Question 39
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?
Aggregating Viewing_Logs to customer–month level first, then joining to Subscriptions and Customer_Master on CustID.
Filtering Viewing_Logs to customers in a given acquisition channel using Customer_Master before computing Active_Months.
Computing Active_Months per customer from Viewing_Logs, then joining once to Subscriptions and Customer_Master on CustID.
Joining Viewing_Logs directly to Subscriptions at the row level and then summing Monthly_Fee × Active_Months without removing duplicate CustID–month combinations.