Katkat
Back

Investigate Instacart Revenue Decline Using Weekly Data

InstacartData Scientist · Technical Screen
MediumData ScientistTechnical ScreenAnalytics & Experimentation

Last updated

Instacart analytics prompt on investigating a 4% weekly revenue decline, covering time-series anomaly detection, seasonality, changepoints, prediction intervals, and richer-data decomposition by orders, AOV, geography, cohorts, and categories.

Scenario

Instacart’s weekly revenue fell 4 % versus the prior week and you only have the historical weekly-revenue time series.

Question

With only weekly revenue data, how would you investigate the 4 % decline? If you later gained access to richer data (orders, geography, AOV, etc.), what additional analyses would you perform?

Hints

Think seasonality, holiday effects, trend breaks, changepoint tests, segmentation and drill-downs.