Implicit Bias and Decision Making
Chapter 5: Implicit Bias and Decision-Making - When the Way We Think Influences the Decisions We Make
"The hardest biases to recognize are often the ones we don't know we have."
Learning Objectives
By the end of this chapter, you should be able to:
-
understand what implicit bias is
-
distinguish implicit bias from intentional discrimination
-
recognise how assumptions influence business decisions
-
identify common forms of bias in case analysis
-
understand how bias can affect customer, employee, and stakeholder analysis
-
recognise how bias can influence team decision-making
-
identify techniques for reducing the influence of bias
-
challenge assumptions without unnecessarily slowing decision-making
-
incorporate more objective evidence into recommendations
-
recognise how diversity of perspectives can improve strategic decisions
Why This Matters
Case competitions are supposed to be analytical exercises.
Teams are given information.
They analyzeanalyse the situation.
- They identify opportunities and problems.
- They evaluate alternatives.
- They make a recommendation.
It sounds objective.
But there is a problem.
The analysis is performed by people.
And people bring assumptions with them.
We make assumptions about:
-
customers
-
employees
-
competitors
-
markets
-
industries
-
demographics
-
motivations
-
behavior
-
risk
-
value
-
leadership
-
success
Many of these assumptions happen automatically.
We may not even realizerealise that we are making them.
That is where implicit bias becomes important.
Discover Your Mad Skills Principle
Good analysis does not eliminate judgment. It makes judgment more visible, testable, and defensible.
The objective isn't to become completely free of bias.
That is unrealistic.
The objective is to recognizerecognise where assumptions may be influencing the analysis and then deliberately test them.
What Is Implicit Bias?
Implicit bias refers to automatic associations or assumptions that can influence our perceptions, judgments, and decisions without conscious intent.
These associations can develop from:
-
experience
-
culture
-
education
-
media
-
social environments
-
professional experience
-
repeated exposure to certain ideas
Importantly, having an implicit bias does not necessarily mean someone consciously believes something discriminatory.
The problem is that automatic assumptions can still influence decisions.
Explicit vs. Implicit Bias
A useful distinction is:
- Explicit
Bias
Bias. A conscious belief or preference.
The person is aware of the belief.
An automatic association that may influence behaviorbehaviour without conscious awareness.
For example, someone may consciously believe that all customers should be treated equally while unconsciously making different assumptions about customers based on age, occupation, appearance, geography, or other characteristics.
This is why good intentions alone aren't enough.
Bias in Business Decisions
Bias can appear throughout the strategic decision-making process.
Consider the sequence:
Problem Definition -->
↓
Data Collection -->
↓
Analysis -->
↓
Alternatives -->
↓
Recommendation -->
↓
Implementation
Bias can enter at every stage.
If the problem is framed incorrectly, the analysis may already be biased before the spreadsheet is opened.
Bias Begins With Problem Definition
Suppose a company says:
"Young customers don't want our product."
That statement contains an assumption.
A better question might be:
"Why is adoption lower among younger customers?"
The first statement assumes the problem.
The second investigates it.
This distinction matters.
Ask Better Questions
- Instead of:
Ask:"Why don't these customers like us?"
Ask:"What evidence explains the difference in customer
behavior?behaviour?" - Instead of:
Ask:"Why are older employees resistant to technology?"
Ask:"What factors are influencing technology adoption among employees?"
- Instead of:
Ask:"Why aren't women applying for these positions?"
Ask:"What factors are influencing the applicant pool?"
Better questions reduce the likelihood that assumptions become conclusions.
Bias in Customer Analysis
Customer analysis is particularly vulnerable to assumptions.
Teams may create a persona such as:
"Sarah is a 35-year-old middle-class professional living in an urban area."
But this tells us relatively little about why Sarah would buy something.
Demographic information can be useful.
But it does not automatically explain:
-
pain points
-
motivations
-
behaviors
-
goals
-
barriers
-
willingness to pay
-
switching behavior
A stronger persona is based on evidence about customer behavior.behaviour.
From Demographics to Motivations
Instead of:
Women aged 25–40
ask:
-
What problem are they trying to solve?
-
What frustrates them?
-
What motivates them?
-
What alternatives do they currently use?
-
What prevents them from switching?
-
What would cause them to pay?
This shifts the analysis from assumptions about people to evidence about behavior.behaviour.
Confirmation Bias
One of the most important biases in case analysis is confirmation bias.
Confirmation bias occurs when people tend to notice, interpret, or prioritizeprioritise information that supports what they already believe.
For example:
The team believes:
"The company should expand internationally."
They begin looking for evidence supporting international expansion.
They may pay less attention to:
-
regulatory barriers
-
competitor strength
-
weak demand
-
capital requirements
-
operational complexity
The analysis starts becoming a search for confirmation rather than an evaluation of alternatives.
The Confirmation Bias Test
Before finalizingfinalising a recommendation, ask:
"What evidence would prove us wrong?"
This is one of the most powerful questions a case team can ask.
If the team cannot answer it, they may not have tested their recommendation sufficiently.
Anchoring Bias
Anchoring occurs when an initial number, idea, or assumption disproportionately influences later judgments.
For example:
The case says:
"The market is expected to grow by 10%."
The team may automatically build its forecast around 10%.
But perhaps the company operates in a segment growing at only 4%.
The original number has become an anchor.
Anchoring in Financial Analysis
Anchoring can influence:
-
market size
-
growth rates
-
pricing
-
margins
-
costs
-
valuation multiples
-
discount rates
-
market share
-
revenue forecasts
The solution is not to ignore the initial information.
It is to ask:
"What evidence supports this assumption?"
Availability Bias
Availability bias occurs when people give greater weight to information that is easy to recall.
For example:
A team recently heard about a successful subscription business.
They may conclude:
"Subscription models are the future."
But one memorable example does not establish a market trend.
Ask:
Is this evidence representative?
Recency Bias
Recent information can also receive disproportionate attention.
Suppose the company had one bad quarter.
The team may conclude:
"The business is declining."
But a five-year trend might show that the quarter was an unusual event.
Always distinguish:
signal
fromfrom
noise.
Stereotyping
Stereotyping occurs when assumptions about a group are applied to individuals.
For example:
"Older customers don't use technology."
That may be an assumption rather than evidence.
A better analysis asks:
Which customer segments are actually adopting the technology, and why?
Segment based on relevant behaviorbehaviour rather than assumptions.
Status Quo Bias
People often prefer existing approaches simply because they are familiar.
In a casecase, this might appear as:
"The company has always done it this way."
But the status quo is itself a strategic choice.
Ask:
What happens if we do nothing?
Doing nothing has:
-
financial consequences
-
competitive consequences
-
operational consequences
-
strategic consequences
Loss Aversion
People often perceive losses as more significant than equivalent gains.
For example:
A team may reject an investment because:
"We could lose $2 million."
Even if the expected value of the investment is strongly positive.
This doesn't mean risk should be ignored.
It means risk should be evaluated systematically.
Ask:
-
What is the probability?
-
What is the magnitude?
-
Can the downside be mitigated?
-
What is the expected value?
-
What is the opportunity cost of not investing?
Overconfidence Bias
Teams can become overly confident in their assumptions.
This is especially dangerous in case competitions because teams have limited time.
They may say:
"We're sure customers will adopt this."
But what evidence supports that confidence?
Strong teams distinguish between:
- What we know
- What we estimate
- What we assume
- What we need to test
That distinction strengthens credibility.
Groupthink
Bias isn't only an individual problem.
It can become a team problem.
Groupthink occurs when the desire for agreement discourages people from challenging the emerging consensus.
For example:
- One team member says:
"I don't think this recommendation works."
- The rest of the team responds:
"We're already committed to it."
The concern disappears.
That may feel efficient.
It can also produce a weak recommendation.
The Devil's Advocate
One useful technique is to deliberately assign someone the role of challenger.
Their job is not to destroy the recommendation.
Their job is to test it.
Ask them to identify:
-
the weakest assumption
-
the biggest risk
-
the strongest competitor response
-
the most vulnerable financial assumption
-
the most important missing evidence
-
the strongest argument against the recommendation
This creates productive disagreement.
The Red Team Exercise
A stronger version is the red team.
One person or subgroup is asked to attack the recommendation.
They should assume:
"We are trying to prove this recommendation is wrong."
They look for:
-
flawed assumptions
-
missing data
-
alternative explanations
-
implementation problems
-
unintended consequences
-
competitor reactions
The original team then has an opportunity to strengthen the recommendation.
Diversity of Perspectives
Different perspectives can improve decision-making.
This isn't simply about demographic diversity.
It also includes diversity of:
-
experience
-
education
-
expertise
-
industry knowledge
-
cultural perspective
-
problem-solving approach
-
risk tolerance
Different perspectives can expose assumptions that a homogeneous team might overlook.
But Diversity Alone Isn't Enough
Simply having different perspectives doesn't guarantee better decisions.
People must feel able to express those perspectives.
This is where psychological safety matters.
Team members need to be able to say:
"I disagree."
without believing that disagreement will damage their position within the team.
Challenge Ideas, Not People
A useful team rule is:
Attack the assumption, not the person.
- Instead of:
"That's a terrible idea."
- Say:
Say:"What evidence supports that assumption?"
- Instead of:
"You're wrong."
- Say:
Say:"What would need to be true for that recommendation to work?"
This makes disagreement productive.
Bias in Stakeholder Analysis
Bias can also affect how teams rank stakeholders.
For example, a team may prioritize:prioritise:
But why?
Is that ranking supported by the case?
Or is it simply an assumption about whose interests matter most?
A stronger approach evaluates stakeholders based on:
-
power
-
interest
-
impact
-
legitimacy
-
urgency
Bias in Competitive Analysis
Teams can also underestimate competitors.
A common mistake is to assume:
"Our solution is better."
But competitors may respond.
Ask:
-
What will competitors do?
-
What capabilities do they have?
-
What advantages do they possess?
-
How easily can they copy the idea?
-
How will customers compare alternatives?
A recommendation should survive competitive response.
Bias in Financial Forecasting
Forecasts are especially vulnerable to bias.
Suppose the team wants the project to work.
They may unconsciously choose:
-
optimistic growth
-
optimistic conversion
-
optimistic margins
-
low costs
-
low discount rates
Individually, each assumption may seem reasonable.
Together, they may create an unrealistic forecast.
Assumption Stacking
This is an important concept.
Imagine a forecast depends on:
-
10% market growth
-
8% market share
-
15% conversion
-
20% margin improvement
Each assumption might be plausible.
But if every assumption is optimistic, the combined result may be highly unlikely.
This is why sensitivity and scenario analysis are so important.
Bias and Sensitivity Analysis
Sensitivity analysis provides a practical defensedefence against biased assumptions.
Instead of asking:
Ask:"What number do we think is right?"
Ask:
"What happens if the number is higher or lower?"
For example:
| Assumption | Downside | Base | Upside |
|---|---|---|---|
| Growth | 2% | 5% | 8% |
| Market share | 3% | 5% | 7% |
| Margin | 10% | 15% | 20% |
The goal is not to eliminate uncertainty.
It is to understand it.
Evidence Hierarchy
Not all evidence should receive equal weight.
A useful hierarchy is:
Strong Evidence
-
case facts
-
reliable historical data
-
credible industry data
-
customer research
-
verified benchmarks
Moderate Evidence
-
comparable organizations
-
expert estimates
-
reasonable assumptions
Weak Evidence
-
anecdotes
-
personal experience
-
intuition
-
memorable examples
This doesn't mean weak evidence is useless.
It means we should know what kind of evidence we are using.
The Assumption Audit
Before finalizingfinalising your recommendation, create three columns:
| Assumption | Evidence | Confidence |
|---|---|---|
| Market grows 6% | Industry benchmark | Medium |
| Conversion is 10% | Comparable program | Medium |
| Customers will switch | Survey evidence | High |
Then ask:
Which assumptions are doing the most work?
Those are the assumptions that deserve the most attention.
The Pre-Mortem
A powerful technique for reducing bias is the pre-mortem.
Imagine:
It is two years
laterlater, and the recommendation failed.
Now ask:
Why did it fail?
Possible answers might include:
-
customers didn't adopt
-
costs were underestimated
-
competitors responded
-
implementation was too slow
-
employees resisted
-
regulation changed
-
the market was smaller than expected
This exercise forces the team to imagine alternatives to its preferred narrative.
Coach's Lens
When coaching teams, one of the most valuable questions is:
"What are you assuming?"
Often the team responds:
"We're not assuming anything. It's in the case."
But case facts and interpretations are different.
For example:
- Fact:
Sales declined 8%.
- Interpretation:
Customers are leaving because of price.
The first is evidence.
The second is a hypothesis.
Treat them differently.
From Assumption to Hypothesis
A useful discipline is to turn assumptions into hypotheses.
- Instead of:
"Customers don't like the current product."
- Say:
Say:"We
hypothesizehypothesise that product complexity is reducing customer adoption." - Then ask:
- What evidence would support this?
- What evidence would contradict it?
Now the team can test the idea.
Bias and the Case Presentation
Bias can also affect how teams present their analysis.
Once a team becomes attached to its recommendation, it may:
-
hide weaknesses
-
minimize risks
-
ignore contradictory evidence
-
emphasize favorable numbers
-
avoid difficult questions
Judges often recognizerecognise this.
A more credible approach is to acknowledge uncertainty.
For example:
"Our recommendation is most sensitive to customer adoption. At 5%
adoptionadoption, the initiative generates positive NPV; below approximately 3%, the economics become unattractive. Our implementation therefore focuses on validating adoption before scaling."
That sounds much more credible.
Handling Bias in Q&A
A judge may challenge your assumption:
"Why do you believe customers will switch?"
Don't defend the assumption automatically.
Instead:
-
acknowledge the uncertainty
-
explain the evidence
-
identify the risk
-
explain the mitigation
For example:
"That's the key uncertainty in our model. We based the assumption on comparable adoption rates and the customer pain point identified in the case. We would validate it through a pilot before committing the full investment."
That demonstrates intellectual honesty.
Common Mistakes
Mistake
1 — - Assuming
Objectivity
Objectivity. Everyone has assumptions.
Mistake 2 —
A conclusion isn't automatically a fact.
Mistake 3 —
Test what could prove you wrong.
Mistake 4 —
Strong teams create space for disagreement.
Mistake 5 —
Intuition can generate hypotheses but shouldn't automatically validate them.
Mistake 6 —
Understand motivations and behaviors.
Mistake 7 —
Use scenarios and sensitivity analysis.
Mistake 8 —
Disconfirming evidence may be the most valuable evidence in the case.
Mad Skills Drill
Prove Yourself Wrong
Take your recommendation.
Complete these sentences:
- Our recommendation fails if...
- The most dangerous assumption is...
- The evidence that would change our mind is...
- The stakeholder most likely to disagree is...
- The competitor response we are most concerned about is...
This exercise forces the team outside its preferred narrative.
Mad Skills Drill
Fact, Assumption, or Hypothesis?
Take five statements from your case.
Label each:
- F — Fact
- A — Assumption
- H — Hypothesis
Then ask:
What evidence would move each assumption or hypothesis closer to fact?
This is a simple exercise that can dramatically improve analytical discipline.
Mad Skills Drill
The Six-Question Bias Check
Before locking in your recommendation, ask:
-
What are we assuming?
-
What evidence supports it?
-
What evidence contradicts it?
-
What would prove us wrong?
-
Who sees this differently?
-
What happens if we're wrong?
If the team cannot answer these questions, the recommendation may not be ready.
Building a Less Biased Recommendation
A strong recommendation should include:
Evidence
- Evidence: What do we know?
- Assumptions:
What are we estimating?
- Uncertainty:
What don't we know?
- Alternatives:
What else could we do?
- Risks:
What could go wrong?
- Mitigation:
How will we respond?
Assumptions
Uncertainty
Alternatives
Risks
Mitigation
This creates a recommendation that is transparent rather than artificially certain.
Chapter Summary
Implicit bias does not mean that people are intentionally making poor decisions.
It means that automatic assumptions can influence how we:
-
interpret information
-
define problems
-
evaluate customers
-
assess stakeholders
-
forecast outcomes
-
compare alternatives
-
evaluate risk
-
make recommendations
The solution isn't to eliminate judgment.
It is to make judgment more deliberate.
Strong case solvers continuously ask:
- What are we assuming?
What evidence supports it?
What could prove us wrong?
Who sees this differently?
Key Takeaways
✓ Everyone brings assumptions into decision-making.
✓ Implicit bias can influence decisions without conscious intent.
✓ Facts, assumptions, and hypotheses should be distinguished.
✓ Confirmation bias can cause teams to search for evidence supporting their preferred answer.
✓ Anchoring can distort forecasts and valuations.
✓ Groupthink can prevent teams from challenging weak recommendations.
✓ Strong teams create space for disagreement.
✓ Sensitivity analysis helps expose the impact of assumptions.
✓ Pre-mortems can reveal risks that teams otherwise overlook.
✓ Better questions often produce better analysis.
✓ Diverse perspectives can expose blind spots.
✓ A strong recommendation should be transparent about uncertainty.
You don't need to eliminate every bias to make a better decision. You need to build a process that makes it harder for bias to control the decision.
Looking Ahead
RecognizingRecognising bias is only the beginning.
The next challenge is deciding what to do when the future itself is uncertain.
Even when our analysis is objective, our assumptions are reasonable, and our team has challenged its own thinking, we still cannot know exactly what will happen.
- Markets change.
- Competitors respond.
- Customers behave differently than expected.
- Regulations change.
- Costs move.
- Technology evolves.
The best decision-makers therefore don't ask:
"Can we predict the future?"
They ask:
"How can we make a strong decision despite uncertainty?"
That is the next challenge in strategic decision-making.