

• Prepared contingency plans

• Reduced overconfidence about the ease of success

When To Use It

Use pre-mortems for:

• Major strategic shifts

• Large resource commitments

• Key hires at the leadership level

• Significant product launches

• Major operational changes

Don’t use it for every decision; that would be analysis paralysis. But for genuinely high-stakes, hard-to-reverse decisions, the 60-90 minutes invested in a pre-mortem is time very well spent.

Practice 2: Actively Seek Disconfirming Evidence

What It Is

Deliberately search for evidence that would prove your belief wrong, not just evidence that confirms it.

Why It’s Hard

Confirmation bias is not a lazy mistake you can correct by trying harder. It’s a systematic feature of how your brain processes information. Even when you consciously try to be objective, you’ll unconsciously seek confirming evidence.

The only reliable way to counteract confirmation bias is to make seeking disconfirming evidence explicit and systematic.

How To Do It

For Every Major Belief, Ask

1. “What evidence would prove this belief wrong?”

2. “Have I actually looked for that evidence?”

3. “If I found disconfirming evidence, would I update my belief or explain it away?”

Example

Belief: “We lose deals primarily because our pricing is too high”

Questions

• What evidence would prove this wrong?

  • If we tested lower pricing, and win rates didn’t improve

  • If systematic win/loss analysis showedthat  other factors were more often cited

  • If we lost to competitors with higher pricing

• Have I looked for that evidence?

  • Have we tested pricing?

  • Have we done rigorous win/loss interviews with a protocol that surfaces real reasons?

  • Have we compared our pricing to competitors we actually lose to?

• Would disconfirming evidence change my mind?

  • If customers said price wasn’t the issue, would I believe them, or would I assume they’re not being honest?

  • If a pricing test showed no impact, would I accept that, or would I explain it away as “the test wasn’t run right”?

The Falsifiability Test

If you can’t articulate what evidence would change your mind, you’re not holding a testable belief; you’re holding an ideology.

Beliefs that can’t be falsified aren’t useful for decision-making because you can’t learn whether they’re right or wrong. You’ll just keep believing them regardless of evidence.

Creating a Culture of Disconfirmation

As a leader, model this explicitly:

“Here’s what I believe. But I want us to actively look for evidence that I’m wrong. If we find [X], that would make me question this belief. So let’s specifically check for [X].”

This does two things:

1. Actually helps you find disconfirming evidence

2. Signals to your team that challenging beliefs is not just okay—it’s what you want

Practice 3: Decision Journaling and Review

What It Is

Keep a written record of major decisions: what you decided, what you believed would happen, why you believed it, and your confidence level. Then review these decisions after sufficient time has passed to see outcomes.

Why It Works

Without explicit review, you’ll never learn to calibrate your confidence. You’ll remember decisions that worked out and forget ones that didn’t (outcome bias). You’ll remember feeling less confident than you actually were (hindsight bias). You’ll unconsciously revise your memory of what you predicted.

A decision journal creates an external record you can’t revise. It forces honest assessment of your decision quality.

What To Capture

At Decision Time

• Date and Context: When you decided, what was happening

• The Decision: What you chose to do

• Alternatives Considered: What else you considered

• Your Reasoning: Why you chose this option

• Your Prediction: What you thought would happen

• Your Confidence: How confident you were (percentage)

• Key Uncertainties: What you were most uncertain about

• What Would Change Your Mind: What evidence would make you reconsider

At Review Time (6-12 months later)

• What Actually Happened: Outcomes vs. predictions

• Were You Right: Did outcomes match predictions?

• Was Your Confidence Justified: If 70% confident, were you right about 70% of things?

• What Did You Miss: What didn’t you consider or what were you wrong about?

• What Did You Learn: What will you do differently next time?

Example Decision Journal Entry

Decision Date: March 15, 2025

Decision: Hire Sarah Johnson as VP Sales

Context: Current VP leaving, need to rebuild sales org, preparing for Series B

Alternatives

• Promote internal sales director (Jane)

• Hire different external candidate (Mike from competitor)

• Contract with fractional VP while searching longer

Reasoning: Sarah has exactly the experience we need, built a team from 5 to 40 reps at a similar-stage company, knows our ICP, culture fit seemed strong in interviews, references were excellent

Prediction: Sarah will succeed in ramping the sales team, improving win rates by Q3, and helping us hit £25M ARR by year-end

Confidence: 75% confident she’ll be successful in the role

Key Uncertainties

• Whether she can adapt from la arger company to our scrappier environment

• Whether her management style will work with our existing team

• Whether our product is actually ready for the aggressive scaling she’s designed for

What Would Change My Mind: If, after 90 days, she hasn’t built rapport with the team, or if she’s struggling with our level of ambiguity and lack of established process

Review Date: December 15, 2025 (9 months later)

What Actually Happened:

• Sarah struggled significantly in the first 6 months

• Team turnover was higher than expected (3 reps left)

• Win rates didn’t improve materially

• We’re at £22M ARR, not £25M

• Sarah and I had a tough conversation in October about fit

• She left in November; we promoted Jane

Was I Right: Partially wrong, she didn’t succeed as I predicted

Was My Confidence Justified: No, I was 75% confident but wrong, suggesting I was overconfident

What Did I Miss

• Underestimated culture adaptation challenge—she was used to more structure

• Didn’t sufficiently validate her ability to thrive in ambiguity

• Didn’t weight “promote from within” option heavily enough

• References were all from larger, more established companies—should have worried about that

What I Learned

• I’m overconfident about leadership hires—need to calibrate down (60% confident should be my baseline)

• Need to weight culture fit/stage fit more heavily than track record

• Should do more thorough reference checks, specifically about how people handle ambiguity

• Promoting strong internal candidates deserves more serious consideration

The Discipline

Decision journaling only works if you actually do the review. This is painful—reviewing past decisions where you were wrong is uncomfortable. But it’s the fastest way to improve judgment.

Commit to reviewing major decisions after 6-12 months. Schedule it. Put it on your calendar as a recurring task.

Practice 4: Create Space Between Impulse and Decision

What It Is

For high-stakes decisions, build in a mandatory delay between your initial reaction and final commitment. This creates space for System 2 to engage.

Why It Works

Your System 1 reaction happens instantly. You feel like you know the right answer. But that immediate conviction is often influenced by recency, emotion, framing, and bias.

By forcing a delay, you give System 2 time to engage and potentially override System 1 if needed.

How To Implement

Create Personal Rules:

“I don’t make major hiring decisions the same day as the final interview, no matter how strong my conviction. I sleep on it and decide the next day.”

“I don’t approve budget increases over $50K in the meeting where they’re requested. I take 24 hours to think about it.”

“I don’t send important, emotionally charged emails immediately. I draft them, then wait an hour before sending.”

Use Forcing Functions

• Sleep on it: “Let me think about this overnight and we’ll discuss tomorrow”

• Seek external input: “I want to talk this through with [mentor/board member] before deciding”

• Request more analysis: “Can you put together a one-page pros/cons before we decide?”

• Set decision date: “We’ll make this decision Friday, which gives us three days to think it through”

When To Use

Use deliberate delay for:

• High-stakes decisions (large resource commitments, strategic shifts)

• Irreversible decisions (can’t easily undo if wrong)

• Emotionally charged decisions (firing, conflict, crisis response)

• Decisions where you feel especially certain (overconfidence check)

Don’t use it for:

• Low-stakes, easily reversible decisions (kills momentum)

• Time-sensitive decisions where delay has real cost

• Decisions where you genuinely have sufficient information and System 2 has already engaged

Practice 5: Cultivate Advisors Who Challenge You

What It Is:

Identify specific people in your life whose explicit role is to challenge your thinking, not to validate it.

Why It’s Essential

No amount of personal discipline will catch all your biases. You need external perspective from people who:

• Are smart and experienced

• Understand your context, but aren’t embedded in it

• Care about your success

• Are you willing to tell you uncomfortable truths

• You trust enough to be vulnerable with

Who They Might Be

• Board member (especially independent directors)

• ZOKRI’s consultants and mentors

• Peer CEOs from this community

• Former colleague you respect

• Advisor with specific domain expertise

How To Use Them

Don’t seek validation. Don’t present your decision and ask “What do you think?” hoping they’ll agree.

Seek challenge. Present your decision and specifically ask:

• “What am I missing?”

• “Where is my thinking weak?”

• “What’s the strongest case against this?”

• “What should I be more worried about?”

• “If you were in my shoes, what would concern you?”

• “What would you do differently?”

Model Receiving Challenge Well

• Thank them for pushback (even if it’s uncomfortable)

• Don’t get defensive or immediately rebut

• Ask clarifying questions to understand their concern

• If you disagree, explain why—but only after genuinely considering their point

Signal That You Want Real Input

“I’m pretty convinced about this decision, but I know I could be wrong. I specifically need you to challenge me, not validate me. Where am I being overconfident or biased?”

This gives them permission to disagree with you, which many people are reluctant to do with CEOs.

The Trap

The trap is seeking advice from people who think like you, will validate your decisions, or are afraid to disagree with you. That’s not advice, that’s an echo chamber.

You need at least one advisor whose job is to make you uncomfortable and make you think harder.

Practice 6: Quantify Uncertainty with Confidence Ranges

What It Is

Instead of single-point predictions, use confidence ranges that express your uncertainty explicitly.

Why It Works

Single-point predictions hide uncertainty and create false precision. “We’ll close $25M this year” sounds definitive but is actually a guess with huge uncertainty around it.

Confidence ranges make uncertainty explicit and improve calibration.

How To Do It

Instead of: “We’ll hit £25M ARR next year”

Say: “I’m 50% confident we’ll be between £22M and £27M, 80% confident we’ll be between £20M and £30M”

This communicates:

• Your best estimate (the midpoint)

• Your uncertainty (the range width)

• Your confidence in different outcomes

For Major Projections

Give three numbers:

• Optimistic case (20% probability)

• Base case (60% probability)

• Pessimistic case (20% probability)

Example: “Our revenue projection for next year is £25M base case, with £30M optimistic and £20M pessimistic, giving us an 80% confidence interval of £20M-$30M.”

This forces you to think about the range of likely outcomes, not just your single-point forecast. And it makes clear to others (board, team) that there’s significant uncertainty.

Calibrating Over Time

Track your confidence ranges. If you say you’re 80% confident something will fall in a range, it should fall in that range 80% of the time.

If you consistently say 80% and are right 60% of the time, you’re overconfident. Your ranges are too narrow. You need to widen them to reflect actual uncertainty.

If you say 80% and are right 95% of the time, you’re underconfident. Your ranges are too wide. You can be more precise.

PART 5: BUILDING A TEAM THAT DECIDES WELL

Individual good decision-making is necessary but not sufficient. You need your entire leadership team to make good decisions. And that requires building team practices and culture that account for human cognitive limitations.

Create Psychological Safety for Dissent

Why This Is First

Everything else depends on this. If people are afraid to challenge assumptions or disagree with you, all the techniques and frameworks in the world won’t help. Biases will go unchecked. Bad decisions will get unanimous support.

Psychological safety means people believe they can speak up with concerns, disagreements, or bad news without being punished, embarrassed, or marginalized.

What It Looks Like

• People voice disagreement with you in meetings (respectfully but clearly)

• People bring up concerns even when no one else has

• People challenge each other’s assumptions, not just yours

• People admit mistakes and uncertainty without defensiveness

• Bad news surfaces quickly because people aren’t afraid to deliver it

What It Doesn’t Mean

Psychological safety is not:

• Everyone is being nice and agreeable

• Avoiding difficult conversations

• Consensus decision-making

• No accountability

You can have high psychological safety and high standards simultaneously. In fact, you need both for good decision-making.

How To Build It

1. Explicitly Invite Disagreement

In meetings: “I need someone to make the case against this. Who can play devil’s advocate?”

Before decisions: “Before we decide, I want to hear concerns. What are we missing? What could go wrong?”

2. Reward People Who Speak Up, Even If They’re Wrong

When someone voices a concern: “I appreciate you raising that. Let’s talk through it.”

When someone admits uncertainty: “Thanks for being honest about what you don’t know. That’s helpful.”

When someone disagrees: “Good point. Tell me more about why you think that.”

3. Never Shoot the Messenger

When someone brings bad news: “Thank you for surfacing this quickly. Let’s figure out what to do.”

Not: “Why didn’t you catch this earlier?” or “This is a problem” (said with anger/frustration).

The fastest way to kill psychological safety is punishing people who deliver bad news. Then bad news goes underground until it’s a crisis.

4. Model Changing Your Mind

When you receive new information or a compelling argument: “You know what, I was wrong about this. Here’s why I’m changing my view.”

This shows that:

• Changing your mind based on evidence is a strength, not a weakness

• You’re genuinely open to persuasion

• It’s safe for others to change their minds too

5. Model Admitting Uncertainty

Instead of projecting certainty: “I’m uncertain about this. I think X is more likely, but I could be wrong.”

This signals:

• Uncertainty is okay

• You want input to reduce uncertainty

• People don’t need to pretend certainty to look competent

What Kills Psychological Safety

• Ridiculing ideas or concerns

• Getting defensive when challenged

• Punishing dissent (even subtly through body language or tone)

• Rewarding agreement and conformity

• Making decisions and then asking for input (pretend consultation)

Watch for these patterns in yourself and actively counteract them.

Use Structured Decision Processes for Major Decisions

Why Structure Helps

Unstructured discussions default to System 1 thinking with minimal bias checking. The loudest voice or the highest status person often wins. Anchoring happens unconsciously. Confirmation bias goes unchallenged.

Structured processes force System 2 engagement and reduce bias.

The Basic Structure for Strategic Decisions

Step 1: Frame the Decision Clearly (5-10 minutes)

What exactly are we deciding? What’s in scope and out of scope? What’s the timeline for the decision?

Bad framing: “Should we change our strategy?” Good framing: “Should we shift from SMB to mid-market as our primary ICP, which would mean changes to product roadmap, sales approach, and marketing? We need to decide by end of month to inform H2 planning.”

Step 2: Generate Options (15-20 minutes)

Brainstorm alternatives. Get multiple options on the table before evaluating any of them.

Rules:

• Don’t critique ideas during generation

• Capture all options, even ones that seem unlikely

• Push for at least 3-4 distinct options (not just yes/no)

• Consider “do nothing” as an explicit option

This counteracts anchoring (first option doesn’t dominate) and confirmation bias (you’re forced to generate alternatives to your preferred choice).

Step 3: Identify Evaluation Criteria (10 minutes)

What factors matter for this decision? How will we evaluate options?

Example criteria:

• Revenue impact potential

• Resource requirements

• Implementation timeline

• Risk level

• Strategic alignment

• Competitive positioning

Get the criteria explicit before evaluating options. Otherwise, you’ll unconsciously weight criteria that favour your preferred option.

Step 4: Seek Disconfirming Evidence for Each Option (20-30 minutes)

For each option, deliberately ask:

• “What’s the strongest case against this?”

• “What could go wrong?”

• “What assumptions are we making?”

• “What would have to be true for this to fail?”

Assign someone to explicitly argue against each option, including your preferred one.

This counteracts confirmation bias by forcing you to genuinely consider downsides.

Step 5: State Confidence Levels (10 minutes)

For your preferred option:

• “How confident are we that this will work?”

• “What are we most uncertain about?”

• “What evidence would increase our confidence?”

If confidence is low (<60%), should you be making this decision now or doing more validation first?

Step 6: Decide and Document (10 minutes)

Decision: What did we decide?

Rationale: Why did we decide this?

Prediction: What do we think will happen?

Confidence: How confident are we?

What We’ll Monitor: What early indicators will tell us if this is working or not?

Review Date: When will we formally review this decision?

This documentation creates the foundation for learning. Without it, you’ll never know if you made a good decision or just got lucky.

Total Time: 70-110 minutes

This seems like a lot of time. But for genuinely strategic decisions (where will we focus? what will we build? how will we restructure?), this time investment is trivial compared to the cost of wrong decisions.

Disagree and Commit

What It Is

A cultural practice where people can fully disagree with a decision during the decision process, but once the decision is made, everyone commits to executing it fully—no undermining, no half-hearted implementation.

Why It Matters

Perfect consensus is rare and often undesirable. Smart people looking at the same data will reach different conclusions. Waiting for everyone to agree causes paralysis.

But making a decision when some people disagree can create problems:

• People undermine decisions they disagreed with (passive resistance)

• Teams re-litigate decisions constantly (never moving forward)

• People say “I told you so” when things go wrong (instead of helping)

Disagree and commit solves this by creating clear phases:

Before Decision

• Everyone’s input is welcome

• Disagreement is encouraged

• The debate is vigorous

• Multiple views are represented

After Decision

• Debate ends

• Everyone commits fully to execution

• No undermining or passive resistance

• If the decision was wrong, we’ll find out through execution and adjust

How To Implement

Make the practice explicit: “We use disagree and commit here. That means we’ll have a full debate before decisions, but once we decide, everyone commits fully. You don’t have to agree, but you do have to commit. Can everyone do that?”

Before deciding, ensure everyone was heard: “Before I make this call, I want to make sure everyone who had concerns gets to voice them. Did anyone not get a chance to share their perspective?”

People can commit to execution even if they disagree, but only if they feel heard.

When deciding, name the disagreement: “I know Sarah and Mike have concerns about this approach. I’ve heard those concerns, and I’m making a different call. But I need you both to commit to executing this fully. Can you do that?”

This makes the disagreement explicit but requires commitment.

After deciding, hold people accountable to commitment: If someone is undermining the decision (through passive resistance, negative comments, or half-hearted execution), address it directly:

“In the meeting, you committed to this approach even though you disagreed. But I’m seeing [specific behaviours] that feel like you’re not fully committed. What’s going on?”

The Alternative

Without disagreeing and committing, you get:

• Endless debate with no decisions

• Decisions were undermined by people who disagreed

• Fake consensus where people say they agree but don’t

• Slow, tentative execution because people aren’t committed

Red Team / Blue Team for Strategic Decisions

What It Is

For major strategic decisions, assign teams to argue opposite sides. One team makes the best possible case for option A. Another team makes the best possible case for option B. Then you debate.

Why It Works

This counteracts confirmation bias and overconfidence by:

• Forcing genuine consideration of alternatives (not just token alternatives)

• Making someone actually develop the case against your preferred option

• Revealing when a decision is closer than you thought (both sides make strong cases)

• Surfacing concerns that might otherwise stay hidden

How To Implement

1. Frame the Decision as a Choice Between Alternatives

“Should we move upmarket (Option A) or stay focused on SMB and optimize (Option B)?”

2. Assign Teams to Each Option

Deliberately assign some people to argue for the option they don’t naturally prefer. This prevents the exercise from being just “people arguing for what they already believe.”

“Sarah and Mike, you’ll make the case for moving upmarket. Jane and Tom, you’ll make the case for staying SMB-focused. I want your absolute best arguments, with data and logic. Pretend you’re trying to convince sceptical board members.”

3. Give Time to Prepare

Give teams 3-5 days to build their case. They should:

• Gather evidence

• Anticipate counterarguments

• Develop their strongest logical case

• Prepare to present in 15-20 minutes

4. Hold the Debate

Each side presents its case (15-20 minutes). Then rebuttals (10 minutes each side). Then open discussion.

5. Make a Decision

After hearing both cases fully developed, decide. The decision-maker (you) should explain:

• What swayed you

• Where the cases were close

• What uncertainties remain

• What you’ll monitor to see if the decision was right

When This Reveals Problems

Sometimes, after Red Team / Blue Team, you realise:

• Both options have serious flaws (need a third option)

• The decision is much closer than you thought (be more humble about prediction)

• You were missing key information (do more diligence before deciding)

• Your initial preference was based on weak reasoning (update your view)

All of these are valuable discoveries before you commit resources.

Run Experiments, Don’t Just Decide

What It Is

When possible, test assumptions before fully committing. Run small experiments that provide evidence before making big bets.

Why It’s Better

Experiments turn beliefs into evidence. They:

• Reduce the risk of being wrong

• Provide data for decisions instead of opinions

• Let you fail small instead of failing big

• Create organisational learning about what works

How To Think About It

Most strategic decisions are really decisions under uncertainty. You have a hypothesis (“moving upmarket will work”), but you don’t know if it’s true.

Traditional approach: Debate the hypothesis, gather opinions, make a decision, commit fully.

Experimental approach: Design a test of the hypothesis, run it small-scale, look at results, and then decide.

Examples:

Hypothesis: “Lower pricing will increase win rates”

Experiment: Test discounted pricing with a segment of prospects for one month. Compare win rates and deal velocity to the control group.

Learn: Does pricing actually affect win rates? If yes, how much? Are the economics still favourable at a lower price?

Hypothesis: “We can successfully sell upmarket”

Experiment: Have two sales reps focus exclusively on mid-market prospects for a quarter. Track: Can they generate a pipeline? What’s the close rate? What’s the sales cycle? What are the objections?

Learn: Can we actually win these deals? What needs to be different to be competitive? What’s the true cost of sale?

Hypothesis: “This new onboarding experience will reduce churn”

Experiment: Implement new onboarding for 20% of new customers. Compare activation rates and retention to the control group.

Learn: Does the new experience actually improve outcomes? By how much? What are the unexpected effects?

When You Can’t Experiment

Obviously, not everything can be tested incrementally. Some decisions are one-way doors, you can’t test “should we fire the VP Sales?” or “should we pivot the product?”

But more things are experimentable than most leaders assume. The bias is toward “we need to decide and commit” when often you could “we could test this first.”

Ask: “Is there a way to test this assumption before we fully commit? What would a small-scale experiment look like?”

Post-Decision Reviews: Organisational Learning

What It Is

Three to six months after major decisions, explicitly review them as a team. What did we expect? What actually happened? Were we right? If not, why not?

Why It’s Essential

Teams that never review decisions:

• Keep making the same mistakes

• Never calibrate their confidence

• Don’t learn what process worked or didn’t work

• Can’t distinguish luck from skill

Teams that review decisions systematically:

• Get better at forecasting

• Identify patterns in what works

• Learn from both successes and failures

• Build organisational capability for judgment

How To Do It

Schedule Reviews in Advance

When you make a decision, schedule the review for 3-6 months from now. Otherwise ,it won’t happen.

Pull Up Your Decision Documentation

Review what you decided, why, what you predicted, and how confident you were.

Compare to Reality

What actually happened? Where were you right? Where were you wrong?

Analyse Without Blame

The goal is learning, not assigning responsibility.

Questions:

• “Why were we overconfident about the timeline?”

• “What early warning signs did we miss?”

• “What evidence should have changed our view that we ignored?”

• “What process failures led to this outcome?”

• “What would we do differently next time?”

Distinguish Luck from Skill

Sometimes good decisions lead to bad outcomes (bad luck). Sometimes bad decisions lead to good outcomes (good luck).

The question isn’t just “did it work?” It’s “Was the decision process good, given what we knew at the time?”

If you made a risky bet with a low probability of success and got lucky, that’s not a good decision, it’s a lucky one. If you made a high-probability bet that failed due to unforeseeable events, that might be a good decision with bad luck.

Document Learnings

Create a simple decision learning log:

• What we learned about our forecasting accuracy

• What we learned about our process

• What we’ll do differently next time

• What we should stop doing/start doing

Over time, this log becomes organisational wisdom about how you make decisions.

PART 6: BALANCING INTUITION AND ANALYSIS

Now we need to address a critical tension: Everything we’ve covered emphasises deliberate, systematic thinking (System 2). But System 1 intuition is valuable. Experienced judgment matters. You can’t analyse everything to death.

How do you balance critical thinking with not overthinking? How do you know when to trust your gut versus when to force more analysis?

When Intuition Is Reliable

Intuition works when

1. You have deep experience in the domain

If you’ve been a CEO for 15 years, your intuition about people, strategy, and execution is valuable. You’ve seen thousands of situations and your brain has built accurate pattern recognition.

If you’ve managed sales teams for 10 years, your intuition about sales dynamics is probably good. You’ve seen what works and what doesn’t many times.

2. You have rapid, clear feedback loops

Intuition improves when you get fast feedback about whether you were right. If you make decisions and immediately see outcomes, your brain learns what works.

Chess grandmasters have excellent intuition because they play thousands of games and see immediately whether their moves were good. Sales reps develop good intuition about whether a deal will close because they see outcomes within weeks.

3. The situation is similar to past situations

Pattern recognition works when patterns repeat. If you’re facing a situation similar to ones you’ve handled before, your intuition is drawing on relevant experience.

If you’re building a sales team from 5 to 20 reps and you’ve done that before, your intuition is probably reliable. If you’re building a sales team from 5 to 200 and you’ve never done that, your intuition is less reliable.

4. The environment is stable and predictable

Intuition assumes the future will be like the past. If your industry, market, and business model are relatively stable, past patterns likely hold.

If you’re in a rapidly changing environment (new technology, new competitive dynamics, pandemic disruption), past patterns may not hold and intuition can mislead.

When Intuition Is Unreliable

Don’t trust intuition when:

1. You lack experience in the domain

If you’re new to a role, industry, or type of decision, you don’t have the pattern exposure needed for good intuition.

First-time CEOs should be especially skeptical of their intuition on CEO decisions. You haven’t built pattern libraries yet.

2. Feedback is slow, noisy, or absent

If you don’t get clear, timely feedback on whether you were right, your intuition can’t improve—and might get worse.

Strategic decisions often have delayed, ambiguous feedback. You make a strategic shift and 18 months later, things are better or worse, but you can’t definitively say whether your strategy caused it or other factors did.

In these domains, intuition doesn’t develop accuracy through experience.

3. The situation is novel or unprecedented

When you face situations you’ve never encountered before, you don’t have relevant patterns to draw on. Your brain will pattern-match to superficially similar situations, but the match may be wrong.

COVID-19 was unprecedented. Leaders who relied on intuition built on pre-pandemic patterns often made mistakes. The situation was unlike anything they’d seen before.

4. You’re emotionally activated

When you’re angry, afraid, or ego-involved, your intuition is strongly biased by emotion. What feels like intuition is often just an emotional reaction.

If you feel strong negative emotion toward a person, your intuition that they’re underperforming may be biased. If you feel excited about an opportunity, your intuition it will work may be biased by wishful thinking.

5. Base rates suggest you’re likely wrong

If most strategic pivots fail, most leadership hires fail, most product launches underperform expectations (all true), then even if your intuition says yours will be different, you should be sceptical.

Outside view: “Most X fail.” Inside view: “But mine will succeed because [reasons].” The outside view is usually more accurate than the inside view.

The Decision Framework: When to Slow Down

Here’s a practical framework for knowing when to override intuition with deliberate analysis:

Trust intuition (make a decision quickly) when

• ✅ Low stakes (easily reversible, limited downside)

• ✅ You have deep experience in this domain

• ✅ The situation is familiar (you’ve seen this before)

• ✅ Time pressure is real (delay has significant cost)

• ✅ You’re not emotionally activated

• ✅ The decision is within your zone of expertise

Force deliberate analysis (slow down) when

• ⚠️ High stakes (hard to reverse, significant resources)

• ⚠️ You lack experience in this domain

• ⚠️ The situation is novel (you haven’t seen this before)

• ⚠️ Your intuition is based on emotional reaction

• ⚠️ You feel especially certain (overconfidence check)

• ⚠️ Others who know the domain disagree with your intuition

• ⚠️ Base rates suggest intuition is likely wrong

Grey zone decisions need judgment

When some indicators say trust intuition and others say slow down, you need to use judgment about which factors matter most.

Generally, err on the side of slowing down for high-stakes decisions even if you feel confident. The cost of wrong high-stakes decisions is much higher than the cost of delay.

Avoiding Overthinking: Analysis Paralysis

The flip side of underthinking is overthinking, analysing endlessly, seeking perfect information, and delaying decisions indefinitely.

Signs you’re overthinking

• You’re waiting for certainty that will never come

• You’re analysing factors that don’t actually change the decision

• You’re studying the problem instead of testing hypotheses

• You’re avoiding the decision because of fear or anxiety

• You’ve gathered sufficient information but keep seeking more

• The cost of delay is exceeding the cost of being wrong

How to avoid analysis paralysis

1. Set decision deadlines

“We’ll make this decision by Friday.” Forces closure and prevents endless analysis.

2. Identify what information would actually change your decision

“If we knew X, would it change our decision? If not, don’t spend time trying to find X.”

3. Run experiments instead of analysing

“We could spend 3 more weeks analysing, or we could test this in 2 weeks and have real data.”

4. Recognise diminishing returns

The first 20% of the analysis usually gets you 80% of the insight. The last 20% of insight takes 80% of the analysis time. Stop at “sufficient” insight, not perfect insight.

5. Accept uncertainty

You’ll never have perfect information. Being 70% confident is often sufficient for decision-making. Waiting to be 95% confident often means waiting forever.

The CEO’s Intuition: Building Your Decision Muscle

Your intuition will get better over time if you:

1. Get feedback on your decisions

Track what you predicted and what happened. This is how intuition calibrates.

2. Study your failures more than your successes

When intuition was wrong, figure out why. What pattern were you matching that was inappropriate? What did you miss?

3. Seek situations where you can test your intuition with low stakes

Make small bets where you can see outcomes quickly and learn whether your intuition was right.

4. Learn from others’ experience

Read case studies, ask peer CEOs about decisions, study what worked and didn’t in similar situations. You can build some pattern recognition vicariously.

5. Stay humble about the limits

Even with experience, your intuition has bounds. Know when you’re outside your domain of expertise and be appropriately sceptical of intuition in those areas.

PART 7: DECISION SYSTEMS THAT WORK WITH HUMAN PSYCHOLOGY

Everything we’ve covered so far is about individual and team practices. But you also need systems and organisational infrastructure that make good decisions more likely and bad decisions less likely.

Decision Rights and Authority Matrices

The Problem

Ambiguity about who can make what decisions creates:

• Bottlenecks (everything escalates to you)

• Slow decisions (people aren’t sure if they can decide)

• Bad decisions (wrong people deciding things outside their expertise)

• Frustration (people thought they could decide but get overridden)

The Solution

Explicit decision rights frameworks like RAPID or RACI that clarify:

• Who recommends (does the analysis)

• Who has input (is consulted)

• Who decides (makes the call)

• Who must agree (has veto)

• Who gets informed (after the fact)

RAPID Framework (Bain)

For each type of decision:

• R – Recommend: Who makes the recommendation?

• A – Agree: Who must agree (veto power)?

• P – Perform: Who implements?

• I – Input: Who provides input?

• D – Decide: Who makes final decision?

Example: Pricing Changes

• Recommend: VP Product and VP Sales jointly

• Agree: CFO (must agree on financial model)

• Perform: Product and Sales teams

• Input: Customer Success (customer reaction), Marketing (positioning)

• Decide: CEO

Why This Helps

Clarifying decision rights:

• Speeds decisions (people know they can act)

• Reduces escalation (appropriate level makes the call)

• Ensures right expertise (people with knowledge decide)

• Creates accountability (clear who decided what)

Implementation

Map your major decision types:

• Pricing

• Product roadmap

• Marketing campaigns

• Hiring (at different levels)

• Budget allocation

• Strategic initiatives

For each, clarify RAPID roles. Document. Share widely. Update as you scale.

Decision-Making Meeting Types

The Problem:

Most meetings mix discussion, debate, and decision-making in unstructured ways. This leads to:

• Decisions made without sufficient input

• Endless discussion without decisions

• Confusion about whether a decision was actually made

The Solution

Separate meeting types with clear purposes:

1. Information Sharing (No Decisions)

Purpose: Share what’s happening, ask clarifying questions Duration: 30-45 minutes Output: Shared awareness

Rules:

• No decisions made

• Updates are brief (5 minutes max per topic)

• Questions are clarifying only, not debating

2. Problem Solving (Working Sessions)

Purpose: Solve specific tactical problems Duration: 60 minutes Output: Solutions to implement

Rules:

• Problems are concrete and scoped

• Focus is on generating solutions, not debating whether problem is real

• Decisions are implementation-level, not strategic

3. Strategic Discussion (No Decisions)

Purpose: Explore strategic topics, surface perspectives Duration: 60-90 minutes Output: Shared understanding, framed choices

Rules:

• Goal is NOT to decide

• Goal is to air views, consider options, identify uncertainties

• Ends with framing of what decision needs to be made

4. Strategic Decision (Decision Meeting)

Purpose: Make strategic decision Duration: 60-90 minutes Output: Decision, rationale, next steps

Rules:

• The decision is framed in advance (what are we deciding?).

• Pre-work is done (analysis, options)

• Meeting is structured (present options, seek disconfirming evidence, decide)

• The decision is documented

Why This Helps

Separating meeting types prevents:

• Making hasty decisions without sufficient exploration

• Endless exploration without ever deciding

• Confusion about whether a decision was made

People know what to expect and how to participate.

Pre-Mortems as Standard Practice

Instead of pre-mortems being ad hoc, make them standard practice for major commitments.

The Rule

“Before we commit to any strategic initiative over $X or any decision that’s hard to reverse, we run a pre-mortem.”

Make it part of your decision process, not something you optionally do when you remember.

Decision Logs and Review Cadence

Create a Decision Log

Simple shared document tracking major decisions:

• Date

• Decision

• Who decided

• Rationale

• Expected outcome

• Review date

Schedule Reviews

Quarterly, review decisions made 6-12 months ago:

• What happened vs. what we expected?

• Were we overconfident? Underconfident?

• What did we learn?

• What patterns are emerging?

This creates organisational memory and learning.

Slowing Down High-Stakes Decisions

Create Forcing Functions

“For any decision involving a budget over £100K or organisational change affecting 20+ people, we wait 48 hours between proposal and decision.”

This simple rule prevents hasty decisions without preventing any good decisions. If something is genuinely time-sensitive, 48 hours won’t matter. If it’s not, the delay enables System 2 thinking.

PART 8: PUTTING IT ALL TOGETHER

Let’s bring this back to where we started: You’re a Growth-Mandate CEO who’s completed diagnostic work and now needs to make decisions that restart growth.

The Meta-Skill: Knowing What You Don’t Know

The single most important takeaway from this entire article is this:

The quality of your decisions depends primarily on your ability to distinguish between:

• What you know (facts)

• What you believe (hypotheses with varying confidence)

• What you’re uncertain about (acknowledged unknowns)

• What you don’t even know you don’t know (blind spots)

Most leaders operate as if they know more than they do. They treat beliefs as facts. They’re overconfident about predictions. They don’t acknowledge blind spots.

The CEOs who successfully restart growth are the ones who:

1. Are honest about uncertainty – “I think X is the problem, I’m 65% confident, here’s why”

2. Actively seek disconfirming evidence – “What would prove me wrong? Let me look for that.”

3. Update beliefs as evidence changes – “I thought X, but now I think Y because of Z”

4. Build teams that challenge assumptions – “I need you to tell me where I’m wrong”

5. Create systems that catch bias – “We use pre-mortems, decision reviews, and structured processes”

This doesn’t mean being indecisive. It means being decisive while being honest about uncertainty.

The Personal Practice Stack

Here’s a simple version of what to actually do:

Daily

• Notice when you feel very certain—that’s an overconfidence flag

• When you feel strong emotion about a decision, pause

• Ask “Is this System 1 or System 2?” before important decisions

Weekly

• Review one major belief: What’s my confidence? What would change my mind?

• Seek out one piece of disconfirming evidence for something you believe

• Check in with at least one advisor who challenges your thinking

Monthly

• Decision journal review: What did I decide last month? Write it down with predictions.

• Team decision quality discussion: How are we making decisions? What can improve?

Quarterly

• Decision review: Look back at decisions from 6 months ago. Were we right?

• Confidence calibration check: Am I overconfident? Underconfident?

• Team bias check: What biases are showing up in our decisions?

The Team Practice Stack

For Major Strategic Decisions

1. Use structured decision process (frame, generate options, seek disconfirming evidence, decide, document)

2. Run pa re-mortem before committing

3. State confidence levels explicitly

4. Set the review date when deciding

Regular Practices

• Quarterly decision reviews (what did we decide 6 months ago? What happened?)

• Red team / blue team for major strategic choices

• Explicit “disagree and commit” culture

• Psychological safety for dissent

Systems

• Decision rights matrix (who decides what)

• Decision logs (tracking major decisions)

• Forcing functions (mandatory delays for high-stakes decisions)

• Experimentation wherever possible (test before committing)

The 90-Day Implementation Plan

Month 1: Foundation

• Introduce these concepts to your leadership team or get ZOKRI to

• Start decision journaling (personally)

• Implement a structured process for the next major strategic decision

• Create a decision rights matrix for key decision types

Month 2: Team Practices

• Run a pre-mortem first before a major commitment

• Implement disagree and commit explicitly

• Start team decision log

• Do the first quarterly decision review (of decisions from 6 months ago)

Month 3: Systems

• Create a decision review cadence

• Implement forcing functions for high-stakes decisions

• Assess and strengthen psychological safety

• Evaluate: How is decision quality improving?

Perfect Decisions Don’t Exist

You’re not going to eliminate bias. You’re not going to be perfectly rational. You’re not going to make the right decision every time.

That’s not the goal.

The goal is incremental improvement. To make slightly better decisions on average. To catch some of your biases before they become strategy. To update your beliefs when evidence changes instead of defending them. To build a team that challenges assumptions instead of reinforcing them.

The CEOs who restart growth aren’t the ones who never make mistakes. They’re the ones who:

• Make mistakes faster (because they test assumptions quickly)

• Learn from mistakes better (because they review decisions honestly)

• Don’t repeat mistakes (because they build systems that catch recurring errors)

• Stay humble about uncertainty (because they know how hard prediction is)

That’s the work. Understanding how you really make decisions. Building practices that make you slightly better. Creating teams that decide well collectively.

It’s not dramatic. It’s not a silver bullet. But it’s the foundation for everything else.

Because if you can’t make good decisions under uncertainty, nothing else matters. The best strategy poorly decided will fail. The best team, poorly organised, will underperform. The best opportunity poorly executed will be wasted.

Decision quality is the meta-capability that determines everything else.

So start with understanding how you really decide. Then build practices that make you incrementally better. That’s the path.