01 / The challenge
Start with the decision.
National TV and radio campaigns could not be understood through digital click data alone. Convertro, later acquired by AOL, needed models that connected broadcast activity, online behavior, and internal client data.
02 / What I built
Connect the pieces.
As the in-house television attribution data scientist, I built multi-touch attribution models from scratch and developed custom algorithms using curve fitting to model direct-response dynamics. I also brokered and managed a data partnership with a major cable distributor.
Broadcast + online + client dataCustom attribution modelsCampaign analysis
03 / What changed
A more useful answer.
The models supported measurement of more than 15 national TV and radio campaigns. I served as the primary client and agency interface, translating the modeling into findings and strategic recommendations.
Attribution requires both technical modeling and a clear account of what the evidence can support. The model is useful when buyers understand its assumptions and the decisions it can inform.