← Back to Home
E-COMMERCE DIAGNOSTICS • CASE STUDY
Product Performance & Conversion Funnel Analysis
1. Dataset, Scope & Business Scenario
Project Type
Personal Project / Simulated Business Case (Maven Fuzzy Factory)
Dataset Source
Maven Fuzzy Factory Teaching Dataset (6 CSVs: orders, items, refunds, products, sessions, pageviews)
Data Scale
32,313 orders • 472,871 sessions • 1,188,124 pageviews
Time Range & Currency
Mar 2012 – Mar 2015 (3 Years) • USD ($)
To investigate a drop-off gap between top-of-funnel site traffic and completed revenue, this analysis audits product-level conversion steps and refund risks across the browsing-to-purchase journey.
- Calculate Net Revenue After Refunds (
Gross Revenue - Refund Amount) across all catalog items. - Construct sequential session-based conversion funnels tracking unique sessions across milestones:
Catalog (/products)→Product Detail Page (PDP)→Cart (/cart)→Order Confirmation (/thank-you-for-your-order). - Compare conversion performance between flagship SKUs: The Original Mr. Fuzzy vs. The Forever Love Bear.
Denominator & Funnel Definitions (Verified in Code):
Each milestone counts deduplicated
website_session_ids (groupby.max()) progressing sequentially from /products (261,231 sessions):
Mr. Fuzzy (162,525 PDP → 69,957 Cart → 23,861 Order Confirmation, 34.1% Cart-to-Order) vs. Love Bear (26,033 PDP → 14,485 Cart → 4,803 Order Confirmation, 33.2% Cart-to-Order).
Sessions landing directly on a PDP without /products are outside this catalog-anchored funnel, and refunds are deducted separately in SKU Net Revenue rather than removed from clickstream progression.
2. Key Findings & The "Love Bear" Paradox
- The "Love Bear" Paradox:
- Top-of-Funnel (Catalog → PDP): Mr. Fuzzy captures 62.2% of catalog sessions, while Love Bear captures only 10.0%—indicating substantially lower catalog-page click-through exposure (placement hypothesis to test).
- Mid-Funnel (PDP → Cart): Once a user views Love Bear, it converts at a higher PDP-to-Cart rate (55.6%) than Mr. Fuzzy (43.0%), making it a high-intent candidate for merchandising visibility tests.
- Cart-to-Order Drop-off Bottleneck: Across both product lines, progression from
/cartto/thank-you-for-your-orderdrops by ~66% (~33–34% completion). - Refund Rate Anomaly: While Mr. Fuzzy is the primary net revenue driver, The Birthday Sugar Panda exhibits the highest refund rate in the catalog, flagging it as a priority candidate for product quality and supplier investigation rather than a proven supplier defect.
3. Technical QA & Engineering Refactoring
- Interactive Funnel Architecture: Migrated legacy Matplotlib horizontal bar plots to
plotly.graph_objects.Funnel, enabling interactive hover states and step-to-step drop-off ratios. - Typecasting Optimization: Replaced repeated datetime casting with a unified ingestion loop utilizing
pd.to_datetime(). - Dependency Deprecation Fixes: Modernized Seaborn plotting parameters (
hue='product_name',legend=False) to eliminate deprecation warnings and ensure compatibility with future releases.
4. Proposed Experiments & Recommendations
- Traffic Injection via Cross-Selling: Place Love Bear as a "Frequently Bought Together" widget on Mr. Fuzzy's high-traffic product page to test catalog-to-PDP lift.
- Checkout UX Overhaul: Audit hidden shipping costs and guest checkout friction to address the observed 66% cart-to-order abandonment rate.
- Automated Abandoned-Cart Recovery: Pilot automated email/SMS recovery reminders at 1 hour and 24 hours post-abandonment.
Validation Plan: All UX and recovery changes should be evaluated via A/B testing measuring Cart-to-Order Completion Rate, Net Revenue per Session, and Refund Rate (monitoring quality controls on higher-refund items such as Birthday Sugar Panda).