Cold Food - Case Study: Diagnose and Reduce Cold Food Complaints

Question

## Round 1 - Data Science Case Study

## Problem

DoorDash receives a spike in customer complaints about food arriving cold. You are asked to diagnose the root cause and propose data-driven solutions.

**Data Available:**
- `deliveries(delivery_id, dasher_id, restaurant_id, pickup_time, delivery_time, food_prep_ready_at, distance_km, weather)`
- `ratings(delivery_id, overall_score, food_temp_score, comments)`
- `restaurants(restaurant_id, category, avg_prep_time_min)`

**Your analysis should cover:**

1. **Wait time at restaurant:** How long does the dasher wait after food is ready? Longer waits = colder food at pickup.
2. **Transit time:** Is delivery distance or traffic the dominant factor?
3. **Segmentation:** Are certain restaurant categories (e.g. sushi, pizza) more affected?
4. **Correlation:** Does `food_temp_score` correlate with `(delivery_time - food_prep_ready_at)`?

```
Key metric: "food exposure time" = delivery_time - food_prep_ready_at
Hypothesis: food_temp_score drops ~0.3 points per 5 extra minutes of exposure
```

## Follow-ups

1. How would you distinguish "cold at pickup" vs. "cooled during transit"?
2. What operational levers does DoorDash have to reduce cold food? Which would you prioritize?
3. How would you design an experiment to test whether insulated bags reduce cold complaints?
4. What metric would you track as a leading indicator before customers complain?

Cold Food - Case Study: Diagnose and Reduce Cold Food Complaints

Question Details

Full Details

Topics

More from DoorDash