In this episode of the Ecommerce Coffee Break Podcast, host Claus Lauter interviews Feifan Wang about the five biggest data challenges plaguing growing e-commerce brands. Feifan draws from his background at Google and as VP of Product & Analytics at Resident Home (which scaled from $7M to $100M+ in one year) to explain why data isn’t just about reporting—it’s about survival at scale.
They discuss the evolution of data needs as a business matures: early-stage brands can survive on directional insights from Google Analytics, but as complexity grows (multiple sales channels, international markets, high ad spend), the margin for error shrinks. Feifan breaks down SourceMedium’s approach to data quality, emphasizing that true accuracy requires handling the messy reality of e-commerce—like refunds, exchanges, and cross-border tax implications—that standard tools often miss.
Key Takeaways
- Data Maturity Curve: Data needs evolve with your business stage. What works at $5M (GA4 + spreadsheets) becomes a liability at $50M when 20% error margins translate to millions in lost efficiency.
- The “Horizontal” Trap: Tools like Supermetrics or Fivetran move raw data from A to B but leave you with the heavy lifting of transformation and normalization. SourceMedium provides a vertical-specific infrastructure that understands commerce logic out of the box.
- Data Quality is Multi-Dimensional: It’s not just accuracy; it’s breadth (integration coverage) and depth (richness of data). Reliable net sales figures must account for discounts, refunds, and exchange loops that platform dashboards often ignore.
- The Human Element: Even the best tech fails without data culture. Feifan emphasizes the importance of standardized taxonomies and consistent naming conventions to move organizations from “land of opinions” to “land of facts.”
- Proactive Onboarding: SourceMedium’s onboarding isn’t just technical setup; it’s a data hygiene audit that catches issues like inconsistent UTMs or untracked landing pages before they corrupt your insights.
- Build vs. Buy: Feifan advises brands to carefully consider the hidden costs of building internal data stacks—maintenance, scaling challenges, and the need for specialized engineering talent—versus buying a purpose-built solution.
