Marketing Analytics with Fexingo · 2026-07-17 · 11 min
Retail media networks like Amazon Ads, Walmart Connect, and Instacart are now a $50 billion-plus ad channel, but most attribution models completely misallocate credit for these placements. In this episode, Lucas and Luna break down why last-click and multi-touch attribution both fail when a retailer owns both the ad and the transaction data - creating a closed-loop illusion that overweights the final click while ignoring upper-funnel influence from social, search, and TV. They walk through a real example from a CPG brand running a holiday campaign across Instacart, Meta, and Google, and show how retailer-walled gardens create double-counting and blind spots. The hosts discuss a practical workaround: using incrementality testing with geo-based holdouts and matched-market experiments to measure true lift, rather than relying on platform-reported attribution. Episode 120 of Marketing Analytics with Fexingo.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.