SEEING THE UNSEEN Data & Decision Science for Leaders Beyond Data. Beyond Dashboards. Beyond Assumptions.

"Great managers read reports. Great leaders read between the reports."

Why Smart Leaders Don't Make Decisions Based on Data Alone

"The biggest business failures rarely happen because of a lack of data. They happen because of the wrong interpretation of the right data."

Imagine you're the Plant Head of an automobile manufacturing company.

Your dashboard tells you:

Production is down.
Overtime is up.
Customer complaints are increasing.
Machine OEE is above 90%.
Inventory is growing.

What would you conclude?

Most managers will immediately recommend:

Increase manpower.
Improve maintenance.
Tighten quality inspections.
Increase production targets.

But what if every one of those decisions is wrong?

What if the real problem isn't visible on the dashboard at all?

Welcome to the world of Decision Science.

We Live in the Age of Data...

Every minute, organizations collect millions of data points.

Factories measure:

Production
OEE
Downtime
Rejections
Energy
Inventory
Delivery
Safety

HR tracks:

Attendance
Attrition
Engagement
Productivity

Sales teams monitor:

Revenue
Conversion
Customer satisfaction
Market share

Yet despite having more data than ever before, organizations continue to make poor decisions.

Why?

Because...

Data doesn't make decisions. People do.

And people are influenced by assumptions, biases, incomplete information, and past experiences.

The Greatest Insights Are Hidden in Plain Sight

During World War II, engineers studied returning aircraft covered with bullet holes.

The obvious decision?

Strengthen the areas with the most bullet holes.

One statistician asked a different question:

"Where are the aircraft that didn't return?"

That single question changed military strategy.

The planes were reinforced where there were no bullet holes, because hits in those areas prevented the aircraft from returning.

The data never changed.

The interpretation did.

The Same Thing Happens Every Day in Business

Imagine these situations:

Customer complaints decrease.
Employee engagement reaches 95%.
Machine utilization exceeds 98%.
Attrition falls.
Production targets are achieved.

Good news?

Maybe.

Or maybe...

The complaint portal stopped working.
Only managers responded to the survey.
Downstream machines are waiting.
Employees stopped reporting issues.
Inventory has quietly doubled.

The dashboard is telling the truth.

But not the whole truth.

The Difference Between Managers and Leaders

Managers ask:

"What happened?"

Leaders ask:

"What are we not seeing?"

Managers read reports.

Leaders question reports.

Managers solve visible problems.

Leaders discover invisible causes.

Think Like a Detective, Not a Reporter

This program isn't about Excel.

It isn't about Power BI.

It isn't about Artificial Intelligence.

It is about developing a new way of thinking.

Through fascinating real-world stories, business cases, and interactive simulations, participants will discover how organizations like:

Toyota
Google
UPS
Target
Rolls-Royce
Walmart
Amazon

transformed ordinary business data into extraordinary competitive advantage—simply by asking better questions.

We'll also explore powerful lessons from movies like:

Moneyball
Vikram Vedha
Maharaja
12 Angry Men
Drishyam

to understand how assumptions, interpretation, and hidden variables shape every decision we make.

What You'll Experience

This isn't a traditional classroom session.

You'll step into the role of:

A Plant Head facing declining productivity.
A Quality Manager investigating rising warranty claims.
A CEO interpreting conflicting dashboards.
A Detective solving business mysteries hidden inside operational data.

Every exercise will challenge your first assumption.

Every discussion will reveal a hidden perspective.

Every case study will change the way you think about data.

The Question That Could Change Your Leadership Forever

Before making your next business decision, ask yourself:

"Am I looking at the data... or am I looking at my interpretation of the data?"

Because sometimes...

The biggest opportunity...

The biggest risk...

The biggest innovation...

And the biggest competitive advantage...

are all hidden in the information everyone else has already seen.

The leaders who succeed are not those with the most data.

They are the ones who see what others don't.

WW-II

Activity - 1 - Which areas to be strengthened in this aircraft?

The Survivor Bias Airplane Diagram

Seeing the Unseen: The Leadership Lesson That Changed the Course of World War II

"The most dangerous decisions are often based on incomplete data—not incorrect data."

Introduction

Imagine that you are the Head of Manufacturing of a world-class automobile company.

Every morning you receive a dashboard containing:

Production Output
OEE
Machine Downtime
Customer Complaints
Employee Productivity
Rejection Rate
Inventory

Everything appears to be under control.

Then suddenly the CEO asks a simple question:

"What are we NOT measuring?"

Silence.

That single question represents the difference between reporting data and understanding data.

More than 80 years ago, a similar question changed military strategy during World War II and gave birth to one of the greatest lessons in statistics, analytics, leadership, and decision science.

It is known today as The Survivor Bias Airplane Case.

The World War II Problem

During World War II, Allied bombers were constantly being attacked.

Many aircraft never returned.

Those that returned had visible bullet holes all over their bodies.

Military engineers collected data from every aircraft that landed safely.

Thousands of observations were recorded.

The objective was simple:

"Where should we add extra armor?"

The Data Collection

Every returning aircraft was inspected.

Engineers mapped every bullet hole.

After hundreds of missions, the aircraft looked like this.

(Show the Survivor Bias Aircraft Diagram here.)

Participants immediately notice:

Wings have many bullet holes
Tail has many bullet holes
Body has many bullet holes

The engines and cockpit have very few.

Interactive Activity (Before Revealing the Answer)

Ask your participants:

If you were the Chief Engineer, where would you strengthen the aircraft?

Allow them to discuss for two minutes.

Collect responses.

Almost everyone answers:

Wings
Tail
Body

Because...

That's where most bullet holes appear.

The First-Level Interpretation

The military initially reached the same conclusion.

Their reasoning was perfectly logical.

More bullet holes

↓

More damage

↓

Add armor there.

Simple.

Logical.

Wrong.

Enter Abraham Wald

A Hungarian mathematician named Abraham Wald was asked to review the analysis.

Instead of looking at the bullet holes...

He looked at something else.

He asked a question nobody else asked.

"Where are the airplanes that didn't come back?"

Everything changed.

The Missing Data

The military had only collected data from

Surviving aircraft.

Nobody had data from

Aircraft that crashed.

The sample itself was biased.

This is called

Survivorship Bias.

The Hidden Insight

Wald realized:

The wings contained many bullet holes.

Yet...

The aircraft still returned.

Therefore,

Those areas could tolerate damage.

Now look at the engines.

Almost no bullet holes.

Why?

Not because they weren't hit.

Because

Aircraft hit in the engines

never came back.

They were never measured.

The Correct Interpretation

Instead of reinforcing

the areas with

the most bullet holes

the military reinforced

the areas with

the fewest bullet holes.

Because those were the fatal impact zones.

The data hadn't changed.

Only the interpretation had.

The Decision Science Framework

This case perfectly demonstrates how leaders should think.

Stage 1

Observation

Aircraft have bullet holes.

↓

Stage 2

Data Collection

Map all bullet holes.

↓

Stage 3

First Assumption

Armor the damaged areas.

↓

Stage 4

Question the Assumption

What data is missing?

↓

Stage 5

New Interpretation

Only survivors were measured.

↓

Stage 6

Decision

Protect areas where damage prevents survival.

Why This Matters in Business

Most organizations unknowingly make the same mistake.

They analyse only

Successful projects
Successful employees
Successful customers
Successful suppliers
Successful plants

Rarely do they analyse

Lost customers
Failed projects
Employees who resigned
Machines that completely failed
Opportunities never pursued

Leadership begins by asking:

"Whose data is missing?"

Automobile Manufacturing Example

Suppose your dashboard shows:

Department	Complaints
Assembly	420
Paint	130
Welding	70

Question

Which department should receive attention?

Most managers answer

Assembly.

Now reveal

Assembly produces

75% of all vehicles.

The complaint rate per 1,000 vehicles is actually lower than the others.

The interpretation changes.

HR Example

HR reports:

Top-performing employees attended leadership training.

Management concludes:

Training creates high performers.

Question

Who attended the training?

Answer

Only already high-performing employees.

The conclusion is invalid.

The missing data?

Employees who never attended.

Sales Example

A company studies only its top 100 customers.

It discovers

Fast delivery
Premium service
Dedicated account managers

Management copies these practices.

Sales don't improve.

Why?

Because no one analysed

the 5,000 customers who left.

Customer Complaint Example

Customer complaints reduced

from

600

Everyone celebrates.

Later they discover

the online complaint portal stopped working.

The data was accurate.

The interpretation wasn't.

Manufacturing KPI Example

Machine A

OEE

96%

Machine B

82%

Everyone wants to improve

Machine B.

Then someone asks

"What machine creates waiting time downstream?"

Machine A causes the largest bottleneck because it produces in oversized batches.

The best-performing machine became the biggest constraint.

The Leadership Mistake

Managers often ask:

Why did production reduce?
Why did quality drop?
Why did employees resign?
Why did sales decline?

Great leaders ask:

What information are we missing?
What assumptions are we making?
Who is not represented in this data?
What would completely change our interpretation?

Modern Examples of Survivorship Bias

Startups

People study successful startups.

Few study

the thousands that failed.

Recruitment

Companies analyse

high performers.

Rarely analyse

why average performers stay average.

Safety

Zero accidents.

Excellent?

Maybe.

Or perhaps

employees stopped reporting incidents.

Employee Engagement

Engagement Score

95%

Outstanding?

Maybe.

Participation

18%.

Only managers completed the survey.

Social Media

People see

successful entrepreneurs.

Not the millions who tried and failed.

The Decision Science Principle

Data

what happened.

Information

explains

what happened.

Insight

reveals

why it happened.

Wisdom

helps decide

what should happen next.

Reflection Activity

Ask participants to think about their own organization.

Complete the following table.

Question	Reflection
What reports do we rely on every week?
What information do these reports not contain?
Which stakeholders are not represented?
What assumptions do we automatically make?
If the opposite were true, what data would prove it?

Group Exercise

Each table receives a dashboard.

The dashboard contains

OEE
Downtime
Production
Inventory
Complaints
Attrition

Task

List

Five conclusions.

Then

List

Five questions you would ask before making a decision.

The second list is more important than the first.

Key Takeaways

By the end of the discussion, participants should understand that:

More data does not guarantee better decisions.
Missing data can be more important than available data.
High-performing areas may hide systemic problems.
Low numbers are not always good news.
Dashboards describe reality—but never the whole reality.
The quality of leadership depends on the quality of the questions asked.

The Four Questions Every Leader Should Ask

Before making any important decision, pause and ask:

1. What does the data tell me?

2. What does the data NOT tell me?

3. What assumptions am I making?

4. What evidence would completely change my conclusion?

Closing Message

"The greatest discoveries in business are rarely hidden by a lack of data. They are hidden by the assumptions we make about the data we already have."

The Survivor Bias Airplane story is not just a lesson in statistics.

It is a lesson in leadership.

Great managers read reports.

Great leaders search for the reports that don't exist.

And that is the essence of Seeing the Unseen – Data & Decision Science for Management.

the famous "Survivorship Bias" illustration from World War II and is one of the best examples to introduce Data & Decision Science to managers.

What the image shows

Red dots = Bullet holes found on aircraft that returned safely from missions.
Blue dashed boxes = Areas that statistician Abraham Wald recommended reinforcing with armor.

At first glance, most people conclude:

"Armor should be added where there are the most bullet holes."

That conclusion is wrong.

The correct insight

The aircraft shown are only the ones that survived.

The planes that did not return were likely hit in the areas where there are few or no bullet holes on this diagram—such as the engines and cockpit.

Those hits were fatal, so those aircraft never made it back to be counted.

Therefore, the correct decision was:

Reinforce the engines
Reinforce the cockpit
Reinforce other critical areas with few bullet holes

Decision Science Principle

Good decisions come from analyzing the data that is missing—not just the data that is visible.

This is called Survivorship Bias.

How to use this in your training

Step 1 – Show only the image (2 minutes)

Ask:

"If you were the Engineering Head, where would you strengthen the aircraft?"

Most participants will answer:

Wings
Tail
Areas with the most bullet holes

Write their answers on a flipchart.

Step 2 – Reveal the twist

Ask:

"What data is missing?"

Pause.

Then explain:

"These are only the aircraft that came back."

Now ask:

"What happened to the aircraft that never returned?"

Participants immediately realize the flaw.

This creates a powerful "Aha!" moment.

Link it to an Automobile Manufacturing Plant

Scenario 1 – Quality

Management observes:

Most defects occur in Assembly Line 3.

Decision:

Increase inspectors in Line 3.

But after deeper analysis:

Line 1 produces the highest-volume model.
Line 3 produces a complex premium model with more inspections.

The real issue:

Inspection frequency, not production quality.

Scenario 2 – Warranty

Data shows:

Model	Warranty Claims
A	1,500
B	300

Management decides:

Model A has poor quality.

Missing data:

Model	Vehicles Sold
A	300,000
B	8,000

The better metric claims per 1,000 vehicles, not total claims.

Scenario 3 – Production

Management sees:

Night shift has more defects.

Immediate conclusion:

Night shift operators need retraining.

Missing data:

Night shift manufactures a new product variant.
New operators were assigned only to that shift.
Material from a new supplier arrived during the night.

The root cause may have nothing to do with operator skill.

Scenario 4 – Safety

Most accidents occur in Assembly.

Conclusion:

Assembly is unsafe.

Missing data:

Assembly has 65% of the workforce.
It also has the highest exposure hours.

A better measure is:

Accidents per 100,000 work hours.

Scenario 5 – HR Analytics

Top-performing employees attend training.

Management concludes:

Training creates top performers.

Missing data:

Only high performers were nominated for training in the first place.

This is another example of survivorship or selection bias.

Trainer's Key Message

Conclude with this statement:

"Managers are not paid to make quick decisions. They are paid to make the right decisions. The right decision comes from asking: What data am I seeing? What data am I missing?"

Debrief Questions

What assumption did you make when you first saw the image?
What information was missing?
Have you made similar decisions at work?
Which reports in your department show only successful outcomes?
What additional data would improve your decisions?

This exercise is an excellent opening activity because it immediately shifts participants from intuition-based thinking to evidence-based decision making, setting the tone for the entire Data & Decision Science program.

Decision Science Model Hidden Inside Vikram Vedha

Vikram Vedha (2022 film) - Wikipedia

Every story follows the same analytical cycle:

Observation

↓

Assumption

↓

Evidence

↓

Contradiction

↓

Re-analysis

↓

New Interpretation

↓

Decision

This is exactly how a good data scientist works.

Story 1 – The Encounter

Initial Data

Vikram is an honest police officer.
Vedha is a dreaded gangster.
Police eliminate gang members.
Public celebrates.

Ask Participants

Who is the hero?

95% will answer

Vikram.

First Assumption

Good Police

↓

Bad Criminal

Simple.

New Data

Vedha surrenders voluntarily.

Question:

"Why would the most wanted criminal surrender?"

Participants become confused.

Hidden Interpretation

Vedha is controlling the investigation.

Not escaping it.

Decision Science Principle

When behaviour contradicts expectations, revisit your assumptions.

Story 2 – Pulli

Vedha narrates Pulli's story.

Initial Data

Pulli was caught carrying drugs.

Question

Is Pulli a criminal?

Everyone says

Yes.

More Data

Pulli was forced.

More Data

He confessed immediately.

More Data

He wanted to leave crime.

Interpretation Changes

Pulli

↓

Victim

instead of

↓

Criminal.

Leadership Lesson

Never classify people using only one event.

Story 3 – Simon

Initially

Simon

Honest police officer.

Trusted friend.

Later

Vedha asks

"What if Simon accepted money?"

Everyone rejects the idea.

New Evidence

Medical expenses.

Financial pressure.

Corruption.

Participants suddenly realise

Good people

sometimes

make bad decisions.

Leadership Lesson

Context changes judgement.

Story 4 – The Encounter Team

Initially

Entire police team

Heroes.

Evidence

Repeated encounters.

Medals.

Recognition.

Hidden Data

Only Vedha's gang

was being eliminated.

Question

Who benefits?

Participants begin thinking differently.

Final Revelation

Entire unit

was compromised.

Leadership Lesson

Patterns reveal systems.

Not individuals.

The Biggest Decision Science Lesson

At the beginning of the movie

Participants think

Police = Good

Gangster = Bad

At the end

Good People

↓

Can make wrong decisions

↓

Bad People

↓

Can have ethical reasons

↓

Reality

↓

Lives in the grey area.

Exactly what Decision Science teaches.

Wonderful Facilitation Exercise

Pause after every Vedha story.

Don't play the next scene.

Ask participants to complete this table.

Stage	Participant Answer
What happened?
What do you think happened?
What evidence supports your opinion?
What evidence is missing?
What assumptions are you making?
If one new fact appears, will your decision change?

Only then

Continue the movie.

They experience

Bayesian thinking

without knowing the term.

Corporate Parallel

Production Delay

Initial Data

Machine stopped.

Participants

Maintenance problem.

New Data

Material delayed.

Interpretation changes.

More Data

Planning changed schedule.

More Data

Customer revised order.

More Data

Supplier delivered mixed batches.

Finally

Machine wasn't the problem.

It was the last visible symptom.

Exactly like

Vikram Vedha.

The Moral Ambiguity Matrix

Draw this during training.

What We Saw	What Actually Happened
Criminal surrendered	He wanted Vikram to investigate deeper
Police encounter	It was manipulated by corruption
Pulli was a criminal	Pulli was exploited
Simon was loyal	Simon acted under personal pressure
Vikram knew the truth	Vikram knew only part of the truth

Then ask:

"What assumptions do we make every day in our plant?"

You'll hear answers like:

Operator error
Maintenance issue
Supplier problem
HR issue

Then ask

"What if Vedha was analysing your factory?"

The room usually goes silent.

The Golden Quote for Your Workshop

I would end this discussion with:

"Vikram solved crimes by collecting evidence. Vedha solved minds by changing interpretations."

And then relate it to management:

"Data rarely changes. Interpretation changes. Great managers analyse data. Great leaders challenge their own interpretation of the data."

Why Vikram Vedha Is Ideal for Senior Managers

The movie teaches, through storytelling, almost every major Decision Science concept:

Decision Science Concept	How the Movie Demonstrates It
Confirmation Bias	Vikram initially interprets all evidence to fit his "good vs evil" worldview.
Incremental Learning	Each of Vedha's stories adds new data that updates earlier conclusions.
Contextual Decision Making	The same action appears right or wrong depending on surrounding circumstances.
Systems Thinking	Individual crimes are linked to a larger network of incentives and corruption.
Root Cause Analysis	The apparent culprit is often not the real cause.
Hypothesis Revision	Vikram repeatedly abandons earlier beliefs when new evidence emerges.

For a Data & Decision Science for Management workshop, Moneyball teaches analytics over intuition, while Vikram Vedha teaches how leaders must continuously update their decisions as new evidence emerges. Used together, they make a remarkably powerful combination.

One-Day Training Program

Data & Decision Science for Management

Duration: 1 Day (6 Hours)

Target Audience

Managers
Assistant Managers
Deputy Managers
Functional Heads
Team Leaders
Cross-functional Project Teams

Objective
Equip managers with practical knowledge of using data, analytical thinking, AI, and decision science techniques to make faster, smarter, and evidence-based business decisions.

Learning Outcomes

Participants will learn to:

✔ Think using data instead of assumptions

✔ Interpret dashboards confidently

✔ Identify the right KPIs

✔ Perform root cause analysis using data

✔ Predict business outcomes

✔ Reduce decision bias

✔ Improve operational efficiency

✔ Apply AI in managerial decision making

Session 1

Why Data is the New Competitive Advantage

Duration: 45 Minutes

Topics

Industry 4.0
Smart Manufacturing
Digital Transformation
Data-driven Organizations
Evolution from MIS to AI
Why decisions fail

Activity

Guess vs Data

Managers make decisions based only on intuition.

Later they receive data.

Compare both decisions.

Learning:

Data changes decisions.

Session 1 Activity

Activity: "The Manager's Decision Challenge – Guess vs Data"

Duration: 20–25 Minutes

Objective
Demonstrate how assumptions can lead to incorrect decisions and how data improves the quality of managerial decision-making.

Step 1: Divide Participants

Divide the participants into teams of 4–6 members.

Each team represents the management committee of an automobile manufacturing plant.

Step 2: Present the Scenario (Without Data)

Scenario

The Managing Director receives complaints that production has dropped by 18% during the last month.

He asks the managers:

"What do you think is causing the problem?"

Each team has 5 minutes to discuss and write down:

Top three possible causes
Their recommended action
Confidence level (High / Medium / Low)

Most teams usually guess:

Employee absenteeism
Machine breakdown
Low motivation
Material shortage
Poor supervision
Supplier delay

Step 3: Reveal the Actual Data

Provide the teams with the following dashboard.

KPI	Value
Production	↓18%
Machine Downtime	1%
Employee Attendance	98%
Material Availability	99%
Customer Orders	+25%
Overtime Hours	+45%
Rejection Rate	11%
New Operators Joined	38
Training Hours	0
Supervisor Ratio	1:32

Step 4: Ask Again

Now ask:

"What is your decision now?"

Most teams realize that the issue is not absenteeism or machine breakdown.

Instead, they identify:

New operators were not trained.
High overtime caused fatigue.
Increased rejection reduced effective output.
Supervisors were handling too many operators.
Higher customer demand exposed capability gaps.

Step 5: Group Discussion

Ask each team:

What changed after seeing the data?
Which assumptions proved incorrect?
Which KPI had the biggest impact?
What additional data would you request before making a final decision?

Facilitator Debrief

Highlight the following messages:

❌ Assumptions can be expensive.

✔ Data reveals the real problem.

✔ Good managers ask for evidence before making decisions.

✔ More data does not always mean better decisions—the right data does.

✔ Decisions supported by data reduce bias and improve organizational performance.

Learning Outcome

Participants will understand that:

Intuition is useful but incomplete.
Data validates or challenges assumptions.
Decisions improve when managers rely on measurable evidence rather than opinions.
Data-driven organizations consistently make faster, more accurate, and more defensible decisions.

Bonus Variation: "The Hidden KPI"

Show only five KPIs initially:

Production
Absenteeism
Downtime
Overtime
Material Availability

Ask teams to identify which additional KPI they need before deciding.

Then progressively reveal:

Training Hours
Rejection Rate
First Pass Yield
Supervisor Ratio
Employee Experience
Customer Complaints

This teaches managers that asking the right questions is as important as analyzing the available data, reinforcing a key principle of decision science: better decisions begin with better information.

Session 2 Activity

Activity: "The Decision Lab – Fix the Right Problem"

Duration: 25–30 Minutes

Objective

Enable managers to apply the Decision Science Framework (Observe → Analyze → Interpret → Decide → Review) while recognizing common decision biases and distinguishing between operational, tactical, and strategic decisions.

Activity Setup

Divide participants into teams of 5–6 members.

Each team acts as the Plant Leadership Team of an automobile manufacturing company.

Assign roles:

Production Manager
Maintenance Manager
Quality Manager
HR Manager
Supply Chain Manager
Plant Head (Decision Maker)

Scenario

Production output has fallen by 15% over the last two weeks.

The Managing Director asks the leadership team to restore production immediately.

Teams are given only limited information initially.

Round 1 – Observe (Without Complete Data)

Provide only the following information.

Information	Status
Production	↓15%
Customer Complaints	Increasing
Overtime	High
Workers Available	100%
Delivery Delays	Yes

Ask teams to answer:

What is happening?
What decision would you make immediately?
Is this an Operational, Tactical, or Strategic decision?
How confident are you? (1–5)

Most teams will recommend:

Hire more workers
Increase overtime
Add another shift
Speed up production

Round 2 – Analyze

Now reveal additional operational data.

KPI	Value
Machine Downtime	22%
Preventive Maintenance	Overdue
Material Availability	100%
Operator Attendance	98%
Machine Utilization	61%
Quality Rejections	Normal

Ask:

What changed?
What is the real bottleneck?
What additional information would you request?

Round 3 – Interpret

Provide the root cause.

A critical CNC machine has been breaking down repeatedly because preventive maintenance was postponed for three weeks. Production teams compensated by increasing overtime, but output continued to decline.

Now ask teams to identify:

Root Cause
Symptoms
Immediate Action
Long-Term Solution

Round 4 – Decide

Each team prepares a one-minute presentation.

Their decision must include:

Operational Decision

What should be done today?

Examples:

Schedule emergency maintenance
Redistribute jobs
Prioritize customer orders

Tactical Decision

What should be done this month?

Examples:

Revise maintenance schedule
Cross-train operators
Improve spare parts availability

Strategic Decision

What should be done permanently?

Examples:

Invest in predictive maintenance
Install IoT sensors
Adopt AI-based maintenance planning
Replace aging equipment

Round 5 – Review

Ask each team:

Was your first decision correct?
Which assumptions were wrong?
Which data changed your thinking?
How would you prevent similar mistakes?

Bias Challenge

Ask each team to identify which biases appeared during their decision-making.

Bias	Example in Activity
Confirmation Bias	"Production is low because we always have manpower issues."
Survivorship Bias	"Increasing overtime worked last year, so it must work again."
Availability Bias	"The last crisis was due to absenteeism, so this must be the same."
Anchoring Bias	Fixating on the initial 15% production drop and ignoring later evidence.
Action Bias	Feeling compelled to "do something" (hire more workers) instead of diagnosing the real issue.

Decision Science Worksheet

Each team completes the following template:

Step	Team Response
Observe	What do we know?
Analyze	What data is available?
Interpret	What is the likely root cause?
Decide	What decision should be taken?
Review	How will we measure success?

Facilitator Debrief

Conclude with these key messages:

Symptoms are not the same as root causes.
Good managers resist the urge to jump to solutions before analyzing evidence.
Operational decisions solve today's problems, tactical decisions improve processes, and strategic decisions build long-term capability.
A structured decision framework reduces bias and leads to more consistent, evidence-based outcomes.

Optional Competitive Twist: "Best Decision Wins"

Award points to each team based on:

Criteria	Points
Correctly identified the root cause	30
Followed the Observe → Analyze → Interpret → Decide → Review framework	20
Correctly classified operational, tactical, and strategic decisions	20
Identified decision biases	15
Proposed sustainable, data-driven solutions	15

The team with the highest score is recognized as the "Decision Science Champions." This gamified approach keeps participants engaged while reinforcing analytical thinking and structured decision-making.

Session 3: Understanding Manufacturing Data

Duration: 60 Minutes

This session should be highly interactive because managers learn best by interpreting real business data rather than memorizing KPI definitions.

Activity 1: KPI Detective

Duration: 20 Minutes

Objective

Help participants understand the relationship between different business functions through data.

Scenario

You are the management committee of an automobile manufacturing company.

The Plant Director has received the following monthly dashboard.

Study the data and identify:

Which department is performing well?
Which department requires immediate attention?
Which KPI worries you the most?
Which KPIs are related to one another?

Manufacturing Dashboard (Dataset)

Production

KPI	Target	Actual
Vehicles Produced	12,000	11,050
Line Output (Vehicles/Hour)	55	49
OEE	85%	72%
Machine Downtime	<4%	11%
Cycle Time	58 sec	72 sec
Shift Efficiency	95%	81%

Quality

KPI	Target	Actual
First Pass Yield	98%	93%
Defects per Vehicle	2	6
Rework Rate	1%	5%
PPM	250	980
Warranty Claims	45	108

Supply Chain

KPI	Target	Actual
Supplier On-Time Delivery	98%	89%
Inventory Days	10	18
Material Stock-outs	0	7
Supplier Defect Rate	0.5%	2.1%
Lead Time	4 Days	8 Days

Human Resources

KPI	Target	Actual
Attendance	98%	95%
Attrition	<1%	3.8%
Training Hours	10 hrs	2 hrs
Productivity	96%	84%
Overtime Hours	8 hrs	26 hrs

Finance

KPI	Target	Actual
Cost per Vehicle	₹5.8 Lakhs	₹6.3 Lakhs
Scrap Cost	₹8 Lakhs	₹18 Lakhs
Energy Cost	₹75 Lakhs	₹96 Lakhs
Maintenance Cost	₹42 Lakhs	₹68 Lakhs

Sales

KPI	Target	Actual
Dealer Satisfaction	92%	80%
Customer Satisfaction	94%	85%
Market Demand	12,500	13,200
Delivery Delay	2 Days	6 Days

Team Discussion Questions

Which KPI would you investigate first?
Which department is creating the biggest business impact?
What could be the root cause?
Which KPIs appear to influence one another?

Activity 2: KPI Domino Effect

Duration: 15 Minutes

Objective

Understand that one KPI affects multiple business functions.

Scenario Card

Show only one KPI.

Machine Downtime = 11%

Ask each team to predict what other KPIs may change.

Expected Answers

Machine Downtime

↓

Lower OEE

↓

Higher Cycle Time

↓

Reduced Production

↓

Delivery Delay

↓

Dealer Dissatisfaction

↓

Customer Dissatisfaction

↓

Higher Cost per Vehicle

↓

Revenue Loss

Teams draw the chain on a flip chart.

Award points for identifying the longest logical chain.

Activity 3: Leading vs Lagging Indicator Game

Duration: 15 Minutes

Provide KPI cards (one KPI per card).

Each team places them under either:

Leading Indicators

Lagging Indicators

KPI Cards

Employee Training Hours
Preventive Maintenance
Attendance
Machine Downtime
Supplier Audit Score
Inventory Accuracy
OEE
Customer Complaints
Warranty Claims
Profit
Revenue
Defect Rate
Rework
Delivery Delay
Employee Engagement
Attrition
Cost per Vehicle
Scrap Cost
Customer Satisfaction
Production Output

Answer Key

Leading Indicators

Employee Training Hours
Preventive Maintenance
Attendance
Supplier Audit Score
Inventory Accuracy
Employee Engagement
Machine Health Score
Calibration Compliance
Safety Audits
Process Capability (Cp/Cpk)

Lagging Indicators

Production Output
OEE
Defect Rate
Customer Complaints
Warranty Claims
Profit
Revenue
Attrition
Scrap Cost
Delivery Delay
Customer Satisfaction
Cost per Vehicle

Activity 4: Find the Story Behind the Data

Duration: 10 Minutes

Instead of asking participants to calculate anything, ask them to tell the business story.

Dataset

Month	Jan	Feb	Mar	Apr	May	Jun
OEE	84	83	81	79	75	72
Downtime (%)	3	4	5	7	9	11
Training Hours	10	8	6	4	3	2
Rework (%)	1	2	2	3	4	5
Warranty Claims	35	42	50	67	81	108
Cost per Vehicle (₹ Lakhs)	5.8	5.8	5.9	6.0	6.1	6.3

Questions

What trends do you observe?
Which KPI changed first?
Which KPIs are consequences rather than causes?
If you were the Plant Head, where would you intervene first?
Which department should lead the corrective action?

Facilitator Debrief

Emphasize these key insights:

Data becomes valuable only when connected across functions.
Leading indicators are predictive—they help prevent problems before they occur.
Lagging indicators measure the outcomes of past decisions.
Managers should focus on improving leading indicators, as lagging indicators naturally improve as a result.
A single operational issue, such as rising machine downtime or declining training hours, can cascade into quality defects, delayed deliveries, increased costs, and reduced customer satisfaction.

These activities encourage managers to think like business leaders by interpreting relationships within data rather than viewing KPIs in isolation.

Session 4: Data Visualization & Dashboard Thinking

Duration: 45 Minutes

Activity 1: Dashboard Doctor (Recommended)

Duration: 20 Minutes

Objective

Teach participants that a dashboard should tell a story, not just display data.

Scenario

Your company CEO receives two dashboards from two different managers.

Only one dashboard helps him make quick decisions.

Your job is to identify:

Which dashboard is better?
What is wrong with the other dashboard?
How would you improve it?

Dashboard A (Poor Dashboard)

KPI	Jan	Feb	Mar	Apr	May	Jun
Production	10000	9800	10100	9600	11000	9500
OEE	78	82	80	76	85	73
Downtime	8	7	9	10	6	11
Defects	3	2	5	4	3	6
Rework	2	2	3	5	2	5
Scrap Cost	15	18	14	19	16	22
Attendance	95	96	95	94	96	93
Energy	84	83	86	87	82	91

Problems

No targets
No colours or alerts
Too many numbers
No trends
Equal importance to every KPI
No indication of performance status

Dashboard B (Good Dashboard)

KPI	Target	Actual	Status
Production	12,000	11,050	🔴
OEE	85%	72%	🔴
Machine Downtime	<4%	11%	🔴
First Pass Yield	98%	93%	🟠
Delivery Performance	98%	91%	🟠
Customer Satisfaction	95%	94%	🟡
Safety Incidents	0	0	🟢

Discussion Questions

Which dashboard would a CEO prefer?
Which dashboard supports faster decisions?
What important information is missing?
Which KPIs deserve the most attention?

Activity 2: Choose the Right Chart

Duration: 15 Minutes

Objective

Help managers understand which visualization best communicates different types of business information.

Each team receives the following scenarios and selects the most appropriate chart.

Business Situation	Best Chart
Monthly Production Trend	Line Chart
Defects by Category	Pareto Chart
Department-wise Attrition	Bar Chart
Temperature vs Defect Rate	Scatter Plot
Machine Utilization Across Plant	Heat Map
Process Stability Over Time	Control Chart
Daily Sales Performance	Line Chart
Supplier Performance Comparison	Bar Chart
Top 10 Customer Complaints	Pareto Chart
Machine Downtime by Shift	Heat Map

Discussion

Ask teams:

Why did you choose that chart?

Activity 3: Wrong Chart Challenge

Duration: 10 Minutes

Objective

Identify misleading visualizations.

Display the following examples.

Example 1

Monthly Production Trend

Shown as a Pie Chart

Question:

❓ Is this correct?

Answer:

❌ No

✔ Use a Line Chart

Example 2

Department-wise Attrition

Shown as a Scatter Plot

Answer:

❌ Wrong

✔ Use a Bar Chart

Example 3

Defects by Category

Shown as a Line Chart

Answer:

❌ Wrong

✔ Use a Pareto Chart

Example 4

Machine Temperature vs Defect Rate

Shown as a Bar Chart

Answer:

❌ Wrong

✔ Use a Scatter Plot

Example 5

Daily OEE Stability

Shown as a Pie Chart

Answer:

❌ Wrong

✔ Use a Control Chart or Line Chart

Bonus Activity: Build the CEO Dashboard

Duration: 10 Minutes

Each team can display only six KPIs on the CEO's dashboard.

Choose from:

Production
OEE
Downtime
First Pass Yield
Defects
Rework
Inventory
Supplier Delivery
Attrition
Attendance
Energy Cost
Cost per Vehicle
Revenue
Customer Satisfaction
Warranty Claims
Safety
Scrap Cost

Questions

Which six KPIs would you choose?
Why are they critical?
Which chart would you use for each KPI?

KPI	Recommended Chart
Production Trend	Line Chart
OEE	Gauge or Bar
Downtime by Machine	Bar Chart
Defects by Type	Pareto Chart
Customer Satisfaction Trend	Line Chart
Cost per Vehicle	Line Chart

Facilitator Debrief

Conclude the session with these key principles:

A dashboard is a decision-making tool, not a data repository.
Every chart should answer a specific business question.
Choose charts based on the nature of the data:

Bar Chart: Compare categories.
Line Chart: Show trends over time.
Pareto Chart: Prioritize the vital few causes (80/20 principle).
Scatter Plot: Reveal relationships or correlations between variables.
Heat Map: Highlight intensity or performance across locations, machines, or time.
Control Chart: Monitor process stability and identify unusual variation.

The best dashboards are simple, actionable, and focused on exceptions rather than overwhelming users with every available metric.

This activity is particularly effective for managers because it simulates the real-world challenge of converting operational data into executive insights, reinforcing the principle that good visualization accelerates good decisions.

Session 5: Root Cause Analysis Using Data

Duration: 60 Minutes

Activity: "The Quality Crisis Investigation"

Objective

Participants will analyze a realistic automobile manufacturing case using multiple root cause analysis tools:

Pareto Analysis (80/20 Rule)
Fishbone (Ishikawa)
5 Why Analysis
Correlation Analysis
Trend Analysis

The goal is to identify:

Root Cause
Corrective Action
Preventive Action

Business Scenario

You are the Cross Functional Investigation Team (Production, Quality, Maintenance, Supply Chain, HR and Customer Service) of an automobile manufacturing company.

During the last month:

Customer complaints increased by 18%
Warranty claims increased by 22%
Social media complaints have started increasing.
One major dealer has threatened to stop accepting deliveries.

The Managing Director has asked your team to investigate the real cause within one hour.

Dataset 1 – Monthly Customer Complaints

Month	Vehicles Sold	Customer Complaints	Complaint %
January	10,200	165	1.62%
February	10,500	172	1.64%
March	10,800	176	1.63%
April	11,000	182	1.65%
May	11,400	190	1.67%
June	11,700	224	1.91%

Observation

Customer complaints increased approximately 18% in June.

Dataset 2 – Complaint Categories

Complaint Type	Number
Door Misalignment	82
Paint Peeling	55
Engine Noise	28
Brake Noise	20
Electrical Issues	18
AC Cooling	12
Dashboard Rattle	9

Activity 1 – Pareto Analysis

Ask participants:

Which complaint should be investigated first?

Expected Answer:

Door Misalignment + Paint Peeling account for nearly 60–65% of all complaints, making them the highest priority.

Dataset 3 – Production Data

Week	OEE	Downtime	Production
Week 1	84%	4%	2,850
Week 2	82%	5%	2,790
Week 3	78%	8%	2,620
Week 4	74%	12%	2,410

Dataset 4 – Quality Data

Week	First Pass Yield	Rework	Defects
Week 1	98%	1.2%	42
Week 2	97%	2.0%	58
Week 3	95%	3.5%	89
Week 4	92%	5.8%	148

Dataset 5 – Maintenance Data

Machine	Downtime (hrs)	Breakdown
Robotic Welding Cell	18	4
Paint Booth	27	6
Assembly Conveyor	8	2
Torque Station	4	1
Inspection Camera	2	1

Dataset 6 – Human Resource Data

KPI	Last Month	Current Month
Attrition	2%	5%
New Operators	4	26
Training Hours	12 hrs	3 hrs
Attendance	97%	94%
Overtime	11 hrs	29 hrs

Dataset 7 – Supplier Quality

Supplier	Defect Rate
Supplier A	0.5%
Supplier B	0.8%
Supplier C	3.8%
Supplier D	0.6%

Supplier C supplies door hinges.

Dataset 8 – Correlation Exercise

Week	New Operators	Training Hours	Defects
1	5	12	42
2	8	10	58
3	15	6	92
4	26	3	148

Discussion

Ask participants:

Which variables seem related?
Does lower training coincide with higher defects?
Could rapid hiring without adequate training be contributing to quality problems?

Expected insight:

As new operators increased and training hours decreased, defects rose sharply, indicating a likely correlation.

Dataset 9 – Trend Analysis

Month	Downtime	Rework	Complaints
Jan	4%	1.2%	165
Feb	5%	1.5%	172
Mar	6%	2.1%	176
Apr	8%	3.2%	182
May	10%	4.4%	190
Jun	12%	5.8%	224

Discussion

What trend do you observe?

Expected Answer:

Downtime steadily increased.
Rework increased alongside downtime.
Customer complaints followed the same upward trend.
This suggests process deterioration over time rather than a sudden isolated event.

Activity 2 – Fishbone (Cause & Effect)

Ask each team to build a Fishbone Diagram under these categories:

Category	Possible Causes
Man	New operators, insufficient training, overtime fatigue
Machine	Paint booth failures, welding robot downtime
Material	Defective door hinges from Supplier C
Method	Poor inspection frequency, maintenance delays
Measurement	Infrequent quality audits, delayed feedback
Environment	High humidity affecting paint curing

Activity 3 – 5 Why Analysis

Problem: Door Misalignment Complaints Increased

Why 1?
Door alignment out of specification.

Why 2?
Door hinges shifted during assembly.

Why 3?
Welding fixture lost calibration.

Why 4?
Preventive maintenance was postponed.

Why 5?
Maintenance team was diverted to emergency breakdowns due to staffing shortages.

Root Cause

Preventive maintenance was neglected, leading to fixture misalignment and increased customer complaints.

Final Team Challenge

Each team presents:

1. Root Cause

Examples:

Inadequate preventive maintenance
Insufficient operator training
Poor supplier quality
Delayed process inspections

2. Corrective Actions (Immediate)

Examples:

Repair and recalibrate welding fixtures
Replace defective hinge inventory
Retrain assembly operators
Increase inspection frequency
Contain suspect vehicles before shipment

3. Preventive Actions (Long-Term)

Examples:

Predictive maintenance program
Automated vision inspection for door alignment
Mandatory onboarding and certification for new operators
Supplier quality improvement plans and audits
Statistical Process Control (SPC) with real-time alerts
Weekly cross-functional quality reviews

Facilitator Debrief

Conclude by reinforcing these lessons:

Symptoms (customer complaints) are outcomes, not causes.
Data from multiple departments must be connected to uncover the real issue.
Effective root cause analysis combines evidence from Pareto, trends, correlation, Fishbone, and 5 Whys rather than relying on assumptions.
Sustainable quality improvement comes from eliminating root causes and implementing preventive systems—not just fixing immediate defects.

This integrated case closely mirrors investigations conducted in automotive manufacturing environments using Lean, Six Sigma, IATF 16949, and Toyota Production System practices, making it highly relevant for managers and cross-functional teams.

Session 6: Predictive Thinking

Duration: 45 Minutes

Activity: "The Future Factory Challenge"

Objective

Help managers understand the evolution of decision-making:

Reactive → Proactive → Predictive → Prescriptive

without using mathematics. Participants will learn how business data can be used to anticipate future events and recommend the best course of action.

Scenario

You are the Management Committee of an automobile manufacturing plant.

The Plant Director asks:

"Based on the available data, what is likely to happen next month, and what should we do today to prevent problems?"

Each team will analyze five real-world business datasets and answer:

What is happening?
What is likely to happen?
What action would you recommend?

Dataset 1: Predicting Machine Failure

Week	Machine Downtime (hrs)	Vibration Level (mm/s)	Temperature (°C)	Maintenance Overdue (Days)
1	2	2.1	48	0
2	3	2.8	51	2
3	5	3.9	56	6
4	8	5.5	64	10
5	12	7.1	73	15

Discussion Questions

What trend do you notice?
Which machine is at risk?
Should maintenance wait until the machine fails?

Expected Business Insight

Reactive

Repair after breakdown.

↓

Proactive

Schedule preventive maintenance.

↓

Predictive

Predict failure based on increasing vibration, temperature, and overdue maintenance.

↓

Prescriptive

Schedule maintenance during planned shutdown and replace the worn bearing before failure.

Dataset 2: Predicting Employee Attrition

Employee	Overtime (hrs/month)	Training Hours	Engagement Score (/100)	Absent Days	Resigned?
A	8	14	90	1	No
B	14	10	82	2	No
C	28	4	60	6	Yes
D	32	2	52	8	Yes
E	18	8	75	3	No
F	36	1	48	9	Yes

Discussion Questions

Which employees appear to be at risk?
What early warning signs do you observe?
How could HR intervene before resignations occur?

Expected Business Insight

High overtime + low training + low engagement + increased absenteeism are warning signs of potential attrition.

Dataset 3: Sales Forecast

Month	Dealer Orders
January	9,800
February	10,200
March	10,850
April	11,600
May	12,400
June	13,300

Questions

What trend do you observe?
If the pattern continues, what might July demand be?
What preparations should production and procurement teams make?

Expected Answer

Demand is steadily increasing.

Managers should:

Increase production planning
Secure raw materials
Prepare logistics
Schedule additional shifts if needed

Dataset 4: Supplier Risk Prediction

Supplier	On-Time Delivery	Defect Rate	Financial Rating	Late Deliveries (Last 6 Months)
A	98%	0.5%	AAA	0
B	96%	0.8%	AA	1
C	82%	3.6%	BBB	7
D	99%	0.4%	AAA	0

Questions

Which supplier is risky?
What could happen if no action is taken?
What should procurement do now?

Expected Business Insight

Supplier C poses a significant risk due to poor delivery performance, high defect rates, and a weaker financial rating. The procurement team should qualify alternate suppliers, increase incoming inspections, and develop a supplier improvement plan.

Dataset 5: Demand Forecast

Month	Festival Season	Marketing Campaign	Demand
January	No	No	10,000
February	No	No	10,400
March	No	Yes	11,500
April	Yes	Yes	13,400
May	Yes	No	12,800
June	No	No	11,300

Discussion

Ask participants:

What factors influence demand?
Besides historical sales, what external factors should managers monitor?

Expected answers:

Festival season
Promotions and advertising
Fuel prices
Interest rates
Competitor launches
Government incentives
Economic conditions

Group Exercise: Match the Business Problem

Give each team the following scenarios and ask them to identify the appropriate predictive technique.

Business Problem	Best Technique
Predict next month's vehicle demand	Forecasting
Predict whether an employee may resign	Classification
Estimate next month's energy cost	Regression
Predict machine failure before breakdown	Classification / Predictive Maintenance
Estimate production output based on overtime	Regression
Forecast inventory requirements	Forecasting

Business Understanding (No Mathematics)

Technique	Manager-Friendly Explanation	Manufacturing Example
Regression	Predicts a number or value	Estimate monthly production, energy cost, or sales
Classification	Predicts a category (Yes/No, High/Low)	Will an employee resign? Will a machine fail? Is a supplier high risk?
Forecasting	Predicts future trends using historical patterns	Forecast vehicle demand, inventory needs, or sales volumes

Final Activity: Predict, Then Prescribe

Each team selects one dataset and completes the following worksheet:

Question	Team Answer
What trend do you observe?
What is likely to happen if nothing changes?
Which business area is at risk?
What action should management take today?
What KPI should be monitored going forward?

Facilitator Debrief

Conclude the session by reinforcing the maturity of decision-making:

Stage	Manager's Question	Example
Reactive	What happened?	"The machine has broken down."
Proactive	How can we prevent it?	"Let's perform preventive maintenance every month."
Predictive	What is likely to happen next?	"Sensor data indicates this machine may fail within two weeks."
Prescriptive	What is the best action to take?	"Replace the bearing during the planned shutdown next Friday to avoid production loss."

Key Learning Points

Data is most valuable when it helps anticipate future events rather than merely explaining the past.
Predictive thinking enables managers to shift from firefighting to prevention.
Modern AI and analytics support managers by identifying patterns and risks, but business judgment remains essential in deciding the best course of action.
Organizations that adopt predictive and prescriptive decision-making reduce costs, improve quality, and respond more effectively to changing business conditions.

Session 7: AI for Decision Making

Duration: 45 Minutes

Activity: "AI Consultant Challenge – Solve the Business Problem"

Objective

Enable managers to understand how AI can assist (not replace) managerial decision-making by solving real manufacturing problems using Generative AI, Machine Learning, Business Intelligence, Predictive Analytics, Microsoft Copilot, and ChatGPT.

The emphasis is on business applications, not technical concepts.

Activity Setup

Divide participants into 6 Cross-Functional Teams.

Each team represents one department:

Team 1 – Production
Team 2 – Quality
Team 3 – Maintenance
Team 4 – Supply Chain
Team 5 – Human Resources
Team 6 – Sales & Customer Service

Each team receives a real-life business scenario and dataset.

Their task:

Understand the problem
Identify useful data
Decide where AI can help
Recommend an AI-powered solution
Explain expected business benefits

Team 1 – Production Planning

Scenario

Customer demand has become unpredictable.

Sometimes the factory produces excess inventory.

Sometimes dealers wait for vehicles.

Dataset

Month	Customer Orders	Production	Inventory	Stock-out
Jan	9,800	10,200	400	No
Feb	10,600	10,200	0	Yes
Mar	11,400	10,700	0	Yes
Apr	10,200	11,000	800	No
May	12,500	11,300	0	Yes
Jun	11,700	12,200	500	No

Questions

What is the business problem?
Which AI solution could help?
What decisions would improve?

Expected Answer

AI can forecast demand and recommend optimal production schedules to balance inventory and customer demand.

Team 2 – Quality Inspection

Scenario

Quality inspectors miss some defects because inspection is manual.

Dataset

Month	Vehicles Produced	Manual Inspection Defects Found	Customer Complaints
Jan	10,000	148	162
Feb	10,400	152	171
Mar	10,900	160	182
Apr	11,100	165	210
May	11,500	170	228

Questions

Can AI improve inspection?
How?

Expected Answer

AI-powered computer vision can inspect paint, welds, gaps, and surface defects in real time, improving consistency and reducing escapes.

Team 3 – Predictive Maintenance

Scenario

Machines stop unexpectedly.

Production losses increase every month.

Dataset

Machine	Temperature	Vibration	Downtime (hrs)
CNC-1	58	2.5	3
CNC-2	72	5.8	11
Robot-1	74	6.3	14
Paint Booth	69	5.2	9

Questions

Which machines need attention?
Can AI predict failures?

Expected Answer

Yes. AI can analyze sensor patterns and recommend maintenance before breakdowns occur, reducing downtime.

Team 4 – Inventory Optimization

Scenario

Warehouse costs are increasing.

Some parts are overstocked while others run out.

Dataset

Part	Stock	Monthly Usage	Lead Time (Days)
Tyres	2,200	2,000	5
Door Handles	8,000	2,000	7
Engine Mounts	500	700	15
Sensors	300	600	21

Questions

Which parts are at risk?
How can AI help?

Expected Answer

AI can forecast demand, optimize reorder points, identify slow-moving inventory, and minimize stock-outs and excess inventory.

Team 5 – HR Analytics

Scenario

Employee resignations have increased.

Dataset

Department	Attrition	Overtime	Engagement Score
Assembly	7%	34 hrs	54
Welding	3%	18 hrs	76
Paint Shop	8%	36 hrs	48
Quality	2%	12 hrs	84

Questions

Which department needs attention?
Can AI identify employees at risk?

Expected Answer

AI can identify attrition patterns, predict high-risk employees, and suggest targeted interventions such as workload balancing, career development, or recognition programs.

Team 6 – Customer Complaints

Scenario

Thousands of customer feedback comments are received every month.

Managers cannot manually read all of them.

Dataset

Complaint Category	Number
Paint Quality	182
Delivery Delay	154
Engine Noise	118
AC Cooling	86
Dashboard Noise	62
Electrical Issues	48

Questions

How can AI help?
Which complaint should management prioritize?

Expected Answer

Generative AI and Natural Language Processing (NLP) can automatically classify complaints, identify recurring themes, perform sentiment analysis, and summarize customer feedback for faster action.

Activity 2 – AI Tool Matching

Duration: 10 Minutes

Ask teams to match the business problem with the most suitable AI technology.

Business Problem	Best AI Tool
Generate meeting minutes	Generative AI (ChatGPT/Copilot)
Predict machine failure	Machine Learning
Create management dashboards	Business Intelligence (Power BI)
Forecast sales	Predictive Analytics
Analyze customer feedback	Generative AI + NLP
Predict employee attrition	Machine Learning
Explain monthly KPIs	Copilot
Inventory forecasting	Predictive Analytics

Activity 3 – Prompt Engineering Challenge

Duration: 10 Minutes

Each team writes a prompt they could use with ChatGPT or Microsoft Copilot.

Example 1 – Production

"Analyze this production report. Identify the top three reasons for lower output, suggest corrective actions, and summarize the findings for the Plant Head."

Example 2 – HR

"Review this employee attrition dataset. Identify high-risk departments, possible causes, and recommend three retention initiatives."

Example 3 – Quality

"Analyze the defect report. Identify recurring defect patterns, rank them by business impact, and recommend preventive actions."

Example 4 – Supply Chain

"Review inventory levels, supplier lead times, and monthly demand. Identify stock-out risks and recommend optimal reorder quantities."

Final Discussion

Each department answers:

Where can AI help your department?

Department	Possible AI Applications
Production	Production scheduling, bottleneck detection, capacity planning
Quality	Visual inspection, defect prediction, root cause analysis
Maintenance	Predictive maintenance, spare parts planning
Supply Chain	Demand forecasting, supplier risk analysis, inventory optimization
HR	Attrition prediction, recruitment screening, workforce planning, learning recommendations
Finance	Budget forecasting, fraud detection, cost variance analysis
Sales	Sales forecasting, pricing optimization, dealer performance analysis
Customer Service	Complaint classification, sentiment analysis, chatbot support

Facilitator Debrief

Conclude the session with the following key messages:

AI supports—not replaces—managerial decision-making.

Managers bring business context, ethical judgment, and strategic thinking. AI accelerates analysis and highlights insights, but people remain accountable for decisions.

Different AI technologies serve different purposes.

Technology	Business Purpose
Generative AI (ChatGPT/Copilot)	Generate reports, summarize information, draft emails, brainstorm ideas, explain trends
Machine Learning	Predict failures, classify risks, detect anomalies, identify patterns
Business Intelligence (Power BI, Tableau)	Monitor KPIs, visualize performance, build interactive dashboards
Predictive Analytics	Forecast demand, production, sales, maintenance, and workforce trends

Final Reflection Question

Ask every participant:

"If you returned to work tomorrow with an AI assistant sitting beside you, what is the first repetitive, time-consuming, or data-heavy task in your department that you would ask it to help with?"

This closing discussion encourages participants to move from understanding AI concepts to identifying practical, high-impact opportunities within their own functions, making the learning immediately applicable.

Session 8: Management Decision Simulation (Capstone Activity)

Duration: 60 Minutes

Activity: "The Plant Management War Room"

Objective

This capstone simulation integrates everything covered in the training:

Data Interpretation
Dashboard Thinking
Root Cause Analysis
Decision Science
Predictive Thinking
AI-assisted Decision Making
Cross-functional Collaboration

Participants work as the Executive Management Committee of an automobile manufacturing plant and make evidence-based decisions under time pressure.

Scenario

You have just joined the Monday Morning Plant Performance Review.

The Managing Director walks into the meeting and says:

"Last month's performance is unacceptable. Customer complaints have increased, production has dropped, dealers are unhappy, and costs are rising. I need a recovery plan within the next 45 minutes."

Each team receives the Plant Performance Dashboard.

Plant Executive Dashboard (Dataset)

Production Dashboard

KPI	Target	Actual	Previous Month
Vehicles Produced	12,000	10,850	11,900
OEE	85%	73%	82%
Machine Downtime	<4%	11%	6%
Line Efficiency	95%	81%	91%
Cycle Time	58 sec	71 sec	61 sec
Overtime Hours	9 hrs	28 hrs	14 hrs

Quality Dashboard

KPI	Target	Actual
First Pass Yield	98%	92%
Rework Rate	1%	6%
Defects per Vehicle	2	7
Customer Complaints	180	212
Warranty Claims	48	69

Supply Chain Dashboard

KPI	Target	Actual
Supplier On-Time Delivery	98%	88%
Inventory Days	10	18
Material Shortages	0	5
Lead Time	5 Days	9 Days

Human Resources Dashboard

KPI	Target	Actual
Attendance	98%	94%
Attrition	<2%	6%
New Operators Joined	5	32
Training Hours	12	3
Employee Engagement	85	59

Finance Dashboard

KPI	Target	Actual
Cost per Vehicle	₹5.8 L	₹6.5 L
Scrap Cost	₹8 Lakhs	₹21 Lakhs
Energy Cost	₹75 Lakhs	₹95 Lakhs

Sales & Customer Dashboard

KPI	Target	Actual
Dealer Satisfaction	92%	78%
Customer Satisfaction	94%	82%
Market Demand	12,500	13,200
Delivery Delay	2 Days	7 Days

Hidden Investigation Data (Reveal After 15 Minutes)

Only after teams complete their initial analysis, distribute this second sheet.

Maintenance

Machine	Breakdowns	PM Overdue
Robotic Welding Cell	6	18 Days
Paint Booth	5	15 Days
Assembly Conveyor	2	4 Days

Supplier Quality

Supplier	Part	Defect Rate
Supplier A	Engine Mount	0.4%
Supplier B	Door Hinges	4.2%
Supplier C	Paint	0.6%

Training

Department	New Operators	Certified
Assembly	18	5
Welding	9	3
Paint Shop	5	2

Customer Complaints

Complaint	Count
Door Misalignment	74
Paint Finish	53
Engine Noise	29
Dashboard Noise	21
Electrical	18

Team Assignment

Each team acts as the Plant Executive Committee.

Complete the following worksheet.

Part A – Interpret the Dashboard

Identify:

Five major business problems
Three KPIs that concern you the most
Which department needs immediate attention?

Part B – Connect the Dots

Draw relationships between KPIs.

Example

Poor Training

↓

Operator Errors

↓

More Defects

↓

Higher Rework

↓

Lower Production

↓

Delivery Delays

↓

Dealer Complaints

↓

Customer Dissatisfaction

Part C – Root Cause Analysis

Identify:

Primary Root Cause

Example:

Preventive maintenance neglected.

Secondary Root Causes

Examples

New operators not certified
Supplier B quality issues
Increased overtime
Low employee engagement

Part D – Management Decisions

Recommend:

Immediate (Today)

Examples

Repair welding robot
Increase inspection
Stop shipment of affected vehicles
Replace defective supplier batches

Short-Term (30 Days)

Examples

Train all new operators
Audit Supplier B
Resume preventive maintenance
Improve production scheduling

Long-Term (6 Months)

Examples

AI Predictive Maintenance
Computer Vision Quality Inspection
Digital OEE Dashboard
Employee Retention Program
Supplier Scorecard
Workforce Planning

Bonus Round – AI Consultant

Ask each team:

If ChatGPT or Microsoft Copilot were part of your management team, what would you ask it to analyze?

Example prompts:

"Analyze this dashboard and identify the top three root causes affecting production."

"Recommend a 90-day improvement plan to reduce customer complaints."

"Predict which KPI is likely to deteriorate next month."

Presentation

Each team has 5 minutes.

Presentation Format

1. Situation

What is happening?

2. Evidence

Which KPIs support your conclusion?

3. Root Cause

Why is it happening?

4. Corrective Actions

What should management do immediately?

5. Preventive Actions

How will this be prevented permanently?

6. AI Recommendation

Where can AI support management?

Scoring Sheet

Evaluation Criteria	Marks
Correct Interpretation of Dashboard	20
Data-Based Root Cause Analysis	20
Cross-Functional Thinking	15
Quality of Corrective Actions	15
Preventive & Strategic Thinking	10
Use of Predictive Thinking	10
AI Recommendation	5
Presentation & Justification	5
Total	100

Expected Ideal Solution

Primary Root Cause

Preventive maintenance was delayed, causing equipment instability and lower OEE.

Contributing Causes

Large intake of new operators with insufficient certification and training.
Poor supplier quality for door hinges leading to misalignment complaints.
Excessive overtime causing fatigue and increased rework.
Weak workforce engagement contributing to higher attrition.

Business Impact

Lower production output despite rising market demand.
Higher defect rates and rework.
Increased scrap and cost per vehicle.
Delivery delays, reduced dealer satisfaction, and lower customer satisfaction.

Executive Action Plan

Immediate (0–7 days): Restore critical equipment, isolate defective parts, reinforce inspections, stabilize production.
Short-Term (30 days): Complete operator certification, audit Supplier B, reintroduce preventive maintenance discipline, rebalance workloads.
Long-Term (3–6 months): Implement predictive maintenance, AI-based visual inspection, integrated Power BI dashboards, supplier performance management, workforce planning, and continuous improvement programs.

Facilitator Debrief

Conclude the workshop with the following message:

"Data does not make decisions—people do. However, managers who can interpret data, distinguish symptoms from root causes, anticipate future risks, and justify their decisions with evidence consistently outperform those who rely solely on intuition. The most effective leaders combine business experience with analytics, structured thinking, and AI to make faster, smarter, and more sustainable decisions."

This capstone simulation brings together all eight sessions into a realistic executive decision-making exercise, closely resembling monthly performance review meetings in leading automobile manufacturing organizations.

Leaders don't just analyze what is visible—they question what is absent, what is assumed, and what is not being measured.

Below are several powerful, real-world scenarios that create the same "Aha!" moment as the WWII aircraft example. These work exceptionally well in leadership development, Lean, Six Sigma, and Decision Science programs.

1. The Hospital Waiting Room

Visible Picture

A hospital has excellent doctors.

Patients still complain.

Management decides:

"Hire more doctors."

Data

Metric	Value
Doctor Consultation	12 minutes
Waiting Time	2 hours

Ask participants:

What is the problem?

Most say

"Need more doctors."

Reveal

Patients spend

Registration = 40 mins
Billing = 25 mins
Pharmacy = 35 mins
Doctor = 12 mins

The doctor was never the bottleneck.

Leadership Lesson

Don't optimize the smallest part of the process.

2. The Iceberg Leadership Exercise

Draw an iceberg.

Visible above water

Sales
Profits
Complaints
Targets

Hidden below water

Culture
Fear
Trust
Leadership
Skills
Processes
Communication
Motivation

Ask

Which part creates the visible problems?

Answer

Everything below the surface.

Lesson

KPIs are symptoms.

Leadership solves causes.

3. The Firefighter Company

Company A

Every week

Employees work overtime

Everyone applauds.

Managers called heroes.

Ask

Is this good?

Most answer

Yes.

Reveal

Overtime exists because

Planning is poor.

The "heroes" created the fire they keep putting out.

Lesson

Never reward firefighting.

Reward fire prevention.

4. The Quiet Employee

Ask

Who is your best employee?

Managers usually answer

"The one who never complains."

Reveal

That employee resigned yesterday.

Exit interview says

"I stopped giving suggestions two years ago."

Lesson

Silence isn't engagement.

Sometimes silence means resignation has already begun.

5. Customer Complaints Reduced

Dashboard

Customer complaints reduced

500

↓

120

Excellent?

Everyone celebrates.

Reveal

Complaint portal crashed.

Customers couldn't complain.

Lesson

Absence of data isn't evidence of success.

6. Empty Factory Floor

Show an image of an almost empty, clean production floor.

Ask

Healthy factory?

Participants answer

Yes.

Reveal

Production stopped because of a supplier shutdown.

Lesson

Context matters more than appearance.

7. High Performer

HR Dashboard

Employee A

100% Target Achievement

Employee B

85%

Manager promotes A.

Reveal

Employee A received

120 leads

Employee B received

40 leads

Who performed better?

Lesson

Normalize data before comparing.

8. The School Ranking

School

100% Pass Rate.

Best school?

Reveal

Only top students were allowed to appear for exams.

Weak students were held back.

Lesson

Selection bias.

9. The Maintenance Hero

Maintenance Manager proudly says

"No machine breakdown for 90 days."

Celebrate?

Reveal

Production demand reduced 60%.

Machines hardly ran.

Lesson

Always relate performance to exposure.

10. Safety Excellence

Factory reports

Zero accidents.

Excellent?

Reveal

Employees stopped reporting near misses because supervisors discouraged reporting.

Lesson

Low reporting doesn't always mean low risk.

11. Best Sales Branch

Branch Chennai

₹12 Crores

Branch Madurai

₹5 Crores

Best branch?

Reveal

Employees

Chennai

120

Madurai

Revenue per employee

Madurai wins.

Lesson

Absolute numbers deceive.

12. AI Hiring

AI selected

95%

Male candidates.

AI biased?

Participants answer

Yes.

Reveal

Training data contained

15 years of historical hiring.

Leadership lesson

AI learns historical decisions.

It doesn't question them.

Leaders must.

13. Employee Engagement Survey

Score 2%

Excellent.

Reveal

Participation 22%

Who answered?

Mostly managers.

Lesson

Response bias.

14. Lean Factory

Before Lean

Inventory ₹75 Crores

After Lean ₹18 Crores

Excellent?

Reveal

Stock-outs increased

Customer delivery delayed.

Lesson

Optimization of one KPI can damage another.

15. The Marathon Winner

Runner wins.

Coach says : Excellent training.

Reveal

All elite runners who dropped out due to injuries were excluded from the analysis.

Lesson

Study failures, not only successes.

16. The CEO Dashboard

Show only:

Revenue ↑

Profit ↑

Market Share ↑

Customer Satisfaction ↑

Ask

Healthy company?

Reveal

Employee Attrition = 38%

Innovation Projects = 0

R&D Budget Cut

Three years later

Company collapses.

Lesson

Leading indicators matter more than lagging indicators.

17. The Empty Suggestion Box

Suggestion Box

No suggestions for six months.

Manager says

Everyone is happy.

Reveal

Employees believe

"No one reads suggestions."

Lesson

No feedback may indicate no psychological safety.

18. The Traffic Signal

A city installs another traffic signal at a congested junction.

Congestion worsens.

Why?

Hidden issue

The nearby shopping mall changed its exit route.

Lesson

Fixing symptoms without understanding the system creates new problems.

19. The Perfect Audit

Audit Result : 100%

No NCs.

Celebrate?

Reveal

Auditors announced the visit one month in advance.

Everything was temporarily corrected.

Lesson

Measure reality, not performance prepared for inspection.

20. The Factory of Mirrors (Grand Leadership Exercise)

Tell participants:

Imagine a factory made entirely of mirrors.

Everywhere you look, you see reflections.

If you see only reflections, can you find the real object?

Leadership is similar.

Reports are reflections.

Dashboards are reflections.

KPIs are reflections.

Complaints are reflections.

Profits are reflections.

The leader's responsibility is to find the real object behind the reflection.

Reflection → Reality

Low productivity → Poor planning? Equipment constraints? Skill gaps?
High absenteeism → Transportation? Leadership? Workload? Shift patterns?
Customer complaints → Product? Process? Communication? Expectations?
High attrition → Compensation? Growth? Manager behavior? Culture?
Low engagement → Fear? Lack of recognition? Unclear purpose?

A Closing Thought for Leaders

Managers look at the dashboard.

Leaders question the dashboard.

Managers solve the problem they can see.

Leaders search for the problem that cannot yet be seen.

Managers ask, "What happened?"

Leaders ask, "What are we not seeing, and why?"

For senior managers in an automobile manufacturing company, these exercises are highly effective because they shift discussions from reporting metrics to systems thinking, cognitive bias awareness, root cause analysis, and evidence-based leadership—the core competencies of modern decision science.

Vanilla Ice Cream Complaint

The "Vanilla Ice Cream Complaint" is one of the most famous business case studies used in Quality Management, Root Cause Analysis, Systems Thinking, Design Thinking, and Data & Decision Science. Although it is often attributed to General Motors (or Pontiac, a GM division), it has become a legendary management story because it demonstrates how leaders should investigate unusual data instead of dismissing it.

Case Study: The Vanilla Ice Cream Complaint

The Complaint

A customer wrote a letter to Pontiac (General Motors):

"Whenever I buy Vanilla Ice Cream, my car refuses to start.

**If I buy Chocolate, Strawberry or any other flavor, it starts perfectly.

Please explain why my car doesn't like Vanilla Ice Cream."**

The complaint sounded ridiculous.

Most companies would have ignored it.

Instead...

Pontiac assigned an engineer to investigate.

The Investigation

The engineer accompanied the customer.

Day 1

Customer bought

Vanilla

Returned to the car.

The car wouldn't start.

Exactly as described.

Day 2

Customer bought

Chocolate.

Returned.

Car started immediately.

Day 3

Strawberry.

No problem.

Day 4

Vanilla.

Again...

Car refused to start.

The pattern repeated several times.

Now the engineer knew:

The customer wasn't imagining it.

There had to be another explanation.

The Real Observation

The engineer didn't focus on the ice cream flavor.

He focused on everything surrounding the purchase.

He noticed:

Vanilla was the best-selling flavor.

It was stored at the front of the freezer.

Customers picked it quickly.

Average purchase time:

30–40 seconds

Other flavors were stored deeper inside.

Customers spent

2–4 minutes

choosing them.

For a management training program, the case study becomes much more powerful if participants receive a realistic observational dataset rather than a story. The table below simulates the engineer's investigation over several days and allows participants to identify patterns through data analysis.

Case Study: The Vanilla Ice Cream Complaint

Engineer's Observation Log

Day	Ice Cream Flavor	Customer Weight (kg)	Fuel Level (%)	Engine Start Time (Home)	Arrival at Store	Engine Stop Time	Purchase Start	Purchase End	Purchase Duration (sec)	Engine Restart Time	Idle Time Before Restart (sec)	Distance Travelled (km)	Avg. Speed (km/h)	Engine Started?
1	Vanilla	78	62	18:00:00	18:12:30	18:12:35	18:12:40	18:13:15	35	18:13:20	45	8.5	42	❌ No
2	Chocolate	78	60	18:01:00	18:13:20	18:13:25	18:13:30	18:16:10	160	18:16:20	175	8.5	42	✅Yes
3	Strawberry	78	58	18:02:00	18:14:10	18:14:15	18:14:20	18:17:05	165	18:17:15	180	8.5	41	✅Yes
4	Vanilla	78	56	18:03:00	18:15:00	18:15:05	18:15:10	18:15:45	35	18:15:50	45	8.5	43	❌ No
5	Mint	78	54	18:04:00	18:16:10	18:16:15	18:16:20	18:19:05	165	18:19:10	175	8.5	42	✅Yes
6	Vanilla	78	52	18:05:00	18:17:00	18:17:05	18:17:10	18:17:40	30	18:17:45	40	8.5	42	❌ No
7	Chocolate	78	50	18:06:00	18:18:20	18:18:25	18:18:30	18:21:10	160	18:21:20	175	8.5	41	✅Yes
8	Vanilla	78	48	18:07:00	18:19:15	18:19:20	18:19:25	18:20:00	35	18:20:05	45	8.5	42	❌ No

Summary Dashboard

Observation	Vanilla	Other Flavours
Average Purchase Time	34 sec	163 sec
Average Idle Time Before Restart	44 sec	176 sec
Engine Restart Success	0%	100%
Fuel Level	No Significant Difference	No Significant Difference
Distance Travelled	Same	Same
Passenger Weight	Same	Same
Average Speed	Same	Same

Hidden Variables Analysis

Variable	Changed?	Relationship with Engine Failure
Ice Cream Flavor	Yes	❌ Coincidental
Passenger Weight	No	❌ No Relationship
Fuel Level	Slightly	❌ No Relationship
Distance Travelled	No	❌ No Relationship
Average Speed	No	❌ No Relationship
Outside Temperature	Same	❌ No Relationship
Purchase Duration	Yes	✅ Strong Relationship
Engine Cooling Time	Yes	✅ Strong Relationship
Time Between Stop & Restart	Yes	✅ Root Cause

Root Cause Analysis Worksheet

Step	Observation
Problem	Engine does not restart after buying Vanilla Ice Cream
Initial Assumption	Vanilla Ice Cream causes engine failure
Data Collected	Fuel, passenger weight, travel distance, speed, purchase time, idle time, restart time
Pattern Found	Failures occur only when purchase duration is very short
Root Cause	Insufficient engine cooling leading to vapor lock
Corrective Action	Wait 2–3 minutes before restarting or redesign the fuel system
Preventive Action	Improve engine/fuel system to eliminate vapor lock during heat soak

Facilitator Questions

Ask participants:

Which variable initially appeared to be the cause?
Which variables can be eliminated from the investigation?
Which variable has the strongest relationship with engine failure?
Why is correlation different from causation?
If you were the engineer, what additional data would you collect before concluding?
What lessons does this case provide for managers solving business problems?

Key Learning Outcome

This exercise teaches one of the most important principles in Data Analytics, Decision Science, Lean Six Sigma, and Root Cause Analysis:

The obvious explanation is not always the correct explanation.

Managers often focus on the most visible factor—in this case, the ice cream flavor—but effective decision-makers investigate the underlying process. The real cause was the time elapsed between engine shutdown and restart, which exposed a heat-related vapor lock issue. This reinforces the importance of collecting comprehensive data, testing assumptions, and distinguishing correlation from true causation before making business decisions.

The Hidden Cause

The car had a newly designed engine.

After switching off the engine,

the fuel system became susceptible to vapor lock (fuel vapor forming in the lines due to heat).

When the customer returned too quickly after buying vanilla, the engine was still heat-soaked and would not restart immediately.

When buying other flavors, the extra time spent inside the store allowed the engine to cool just enough for normal restarting.

So the real relationship was:

Vanilla ❌

Time between engine shutdown and restart ✅

Decision Science Lessons

Lesson 1

Never laugh at unusual data.

Sometimes strange data reveals major design flaws.

Lesson 2

Correlation is not causation.

Ice cream flavor

didn't cause

the engine failure.

Time did.

Lesson 3

Observe the process.

The engineer didn't investigate ice cream.

He investigated

the customer's journey.

Lesson 4

The problem wasn't visible.

Visible

Vanilla

Hidden

Engine temperature

Lesson 5

The best leaders ask

"What else changed?"

instead of

"Who is wrong?"

How to Facilitate This in Your Training

Step 1

Ask participants:

If you were the Service Manager, how would you respond to this complaint?

Most answers:

Customer is joking.
Psychological issue.
Coincidence.
Ignore it.

Write all responses.

Step 2

Ask

"What data would you collect?"

Participants usually say:

Vehicle model
Mileage
Fuel
Weather
Driver

Then ask

"What are you NOT asking?"

Eventually someone mentions

"Time."

Step 3

Reveal the investigation.

Watch the "Aha!" moment.

Connect to Automobile Manufacturing

Example 1

Complaint

"Cars produced every Monday have more defects."

Wrong conclusion

Monday workers are careless.

Real issue

Machines remained idle over the weekend.

Lubrication wasn't effective.

Example 2

Complaint

Vehicles painted in White have more rework.

Wrong conclusion

White paint is defective.

Real issue

White paint highlights tiny surface imperfections that are less visible in darker colors.

Example 3

Complaint

Night shift has lower productivity.

Wrong conclusion

Night shift workers are inefficient.

Real issue

Material replenishment occurs during the night, interrupting production.

Example 4

Complaint

Dealer complaints increase after festivals.

Wrong conclusion

Dealers are dissatisfied.

Real issue

Transport delays due to holiday traffic.

Discussion Questions

Why did the engineer believe the customer instead of dismissing the complaint?
What assumptions did everyone initially make?
What was the hidden variable?
What additional data would you collect before reaching a conclusion?
Have you encountered a "Vanilla Ice Cream" problem in your own plant?

Key Learning Slide

Visible Problem

Vanilla ice cream
Car won't start

⬇️

Hidden Variable

Short purchase time
Heat soak
Vapor lock

⬇️

Root Cause

Engine design issue

⬇️

Leadership Principle

Extraordinary leaders investigate extraordinary complaints. They don't dismiss them because they sound impossible.

Trainer's Closing Message

"Data rarely lies—but our interpretation often does. Great managers see patterns. Great leaders question patterns. Exceptional leaders search for the hidden variable that explains the pattern."

This case is an excellent opener for a Data & Decision Science for Management program because it demonstrates, in a memorable way, that the first explanation is rarely the correct one. It reinforces critical thinking, systems thinking, hypothesis testing, and evidence-based decision making—exactly the mindset leaders need in a modern automobile manufacturing environment.

The Story: "Ford Wanted to Reduce Vehicle Weight"

During the development of a new Ford model (versions of this story are also associated with the Ford Taurus program), management set an aggressive target:

"Reduce the vehicle weight to improve fuel efficiency."

The engineering teams worked hard to:

Use thinner steel
Reduce component sizes
Eliminate unnecessary material
Optimize the chassis

After months of effort, the vehicle weight had reduced—but not nearly enough.

Management was disappointed.

Then Someone Asked a Different Question

Instead of asking:

"How do we make the car lighter?"

One engineer asked:

"What is making the car heavy?"

This subtle change shifted the team's thinking from optimizing parts to challenging assumptions.

The Hidden Discovery

The investigation showed that one of the biggest contributors to the increased vehicle weight wasn't the chassis or engine.

It was the large number of optional features and accessories that had accumulated over successive model years:

Larger seats
Additional sound insulation
Bigger entertainment systems
Extra wiring harnesses
Multiple brackets
Reinforcements added after previous issues
Luxury options becoming standard

Each engineering team had added "just a little more" over the years.

Individually, each addition seemed insignificant.

Collectively, they had added tens of kilograms.

Leadership Lesson

The problem wasn't:

The car is too heavy.

The real problem was:

Nobody questioned why weight kept increasing in the first place.

Training Activity

Ask Participants

Your plant's manufacturing cost has increased by 15%.

What would you do?

Most answers:

Negotiate with suppliers
Reduce manpower
Improve productivity
Reduce waste

Now reveal additional information.

Every department had independently introduced:

One extra inspection
One additional approval
One extra report
One additional packaging layer
One more software license
One more meeting every week

Each decision was justified locally.

Collectively, they created enormous cost.

Automobile Manufacturing Parallel

Case: Assembly Line Cycle Time

Cycle time increased from 48 seconds to 58 seconds.

Production blames operators.

Industrial Engineering blames layout.

Maintenance blames equipment.

Now reveal:

Over the past five years:

14 new inspection points were added.
8 customer-specific checks became permanent.
6 paperwork requirements remained even after digitization.
4 optional features became standard.

Nobody ever removed anything.

The problem wasn't operator speed.

It was process weight.

Data & Decision Science Principle

This illustrates accumulation bias.

Organizations rarely fail because of one major decision.

They become inefficient because of hundreds of small decisions that are never reviewed.

Facilitation Questions

Ask your participants:

What processes in your department have become "heavier" over the years?
Which reports are still produced only because "we've always done it"?
Which approvals no longer add value?
Which KPIs are measured but never used?
If you could remove 20% of your process tomorrow, what would it be?

The Hidden Picture Exercise

Draw two circles on the board.

Circle 1 – Visible

Heavy vehicle
High cost
Long cycle time
Slow approvals
More inspections

Circle 2 – Invisible

Legacy decisions
Fear of removing controls
Departmental optimization
Scope creep
"Temporary" changes that became permanent

Ask:

Which circle should a leader spend more time investigating?

The answer is the second one.

A Powerful Closing Quote

Managers improve the process they inherit.

Leaders question whether the process should exist in its current form at all.

A note on historical accuracy

Unlike the Vanilla Ice Cream story (which has become a management legend) and the Abraham Wald survivorship bias case (which is historically documented), this "Ford weight" story exists in several versions and is often used as a leadership teaching case rather than a precisely documented historical event. For training, it's best to present it as an illustrative management case rather than claiming it happened exactly as described at Ford.

For your Maruti Suzuki Data & Decision Science workshop, this story is particularly effective because it introduces a powerful concept:

"Leaders don't just reduce weight—they identify what has been silently adding weight for years."

This concept applies equally to manufacturing processes, quality systems, approvals, reports, meetings, inventories, and organizational bureaucracy.

1. Toyota – "Why do we inspect quality?" → "Why don't we build quality?"

Traditional Thinking

Inspect products after manufacturing.

↓

Reject defective vehicles.

Toyota's Thinking

Don't inspect quality.

Build quality into the process.

This led to

Jidoka
Poka-Yoke
Andon
Stop-the-line culture

Leadership Lesson

Don't improve inspection.

Improve the process so inspection becomes less necessary.

2. Netflix – "How do we eliminate late fees?" → "Why have due dates at all?"

Blockbuster

Business depended on

Late fees.

Netflix

Questioned the assumption.

Instead of

Late fees

↓

Monthly subscription.

The entire industry changed.

Lesson

Sometimes the biggest opportunity is hidden inside the business model everyone accepts.

3. Southwest Airlines – "How do we provide better airline meals?" → "Why serve meals?"

Traditional airlines competed on

Food
Luxury
Multiple classes

Southwest asked

What does the customer actually value?

Answer

Affordable

On-time

Frequent flights

They removed

Meals
Multiple aircraft types
Complex seating

Result

Lower costs

Faster turnaround

Higher aircraft utilization.

Lesson

Competitive advantage often comes from removing complexity.

4. Amazon – "How do we sell more?" → "How do we help customers buy?"

Most retailers focused on

Selling.

Amazon focused on

Buying experience.

One-click ordering
Reviews
Recommendations
Fast delivery

The perspective shifted from

Company

↓

Customer

Lesson

Change the point of view.

5. Apple – "How do we build a better MP3 player?" → "How do people experience music?"

Others improved hardware.

Apple built

iPod
iTunes
Integrated ecosystem

The innovation wasn't only the device.

It was the entire experience.

Lesson

Customers buy experiences, not products.

6. McDonald's – Ray Kroc

Everyone thought

McDonald's sells hamburgers.

Ray Kroc believed

McDonald's sells

Consistency.

The product became secondary.

The process became the competitive advantage.

Lesson

Operational excellence beats occasional excellence.

7. Xerox

Problem

Customers called frequently for repairs.

Initial solution

Increase service engineers.

Instead

Xerox analyzed

Machine usage data.

They redesigned

Preventive maintenance.

Calls reduced dramatically.

Lesson

Fix causes, not consequences.

8. 3M – Post-it Notes

A scientist accidentally created

A weak adhesive.

Everyone thought

Failure.

Another employee asked

Where can weak glue become useful?

Result

Post-it Notes.

Lesson

Innovation often starts where others see defects.

9. Honda (US Market)

Conventional wisdom

Sell large motorcycles.

Reality

Small motorcycles became unexpectedly popular.

Honda shifted strategy.

Instead of forcing the original plan,

they followed customer behavior.

Lesson

Listen to the market, not your assumptions.

10. IKEA

Question

How do we reduce furniture cost?

Traditional answer

Cheaper materials.

IKEA asked

Why assemble furniture in the factory?

Customer assembles.

Transport costs reduced dramatically.

Lesson

Question every assumption.

11. Starbucks

Coffee wasn't the innovation.

Howard Schultz asked

Can a coffee shop become

A third place

between

Home

and

Office?

Result

People paid for

Experience

Community

Ambience

Lesson

Value isn't always in the product.

12. Uber

Traditional thinking

Increase taxis.

Uber asked

How do we use existing cars?

Lesson

Sometimes the resource already exists.

The business model doesn't.

13. Zara

Fashion companies asked

How do we forecast next year's trends?

Zara asked

How do we respond within two weeks?

Lesson

Speed can outperform prediction.

14. Intel

Intel asked

Should we continue making memory chips?

They realized Japanese firms had become dominant.

Andy Grove famously asked:

"If the board fired us and hired new management, what would they do?"

Answer:

Exit memory.

Focus on microprocessors.

This strategic pivot transformed Intel.

Lesson

Leaders must sometimes act as if they are outsiders looking at their own business.

15. Tesla

Traditional manufacturers

Sell cars.

Tesla built

Software-defined vehicles.

Over-the-air updates.

The car kept improving after purchase.

Lesson

Products can continue creating value after delivery.

16. Domino's Pizza

Instead of claiming

"Our pizza is the best,"

they admitted

"Our pizza needs improvement."

They redesigned the recipe and invited customers to judge the results.

Lesson

Transparency can build trust more effectively than perfection.

17. GE Aviation

Instead of selling engines,

GE increasingly sold

"Power by the Hour."

Customers paid for engine uptime and performance rather than simply buying equipment.

Lesson

Shift from selling products to delivering outcomes.

18. Michelin

Instead of selling more tires,

Michelin developed fleet management services to help customers extend tire life and reduce operating costs.

Paradoxically, helping customers use fewer tires strengthened long-term loyalty and profitability.

Lesson

Long-term customer success can be a stronger business strategy than maximizing short-term sales.

19. Adobe

Traditional model

Sell software once.

Adobe shifted to

Creative Cloud subscriptions.

Revenue became predictable.

Customers always had the latest version.

Lesson

Rethink how value is delivered, not just what is delivered.

20. The "Empty Chair" at Amazon

Jeff Bezos famously left an empty chair in meetings to represent the customer.

Every important discussion included the question:

"What would the customer say if they were sitting here?"

Lesson

The most important stakeholder is often the one not present in the room.

Facilitation Activity: "The Leadership Lens"

After each story, ask your participants to complete the following table:

Traditional Question	Leadership Question
How do we reduce defects?	Why do defects occur at all?
How do we reduce costs?	What activities create no customer value?
How do we increase production?	What limits production flow?
How do we hire more people?	Why is productivity low?
How do we improve inspection?	How do we eliminate the need for inspection?
How do we sell more vehicles?	How do we create more customer value?
How do we solve today's problem?	How do we prevent tomorrow's problem?

The Common Pattern

Every breakthrough began with a different question:

Toyota: "How do we prevent defects?" instead of "How do we find defects?"
Netflix: "Why have due dates?" instead of "How do we manage late returns?"
Amazon: "How do customers want to buy?" instead of "How do we sell?"
Intel: "If we were starting today, what business would we choose?"
IKEA: "Why must the factory assemble the furniture?"
Zara: "How do we respond faster?" instead of "How do we predict better?"

This is the essence of Decision Science for Leaders: the quality of an organization's decisions is often determined not by the answers it finds, but by the questions its leaders choose to ask.

7-Step Decision Science Framework.

The Decision Science Interpretation Framework

For every case study, reveal the information in this order:

Step 1 – Business Need

Why was the data originally collected?

↓

Step 2 – Raw Data Captured

What data was available?

↓

Step 3 – First-Level Assumption

What conclusion would most managers make?

↓

Step 4 – Hidden Question

What question did someone ask that nobody else asked?

↓

Step 5 – Deeper Analysis

What additional relationships or patterns were discovered?

↓

Step 6 – The "Seeing the Unseen" Moment

What invisible insight emerged?

↓

Step 7 – Decision & Business Impact

What changed because of this insight?

Below are the first five cases fully developed in that format. (The remaining cases can be developed in the same style into a training workbook.)

CASE STUDY 1 – TARGET: Predicting Pregnancy

Step 1 – Business Need

Target introduced loyalty cards to understand customer buying behaviour.

Purpose:

Billing
Promotions
Inventory Planning
Customer Loyalty

Nobody intended to predict pregnancies.

Step 2 – Raw Data Captured

Customer ID	Product	Date	Quantity
C101	Cotton Balls	Jan 2	2
C101	Unscented Lotion	Jan 5	1
C101	Calcium Tablets	Jan 8	1
C101	Magnesium	Jan 11	1

Millions of such transactions existed.

Step 3 – First-Level Assumption

Ask participants:

"What does this data tell you?"

Typical answers:

Customer preferences
Inventory requirements
Product popularity

Nobody says

Pregnancy.

Step 4 – Hidden Question

A data scientist asked:

"What life event causes this buying pattern?"

Not

"What products are selling?"

Step 5 – Deeper Analysis

They correlated

Thousands of pregnant customers

against

their buying history.

Patterns emerged.

Customers bought

Unscented lotion
Vitamins
Cotton products

during predictable stages.

Step 6 – Seeing the Unseen

The products were never important.

The sequence was.

Buying behaviour predicted

Life stage.

Step 7 – Business Decision

Target began personalized promotions

before competitors even knew customers were expecting.

Leadership Principle

Leaders don't analyse transactions.

They analyse life events hidden inside transactions.

CASE STUDY 2 – UPS: The Left Turn Problem

Step 1 – Business Need

GPS installed to

Track deliveries
Improve routes
Reduce fuel costs

Step 2 – Raw Data

GPS coordinates
Route maps
Fuel consumption
Delivery time
Driver behaviour

Step 3 – First-Level Assumption

Shortest route

Best route

Most participants agree.

Step 4 – Hidden Question

Data scientist asked

"What happens during left turns?"

Nobody had analysed turns separately.

Step 5 – Analysis

Every left turn created

Waiting
Idling
Fuel burn
Accident risk

Thousands daily.

Step 6 – Seeing the Unseen

Road distance wasn't expensive.

Waiting at intersections was.

Step 7 – Decision

UPS redesigned routes.

Millions saved.

Leadership Principle

Don't optimise distance.

Optimise interruptions.

CASE STUDY 3 – GOOGLE PROJECT OXYGEN

Step 1

Google collected

Performance reviews.

Purpose

Employee appraisal.

Step 2

Collected

Promotions
Salary
Attrition
Feedback
Manager ratings

Step 3

Assumption

Best managers

Best technical experts.

Step 4

Hidden Question

"What actually predicts high-performing teams?"

Step 5

Analysis

Compared

High-performing teams

Low-performing teams.

Step 6

Finding

Managers who coached,

listened,

removed barriers

outperformed technical experts.

Step 7

Google redesigned leadership training.

Leadership Principle

Measure behaviours,

not resumes.

CASE STUDY 4 – WALMART & HURRICANES

Step 1

Collected

POS sales.

Purpose

Inventory.

Step 2

Millions of bills.

Products sold.

Weather.

Dates.

Step 3

Managers predicted

Water

Bread

Milk

Medicine

Step 4

Data scientist asked

"What products increase together before hurricanes?"

Step 5

Association analysis.

Unexpected products

Beer

Pop-Tarts

Step 6

The unseen

Customers prepared emotionally,

not just practically.

Step 7

Stores stocked those items.

Sales increased.

Leadership Principle

Customer behaviour

is rarely logical.

It is contextual.

CASE STUDY 5 – ROLLS-ROYCE JET ENGINES

Step 1

Sensors installed.

Purpose

Maintenance.

Step 2

Collected

Temperature
Pressure
Oil
Vibration
RPM

Millions of readings.

Step 3

Traditional thinking

Machine fails.

↓

Repair.

Step 4

Hidden Question

"What changes before failure?"

Step 5

Data compared

Healthy engines

Failed engines.

Tiny vibration changes appeared

weeks earlier.

Step 6

Machines were communicating.

Nobody was listening.

Step 7

Predictive maintenance born.

Leadership Principle

Don't study failure.

Study the signals before failure.

Why This Framework Creates an "Aha!" Moment

Every case deliberately leads participants through a cognitive journey:

Stage	Participant's Thinking
Raw data	"This looks like ordinary operational data."
First assumption	"I know what this means."
Hidden question	"I hadn't thought of asking that."
Deeper analysis	"The same data tells a different story."
Seeing the unseen	"The real insight wasn't visible at first."
Decision	"A completely different business action emerges."

"The first interpretation is rarely the correct interpretation."

Since your audience will be senior managers from an automobile company, keep the humor clean, intelligent, and relatable. Every example should end with a Decision Science principle.

1. What the Wife Says vs What She Means 😄

Wife Says	First Interpretation	What She Actually Means
Do whatever you want.	I have complete freedom.	There is only one correct answer. Find it.
I'm fine.	Everything is okay.	You should know why I'm not fine.
It's okay.	Issue resolved.	The discussion has only begun.
Five more minutes.	Exactly 5 minutes.	Time is undefined.
I don't want anything.	No action needed.	You should have already understood what I wanted.
Go with your friends.	Permission granted.	Your priorities are under observation.
You never listen.	I missed one sentence.	I feel unheard over time.

Trainer's Punchline

Data (words) and information (meaning) are different.

Decision Science begins where literal interpretation ends.

2. What Managers Say vs What Employees Hear

Manager Says	Employee Interprets
We are one family.	Overtime may be expected.
This is a small task.	This will consume my entire day.
Can we discuss?	Something is wrong.
No pressure.	There is definitely pressure.
This won't take long.	Cancel your lunch plans.
We value work-life balance.	Unless there is a deadline.

Learning

Same sentence.

Different interpretation.

3. Company Email Translation 😂

Company Email	What Employees Think
We are restructuring.	Someone is leaving.
Exciting changes ahead.	More work is coming.
Cost optimization.	Budget cuts.
Lean organization.	Fewer people.
New reporting structure.	More approvals.
Mandatory fun activity.	It's no longer fun.

Discussion

Ask

"Which interpretation is based on data?

Which is based on experience?"

4. HR Translation

HR Says	Employee Thinks
We'll get back to you.	Probably rejected.
Competitive salary.	Less than expected.
Fast-paced environment.	Long working hours.
Learning opportunity.	You'll figure it out yourself.
Dynamic role.	Responsibilities keep changing.

5. Automobile Plant Translation

Plant Head Says	Production Team Thinks
Let's improve OEE.	More reports are coming.
We need higher quality.	More inspections.
Customer first.	Weekend work.
Safety is our priority.	Another audit.
Let's reduce inventory.	Material shortages ahead.

Now ask

"What if management actually meant process simplification?"

Participants laugh because assumptions drive behaviour.

6. Customer Feedback

Customer says

"The car is good."

Sales team celebrates.

Customer actually means

"It met my minimum expectation."

Customer says

"I'll think about it."

Salesperson thinks

Future sale.

Reality

Usually no decision yet—or no purchase.

7. Data Scientist vs Manager

Manager

"The machine failed because the operator was careless."

Data Scientist

"What evidence supports that?"

Manager

"Everyone knows."

Data Scientist

"That's an opinion."

Entire class laughs.

8. Traffic Police Example

Traffic police stop

20 cars.

15 are motorcycles.

Question

Are motorcyclists breaking more rules?

Most answer

Yes.

Reveal

90% of vehicles passing were motorcycles.

Lesson

Percentages matter.

Absolute numbers deceive.

9. Restaurant Example

Restaurant owner

"No complaints today."

Manager

Excellent.

Waiter

Sir...

Nobody came today.

Everyone laughs.

Lesson

No data

≠

Good data.

10. School Example

Principal

100%

Pass result.

Parents clap.

Reveal

Only top students appeared.

Weak students weren't allowed to write.

Lesson

Selection bias.

11. Company KPI

CEO

Accidents

↓

Zero.

Excellent.

Safety Head

Near-miss reporting

↓

Zero too.

Question

Excellent?

Actually

Employees stopped reporting.

12. Manufacturing Meeting

Production

Maintenance caused delay.

Maintenance

Production overloaded machine.

Quality

Both ignored standards.

Nobody attended training.

Finance

Everyone exceeded budget.

CEO

"So...

Who made the vehicle?"

😂

13. WhatsApp Double Tick

Single Tick

Message sent.

Double Tick

Delivered.

Blue Tick

Read.

No Reply

Participants laugh.

Ask

"What conclusion do you make?"

Then ask

"What data are you missing?"

Maybe

Driving
In meeting
Battery dead
Sleeping
Network issue

Lesson

Data

≠

Meaning.

14. Doctor Example

Patient

"I have fever."

Doctor

Doesn't prescribe medicine immediately.

Instead asks

Since when?
Temperature?
Travel?
Other symptoms?

Lesson

Professionals ask questions before giving answers.

Managers should too.

15. Cricket Example (Perfect for India)

Player scored

100 runs.

Best player?

Reveal

Balls faced

198

Strike rate

Another player

45 runs

20 balls

Strike rate

225

Question

Who created more match impact?

Lesson

Context changes interpretation.

16. Factory Clock

Manager

Everyone came on time.

Attendance 100%.

Production

Output down.

Question

What happened?

Reveal

Workers came.

Machines didn't.

17. Elevator Problem (Classic)

Hotel receives complaints

"Elevator is too slow."

Management

Should we buy a faster lift?

Engineer installs mirrors.

Complaints reduce.

The lift speed never changed.

Lesson

Sometimes the problem is perception, not performance.

18. The Three Watch Problem

A man with one watch knows the time.

A man with two watches is never sure.

Ask participants:

Which manager has more dashboards?

Usually

The confused one.

Lesson

More data doesn't always produce better decisions.

Better questions do.

19. The Coffee Test ☕

Trainer asks:

"Who here drinks coffee every day?"

Many hands go up.

Then ask:

"Does coffee cause success?"

Everyone laughs.

Explain:

Many successful people drink coffee.

Many unsuccessful people also drink coffee.

The relationship is correlation, not causation.

Lesson

Never confuse association with cause.

20. The Banana Story (A Memorable Closing)

Show this sequence:

Monkey climbs ladder.

Monkey gets banana.

Monkey is rewarded.

Participants understand.

Now change it:

Manager increases meetings.

Reports increase.

Performance decreases.

Ask:

"What caused the problem?"

Was it the meetings?

Was it the lost production time?

Closing Line

In Decision Science, the first explanation is usually the easiest.

The best explanation is usually the one discovered after asking one more question.

A Great Energizer You Can Use Every Hour

I call it "Interpretation vs Reality."

Display one statement and ask the participants to vote on its meaning before revealing the context.

Examples:

Statement	Initial Assumption	Reality
"Customer complaints dropped by 80%."	Quality improved.	The complaint portal was down.
"Employee engagement is 95%."	Highly engaged workforce.	Only managers completed the survey.
"Machine utilization is 98%."	Excellent performance.	Downstream stations are starved because of batch scheduling.
"Production met the target."	Success.	Inventory increased because demand fell.
"Warranty claims reduced."	Better quality.	Fewer vehicles reached the age where failures typically occur.

These mini-cases take just 2–3 minutes, create laughter and discussion, and reinforce the workshop's central theme:

"Never confuse observation with interpretation. Great leaders don't just collect data—they challenge the story the data appears to tell."

This style of energizer works exceptionally well between technical sessions because it refreshes participants while strengthening the core mindset of Data & Decision Science.

Drishyam (Hindi / Malayalam / Tamil)

Drishyam (2015 film) - Wikipedia Drishyam 3 Worldwide Release Announced By Mohanlal On THIS Date Papanasam DVD (Destruction of sins) (Malaysia)

Decision Science Principle: Incomplete Data Creates Wrong Conclusions

Initial Assumption

Police believe

Vijay Salgaonkar murdered a boy.

All evidence appears to support this.

Hidden Data

The timeline has been manipulated.

The family's movements have been carefully planned.

Twist

The investigation is built on assumptions rather than verified chronology.

Leadership Lesson

The strongest evidence can still lead to the wrong conclusion if the timeline is incorrect.

Corporate Example

Production delay.

Everyone blames Maintenance.

Timeline later shows

Supplier delay occurred first.

Kahaani (Bollywood)

Kahaani - Wikipedia

Everyone assumes

Pregnant woman

↓

Victim.

Reality

She is the strategist controlling the investigation.

Lesson

Never underestimate

outliers.

PK (Bollywood)

Everyone assumes

PK is strange.

Reality

Society's assumptions are strange.

Lesson

Sometimes the outsider

sees reality better.

Decision Science

Challenge assumptions.

Lagaan

REVISIT – LAGAAN: ONCE UPON A TIME IN INDIA – meri maaa, CINEMAAA

Everyone assumes

British are unbeatable.

Bhuvan analyses

Strengths

Weaknesses

Resources

and creates

a completely different strategy.

Lesson

Winning isn't about

resources.

It's about

using available data differently.

Apollo 13

Amazon.com: Apollo 13 (4K UHD + Blu-ray + Digital) : Tom Hanks, Ed Harris, Bill Paxton, Kevin Bacon, Gary Sinise, Kathleen Quinlan, Mary Kate Schellhardt, Emily Ann Lloyd, Miko Hughes, Max Elliott

Everyone asks

How do we complete the mission?

NASA changes question

How do we bring them home alive?

Problem changes.

Solution changes.

Lesson

Correctly defining the problem

is half the solution.

Special 26 (Bollywood)

Special 26 Full Movie | Akshay Kumar | Kajal Aggarwal | Latest Bollywood Hindi Action Hd Movies

Everyone assumes

They are genuine CBI officers.

Reality

They are con artists.

Lesson

Authority

creates cognitive bias.

Dhoom 3

Police assume

Simple robbery.

Reality

Psychological condition

changes entire investigation.

Lesson

Behaviour needs context.

Trainers Treasure - All that makes a perfect trainer

Thursday, August 6, 2026