Most of the conversions are logged after their "last touch" with the display channel. However, we refer to our attribution through a campaign-wide lens, considering each touchpoint as a contributor to the conversion. Therefore, we consider our attribution as a "multi-touch" and “multi-channel” model, attributing a conversion to the entire collection of streaming impressions (Connected TV, digital audio, and display).

Because Connected TVs and Bluetooth devices are unclickable mediums, tracking conversions gets tricky. To combat this visibility gap, EMG uses the Household graph (which tethers all internet devices from one household) and display channel (a clickable medium) to deterministically track cross-channel conversion activities (website visits, offline dealership arrivals).

We track every arrival to every dealership, so long as the individual tracked has completed an impression from said dealership. Once an individual completes that dealership's ad, we can track them to whichever dealership they enter, whether it is to the campaign’s personal dealership, or another dealership. This data allows us to calculate proper market share penetration (via our GMS iQ).

EMG attribution is a constant pursuit of the truth. EMG data scientists ensure dealership location records are exact via match back mapping and custom conversion zone generation, and conduct a routine data filtration procedure, requiring dwell times and recency standards to remove potential employees, delivery drivers, drive bys, pick ups, and street traffic interference. Unlike many other solutions which can only approximate an arrival, EMG’s one-to-one attribution puts us in a position to perform optimizations on our models to drive superior performance and report deterministic arrivals.