Build dashboard; diagnose engagement–purchase gap
Overview
This question evaluates a data scientist's competency in product analytics and experimentation, focusing on dashboard specification, precise event- and user-level metric definitions, core dimensions, data freshness SLAs, and instrumentation within the Analytics & Experimentation domain.
Meta
Oct 13, 2025, 9:49 PM
Data Scientist
Onsite
Analytics & Experimentation
0
0
Build a Comprehensive Dashboard for the Shopping Tab (Organic Only)
Context
Assume the Shopping tab is an in-app surface for organic product discovery and purchase. Exclude paid ads and sponsored placements from all metrics. Purchases may occur via in-app checkout or via offsite merchant redirect with server/pixel attribution.
Task
-
Specify a comprehensive dashboard for the Shopping tab (organic only):
- Primary metrics (e.g., GMV, purchases, unique buyers, PDP CTR).
- Secondary metrics and guardrails (e.g., bounce rate, search exits, creator engagement).
- Define each metric precisely at both the event-level and user-level where relevant.
- List core dimensions (e.g., country, device, acquisition channel).
- Provide data freshness and SLAs.
-
Suppose engagement with the tab is high but purchase rate is low. Propose a diagnosis plan that includes:
- Segmentation (new vs. returning, search vs. feed entry, product category).
- Funnel drop-off analysis.
- Instrumentation checks.
- Experiments you would run to isolate UX vs. supply (inventory/price/trust) issues.
Solution
Login required
Show