Business Data Management End Term: 21 December 2025 (September 2025 term)

A beverage company wants to estimate the annual “Addressable Market” (in ₹) for a new premium coffee brand in India. They purchase a national survey with 10,000 households and the following summary:
● 70% of households consume coffee at least once a week. ● 40% of coffee-drinking households have monthly household income ≥ ₹80,000. ● Among those higher-income coffee-drinking households, the average is: ● 20 cups of coffee per month per household, ● Expected willingness to pay of ₹50 per cup for premium coffee.
Assume this survey is nationally representative of 300 million households in India.
Which is the BEST first step to avoid overestimating the “Addressable Market” for this premium brand?

A "Left Join" between Table A (100 rows) and Table B (100 rows) results in a table with 105 rows. What does this definitely prove?

You are analyzing market share data for the smartphone industry.
● Last year, total market volume was 100 million units; this year it is 105 million units. ● Last year, Company X sold 20 million units; this year its sales volume remained 20 million units.
Which statement best describes Company X’s performance in terms of market share?

A B2B SaaS company tracks leads through this funnel:
● Stage 1: MQL (Marketing Qualified Lead) ● Stage 2: SQL (Sales Qualified Lead) ● Stage 3: Demo Completed ● Stage 4: Closed Won
You have two tables: ● Leads(LeadID, CreatedDate, Acquisition_Channel, Region) ● Opportunities(LeadID, Stage, Stage_Change_Date, Deal_Value)
You are asked: “Which acquisition channel has the highest conversion rate from Demo Completed to Closed Won in the last 12 months?”
An analyst writes the following query:
● Filters Opportunities to rows where Stage IN ('Demo Completed','Closed Won') in the last 12 months. ● Groups by Acquisition_Channel. ● Computes: Conversion_Rate = (COUNT(DISTINCT LeadID WHERE Stage='Closed Won')) / (COUNT(DISTINCT LeadID WHERE Stage='Demo Completed')).
Which subtle issue is most likely to bias this conversion rate upward for some channels?

A business analyst is studying how “Macro-economic indicators” relate to the sales of luxury cars. Over the last 3 recessions:
● Each recession began in Q4 of Year 3, Q2 of Year 7, and Q1 of Year 12 (as dated by the central bank). ● Luxury car sales started declining sharply in Q2 of Year 3, Q4 of Year 6, and Q3 of Year 11. ● Unemployment rates peaked in Q2 of Year 4, Q1 of Year 8, and Q3 of Year 13.
In this context, luxury car sales are best classified as:

You are visualizing a funnel for an e‐commerce site:
● Visits → Add to Cart → Checkout → Payment → Order Confirmed
You observe the following over the past week: ● 10,000 users reached Checkout. ● 5,000 users reached Payment. ● 4,800 users reached Order Confirmed. ● Error logs show <0.5% payment gateway errors. ● Session replays show many users dropping off on the Payment page before submitting their card details.
What is the most reasonable data‐driven hypothesis to investigate first?

You are given a 100% stacked bar chart comparing market share of 5 competitors over 3 years. Each year’s bar is normalized so that the total height is 100%, and only percentage splits are shown. What key business insight cannot be inferred from this chart alone?

A business analyst is studying how “Macro-economic indicators” relate to the sales of luxury cars. Over the last 3 recessions:
● Each recession began in Q4 of Year 3, Q2 of Year 7, and Q1 of Year 12 (as dated by the central bank). ● Luxury car sales started declining sharply in Q2 of Year 3, Q4 of Year 6, and Q3 of Year 11. ● Unemployment rates peaked in Q2 of Year 4, Q1 of Year 8, and Q3 of Year 13.
In this context, luxury car sales are best classified as:

An external agency provides you with data on both Customer Lifetime Value (CLV) and Share of Wallet (SOW) for your retail clients. Customer R shows high CLV (top 5% of your portfolio), but only 20% SOW in the relevant product category. A recent market survey finds the average SOW for your top CLV segment is 55%. (Share of wallet is the percentage of a customer's total spending in a specific product or service category that goes to a particular company.)
What is the most strategic action to maximize future revenue from this customer?

You are analyzing market share data for the smartphone industry.
● Last year, total market volume was 100 million units; this year it is 105 million units. ● Last year, Company X sold 20 million units; this year its sales volume remained 20 million units.
Which statement best describes Company X’s performance in terms of market share?

You are auditing a “Sales Transaction” dataset. For true Beverage items, the “Product Category” column contains:
● “Bev.” in 40% of rows ● “Beverages” in 30% ● “Drinks” in 20% ● “Soft Drinks” in 5% ● Blank in 5%
Your BI dashboard defines “Beverage revenue” using: ● A filter: Product Category IN ("Beverages", "Drinks")
Last month, the dashboard reported ₹80 crore as “Beverage revenue.” You run a corrected query that groups all 4 labels together as “Beverage” and find actual Beverage revenue is ₹120 crore.
Which data quality dimension is the root cause of this ₹40 crore under-reporting?

You have the following Tables:
● Customer_Master(CustID) with 10,000 active customers. ● Transactions(CustID, TxnDate) with all purchase records.
You want a list of customers who have been active in the past (at least one transaction ever) but have made no purchases in the last 6 months. Conceptually, which operation on sets best describes this logic?
Let: ● A = set of customers with at least one transaction in the last 6 months. ● B = set of customers with at least one transaction more than 6 months ago.

A retail manager wants to analyze Daily Inventory Turnover for SKU A to identify days with possible stockout risk.
The IT team provides this dataset for the last month (30 days): ● Total units sold in the month: 3,000 ● Stock level on Day 1: 400 units ● Stock level on Day 30: 400 units ● No data on stock levels or sales by day in between
The manager suggests estimating Daily Inventory Turnover by: Daily Turnover = (3,000/30) / ((400+400)/2)
Why is this approach fundamentally flawed for detecting stockout risk?

A beverage company wants to estimate the annual “Addressable Market” (in ₹) for a new premium coffee brand in India. They purchase a national survey with 10,000 households and the following summary:
● 70% of households consume coffee at least once a week. ● 40% of coffee-drinking households have monthly household income ≥ ₹80,000. ● Among those higher-income coffee-drinking households, the average is: ● 20 cups of coffee per month per household, ● Expected willingness to pay of ₹50 per cup for premium coffee.
Assume this survey is nationally representative of 300 million households in India.
Which is the BEST first step to avoid overestimating the “Addressable Market” for this premium brand?

In a tidy retail dataset, you track weekly sales for 5 stores over 52 weeks. You want a structure that:
● Allows you to easily compute total sales per store, ● And also average sales per week across stores.
Which layout is most consistent with tidy‐data principles?

You are presenting to the CEO. She asks: “Which region contributed the most to our profit growth this year compared to last year?”
You have the following data for each region: ● Previous Year Profit ● Current Year Profit ● Profit Change (Current − Previous)
Which visualization is most effective to clearly answer her question?

You have the following Tables:
● Customer_Master(CustID) with 10,000 active customers. ● Transactions(CustID, TxnDate) with all purchase records.
You want a list of customers who have been active in the past (at least one transaction ever) but have made no purchases in the last 6 months. Conceptually, which operation on sets best describes this logic?
Let: ● A = set of customers with at least one transaction in the last 6 months. ● B = set of customers with at least one transaction more than 6 months ago.

In a tidy retail dataset, you track weekly sales for 5 stores over 52 weeks. You want a structure that:
● Allows you to easily compute total sales per store, ● And also average sales per week across stores.
Which layout is most consistent with tidy‐data principles?

A SKU (Stock Keeping Unit) has a "Long Tail" sales distribution. In an ABC analysis, this SKU would likely fall into Category C. What is the primary risk of managing this data using simple "Average Demand"?

You are auditing a “Sales Transaction” dataset. For true Beverage items, the “Product Category” column contains:
● “Bev.” in 40% of rows ● “Beverages” in 30% ● “Drinks” in 20% ● “Soft Drinks” in 5% ● Blank in 5%
Your BI dashboard defines “Beverage revenue” using: ● A filter: Product Category IN ("Beverages", "Drinks")
Last month, the dashboard reported ₹80 crore as “Beverage revenue.” You run a corrected query that groups all 4 labels together as “Beverage” and find actual Beverage revenue is ₹120 crore.
Which data quality dimension is the root cause of this ₹40 crore under-reporting?

You are presenting to the CEO. She asks: “Which region contributed the most to our profit growth this year compared to last year?
You have the following data for each region:
● Previous Year Profit ● Current Year Profit ● Profit Change (Current − Previous)
Which visualization is most effective to clearly answer her question?

You are analyzing "Basket Co-occurrence" (Market Basket Analysis). You find that [Bread, Milk] has high lift, but [Bread, Butter] has low lift. However, the raw sales volume of [Bread, Butter] is higher than [Bread, Milk]. What does this imply?

In a Pivot Table, you drag "Region" to Rows, "Year" to Columns, and "Sales" to Values. You then drag "Sales" into the Values area a second time. A business analyst sets:
● The first "Sales" field to “Sum” (normal). ● The second "Sales" field to “% of Row Total”.
What is the main reason for doing this?

A chain of gyms analyzes member data and finds:
● Strong positive correlation between “Number of Personal Training (PT) Sessions purchased per month” and “Likelihood of renewing annual membership”. ● Members who buy ≥4 PT sessions per month have a 90% renewal rate. ● Members who buy 0 PT sessions per month have a 50% renewal rate.
The sales director concludes: “If we push every member to buy at least 4 PT sessions per month, our renewal rate will jump to 90%.”
What is the most important concern you should raise about this conclusion?

You are visualizing a funnel for an e‐commerce site:
● Visits → Add to Cart → Checkout → Payment → Order Confirmed
You observe the following over the past week: ● 10,000 users reached Checkout. ● 5,000 users reached Payment. ● 4,800 users reached Order Confirmed. ● Error logs show <0.5% payment gateway errors. ● Session replays show many users dropping off on the Payment page before submitting their card details.
What is the most reasonable data‐driven hypothesis to investigate first?

You are analyzing seasonality for an Ice Cream brand. December sales (in 1,000 units) over the last 5 years were:
Year 1: 80 Year 2: 120 Year 3: 60 Year 4: 140 Year 5: 100
You calculate the average December sales as 100,000 units and share this with the production team. They say: “This doesn’t help us decide how much risk we have of over or under‐producing in December.”
What other metric would satisfy them? Answers Case Sensitive : No

A small shop owner keeps data in a notebook. For January he writes entries like:
● "15th Jan: Sold 10 kg Rice, 5 kg Dal. Purchase: 50 kg Rice." ● "20th Jan: Sold 5 kg Rice. Purchase: 20 kg Rice."
You are asked to digitize this to calculate the current stock of Rice on 31st Jan. The notebook is missing one critical piece of information that makes accurate historical stock reconstruction difficult.
Which data attribute is essential but missing?

An external agency provides you with data on both Customer Lifetime Value (CLV) and Share of Wallet (SOW) for your retail clients. Customer R shows high CLV (top 5% of your portfolio), but only 20% SOW in the relevant product category. A recent market survey finds the average SOW for your top CLV segment is 55%. (Share of wallet is the percentage of a customer's total spending in a specific product or service category that goes to a particular company.)
What is the most strategic action to maximize future revenue from this customer?

You are analyzing "Basket Co-occurrence" (Market Basket Analysis). You find that [Bread, Milk] has high lift, but [Bread, Butter] has low lift. However, the raw sales volume of [Bread, Butter] is higher than [Bread, Milk]. What does this imply?

A SKU (Stock Keeping Unit) has a "Long Tail" sales distribution. In an ABC analysis, this SKU would likely fall into Category C. What is the primary risk of managing this data using simple "Average Demand"?

In a "Churn Prediction" model, you include a variable "Customer calls to support line". You find a positive correlation (More calls = Higher Churn). However, for a subset of high-value customers, More calls = Lower Churn. This phenomenon is known as:

A chain of clinics measures “Doctor Utilization Rate”:
Utilization = Total Consultation Minutes / Available Working Minutes
Data for last quarter: ● Overall across all clinics: Utilization = 80%. ● You then segment by Clinic_Type: ● Urban clinics: Utilization = 78% ● Rural clinics: Utilization = 76%
A senior manager concludes: “Segmentation doesn’t change the picture; utilization is around 77–80% everywhere.”
You dig deeper and find: ● Urban clinics account for 90% of Available Working Minutes. ● Rural clinics account for 10%. ● Within Urban, junior doctors average 60% utilization, senior doctors 95%. ● Within Rural, junior doctors average 85%, senior doctors 70%.
Which statement best describes the main analytical risk in the manager’s conclusion?

A "Left Join" between Table A (100 rows) and Table B (100 rows) results in a table with 105 rows. What does this definitely prove?

In a Pivot Table, you drag "Region" to Rows, "Year" to Columns, and "Sales" to Values. You then drag "Sales" into the Values area a second time. A business analyst sets:
● The first "Sales" field to “Sum” (normal). ● The second "Sales" field to “% of Row Total”.
What is the main reason for doing this?

Why is "Metadata" critical for Business Data Management?

A retail manager wants to analyze Daily Inventory Turnover for SKU A to identify days with possible stockout risk.
The IT team provides this dataset for the last month (30 days): ● Total units sold in the month: 3,000 ● Stock level on Day 1: 400 units ● Stock level on Day 30: 400 units ● No data on stock levels or sales by day in between
The manager suggests estimating Daily Inventory Turnover by: Daily Turnover = (3,000/30) / ((400+400)/2)
Why is this approach fundamentally flawed for detecting stockout risk?

You are analyzing seasonality for an Ice Cream brand. December sales (in 1,000 units) over the last 5 years were:
Year 1: 80 Year 2: 120 Year 3: 60 Year 4: 140 Year 5: 100
You calculate the average December sales as 100,000 units and share this with the production team. They say: “This doesn’t help us decide how much risk we have of over or under‐producing in December.”
What other metric would satisfy them? Answers Case Sensitive : No

A retail chain runs a new loyalty program in only some stores. After 3 months, you compare:
● Program stores: average bill size = ₹1,200 ● Non-program stores: average bill size = ₹900
The marketing head says: “The loyalty program increased bill size by ₹300.”
You dig into the setup and find: ● Program stores were all high-income urban locations. ● Non-program stores were mostly mid-income suburban locations. ● No random assignment was used.
What is the most accurate data-driven response?

A subscription video platform wants to understand “Customer Engagement Value” (CEV) by combining:
● Customer_Master(CustID, Region, Acquisition_Channel) ● Subscriptions(CustID, Plan_Type, Monthly_Fee) ● Viewing_Logs(CustID, Date, Minutes_Watched)
They define: ● Active_Months = number of distinct months a customer watched ≥ 30 minutes. ● CEV = Active_Months × Monthly_Fee.
You are asked: “Which acquisition channel is driving the highest average CEV per customer?”
Which of the following strategies is most likely to produce a systematically biased (overstated) estimate of average CEV per customer for a given acquisition channel?

A small shop owner keeps data in a notebook. For January he writes entries like:
● "15th Jan: Sold 10 kg Rice, 5 kg Dal. Purchase: 50 kg Rice." ● "20th Jan: Sold 5 kg Rice. Purchase: 20 kg Rice."
You are asked to digitize this to calculate the current stock of Rice on 31st Jan. The notebook is missing one critical piece of information that makes accurate historical stock reconstruction difficult.
Which data attribute is essential but missing?

A B2B SaaS company tracks leads through this funnel:
● Stage 1: MQL (Marketing Qualified Lead) ● Stage 2: SQL (Sales Qualified Lead) ● Stage 3: Demo Completed ● Stage 4: Closed Won
You have two tables: ● Leads(LeadID, CreatedDate, Acquisition_Channel, Region) ● Opportunities(LeadID, Stage, Stage_Change_Date, Deal_Value)
You are asked: “Which acquisition channel has the highest conversion rate from Demo Completed to Closed Won in the last 12 months?”
An analyst writes the following query: ● Filters Opportunities to rows where Stage IN ('Demo Completed','Closed Won') in the last 12 months. ● Groups by Acquisition_Channel. ● Computes: Conversion_Rate = (COUNT(DISTINCT LeadID WHERE Stage='Closed Won')) / (COUNT(DISTINCT LeadID WHERE Stage='Demo Completed')).
Which subtle issue is most likely to bias this conversion rate upward for some channels?

Why is "Metadata" critical for Business Data Management?

A chain of clinics measures “Doctor Utilization Rate”:
Utilization = Total Consultation Minutes / Available Working Minutes Data for last quarter:
● Overall across all clinics: Utilization = 80%. ● You then segment by Clinic_Type: ● Urban clinics: Utilization = 78% ● Rural clinics: Utilization = 76%
A senior manager concludes: “Segmentation doesn’t change the picture; utilization is around 77–80% everywhere.”
You dig deeper and find: ● Urban clinics account for 90% of Available Working Minutes. ● Rural clinics account for 10%. ● Within Urban, junior doctors average 60% utilization, senior doctors 95%. ● Within Rural, junior doctors average 85%, senior doctors 70%.
Which statement best describes the main analytical risk in the manager’s conclusion?

A subscription video platform wants to understand “Customer Engagement Value” (CEV) by combining:
● Customer_Master(CustID, Region, Acquisition_Channel) ● Subscriptions(CustID, Plan_Type, Monthly_Fee) ● Viewing_Logs(CustID, Date, Minutes_Watched)
They define: ● Active_Months = number of distinct months a customer watched ≥ 30 minutes. ● CEV = Active_Months × Monthly_Fee.
You are asked: “Which acquisition channel is driving the highest average CEV per customer?”
Which of the following strategies is most likely to produce a systematically biased (overstated) estimate of average CEV per customer for a given acquisition channel?

A CEO dashboard shows monthly revenue for the last 3 years. You notice:
● The y‐axis starts at ₹95 crore instead of 0. ● The actual revenue range is from ₹100 crore to ₹105 crore. ● The line chart visually appears to have huge swings (steep peaks and valleys).
The CEO says: “Our revenue is extremely volatile; we need an urgent cost‐cutting task force.”
As the data analyst, what is the most appropriate first response?

A CEO dashboard shows monthly revenue for the last 3 years. You notice:
● The y‐axis starts at ₹95 crore instead of 0. ● The actual revenue range is from ₹100 crore to ₹105 crore. ● The line chart visually appears to have huge swings (steep peaks and valleys).
The CEO says: “Our revenue is extremely volatile; we need an urgent cost‐cutting task force.”
As the data analyst, what is the most appropriate first response?

A retail chain runs a new loyalty program in only some stores. After 3 months, you compare:
● Program stores: average bill size = ₹1,200 ● Non-program stores: average bill size = ₹900
The marketing head says: “The loyalty program increased bill size by ₹300.”
You dig into the setup and find: ● Program stores were all high-income urban locations. ● Non-program stores were mostly mid-income suburban locations. ● No random assignment was used.
What is the most accurate data-driven response?

You are given a 100% stacked bar chart comparing market share of 5 competitors over 3 years. Each year’s bar is normalized so that the total height is 100%, and only percentage splits are shown. What key business insight cannot be inferred from this chart alone?

A chain of gyms analyzes member data and finds:
● Strong positive correlation between “Number of Personal Training (PT) Sessions purchased per month” and “Likelihood of renewing annual membership”. ● Members who buy ≥4 PT sessions per month have a 90% renewal rate. ● Members who buy 0 PT sessions per month have a 50% renewal rate.
The sales director concludes: “If we push every member to buy at least 4 PT sessions per month, our renewal rate will jump to 90%.”
What is the most important concern you should raise about this conclusion?

In a "Churn Prediction" model, you include a variable "Customer calls to support line". You find a positive correlation (More calls = Higher Churn). However, for a subset of high-value customers, More calls = Lower Churn. This phenomenon is known as: