Hi,

Not sure about BCG Gamma specifically, but here are some examples of the data science cases I've seen

• "What are the different metrics that will change for a ride-sharing company if it is raining?"
• "Digital bank has a decrease in the pass rate (Positive results in both Face ID and Document scan in the app for new users). What may be the root cause?"

The approach to this is still about structuring the problem, but it will be more like an issue tree:

1. Make the issue tree prioritizing the data analysis (e.g. completed rides = number of orders * completion rates; Completion depends on customer cancellations, driver cancellations and rejects (driver not found); and so on)
2. Narrow down to a branch with a problem (the one where you could see the change)
3. Build the hypothesis on what may have led the change (e.g. the pass rate for the new users was constantly decreasing since we changed the acquisition channel and acquired users of lower quality)

Best,