A mid-size CPG brand used AMC to compare SP vs SB on the same queries, quantify halo sales on non-promoted ASINs, and pinpoint funnel leakage at search term × ASIN level — enabling smarter budget shifts and clearer decision-making.
Mid-size CPG (Health & Household)
Protective gloves + nutrition supplements
SP + SB | 1 month AMC data
The brand ran Sponsored Products heavily, with Sponsored Brands “on the side.” Performance was judged mainly via console ROAS on promoted ASINs and last-touch conversion-time reporting.
The team suspected two things but couldn’t prove them:
Goal: replace “promoted-only ROAS” thinking with a portfolio-aware, funnel-aware operating model.
We built the AMC workspace around three analytical domains and ran three core modules.
SP vs SB by search term: spend, clicks, sales, ROAS, halo share.
Search term × tracked ASIN: DPV, purchases, DPV→purchase rate.
Lag-aware measurement to avoid cutting winners too early.
Clear evidence of underfunded SB, major halo contribution, and where the funnel really leaked.
A console-only view would systematically undervalue the real growth drivers.
Implication: SB is underfunded and halo is not optional. If you optimize promoted-only ROAS, you will cut campaigns that actually grow the portfolio.
On shared search terms, SB often beat SP by +1.5–4.0 ROAS points.
Strategic takeaway: stop the default “SP first, SB if budget allows.” Instead, shift budget on specific generic terms where SB clearly outperforms — and measure on total + halo sales.
In this dataset ATC was zero, so we used DPV → purchases as the funnel.
Key point: many “performance problems” were not keyword problems — they were PDP/offer problems that PPC alone can’t fix.
One promoted ASIN drove ~$1.2K sales with 100% halo share.
Some generic terms were “laggy” and under-reported in short conversion windows.
SP DPV/ATC click fields returned zero in the export.
We translated findings into a Monday-morning execution plan:
Result: a defensible, CMO-friendly narrative about where generic search truly works, where the funnel breaks, and where SB deserves budget over SP on the same query.