Orange Juice


Attribution Modeling


Written on 11/8/17 9:29 PM


Attribution Modeling – Holy Grail or …

hashtag_large_thumbAttribution. It’s the word on every marketer’s lips, and it’s impacting how media budgets are developed and optimized. But in today’s multi-channel environment, attribution “science” may not be telling the whole story. While attribution modeling provides increased visibility into a customer’s journey, models often tell only part of the story – on-line conversions and channels. If your data shows that a significant portion of conversions have long, multi-step paths – meaning customers are engaging across multiple channels – you may have an attribution gap.

Attribution – On-line-to-off-line

If you don’t understand how online channels impact on offline revenue, you’re only seeing part of the story. Only looking at digital conversions and revenue fails to take into account the effect on offline sales and in-store visits. Of course, one way to address this is with unique promotion codes by channel to fill the gaps in the story, which allows you to track conversion. Google’s “In-Store Visits” metric uses multiple signals to determine online/offline impact, including Google Maps data, WiFi, GPS and more to create in-store visit estimates, but this only applies to the million opted-in users.

Platforms that leverage mobile device data are also a step in the right direction, although they don’t track a specific user’s activity.

Attribution – Cross-device

The proliferation of devices has created another attribution challenge – measuring consumer behavior and engagement across devices and screens. The migration of search away from taps on the screen and toward voice-recognition tools will only make this more difficult. Although some companies can begin to make headway through multiple screen and device logins, it remains an area of opportunity.

Attribution modeling will only continue to improve as tools emerge to support a holistic approach to campaigns – across channels and across devices. Working with an attribution model that supports holistic marketing campaigns and efforts while avoiding simplistic “first- or last-click” approaches will help ensure multi-channel campaign optimization. Tracking the interplay of on- and off-line behavior and activity isn’t easy, but it can be addressed with unique promotional codes tying back to the individual channel. Regardless of attribution model, challenges and opportunities remain for improvement as we seek to understand consumer behavior and invest in the areas of greatest value.