• by Rahul Bafna , Op-Ed Contributor, October 1, 2015

Up until about 2010, when consumers relied on a single device to consume media and make digital purchases, it was fairly easy and made sense for marketers to use a last-click model for attributing conversion credit to specific channels. One could argue that there are still multiple touch-points on that single device that should each receive partial credit, and in fact many companies did.

The rise of multi-touch attribution vendors like Converto (AOL), Adometry (Google), MarketShare and VisualIQ shows that large marketers understand that last-click attribution is a poor and outdated model. These companies are able to determine the true contribution of each channel and assign proper credit, thereby helping marketers optimize their media mix.

For some of these companies, this goes beyond just digital to include TV, radio, and print as well.

But today, consumers have two, three or more digital devices, and use them interchangeably as they research and purchase products online. With mobile accounting for 60% of all digital retail engagement, but only 13% of transaction dollars, the purchase funnel is more fragmented than ever.

Modern marketers must consider multiple touch-points, channels, and devices in the path to purchase, and are reacting accordingly. This year, mobile ad spend will catch up to desktop, and will overtake it in 2016. So we’ve solved half the problem – reaching consumers wherever they are – but we still need to figure out the attribution side.

That’s not to say there hasn’t been development. As marketers advertise across channels and devices in order to keep up with consumers, their desire for multi-touch attribution has grown as well. Fortunately the technology providers are eager to fill this need and help marketers understand cross-channel and cross-device purchase paths.

In addition to those listed above, Facebook's Atlas rolled out cross-device conversion attribution in 2014, and Google's DCM is beta-testing it right now. The major ad servers realize that single-device, last-click attribution is no longer sufficient in a multi-device world. Even search marketing platforms like Kenshoo are licensing cross-device identity solutions to more accurately attribute conversions based on activity across devices.

The trick is finding the balance.

Obviously last-click is too simplistic, and single-device measurement is insufficient. Should we attribute credit evenly across every interaction? Should we time-decay the attribution, so that recent interactions are worth more than earlier interactions? Or should marketers be able to customize their own multi-touch, cross-channel attribution models?

The ultimate goal is to understand how consumers are influenced by ads as they move through the purchase funnel across devices, and it’s clear that more sophisticated attribution models are needed. The path to purchase is obviously long and complex, but it is beneficial for marketers to accurately understand the contribution and influence of each device and channel as they effectively optimize marketing mix.

It’s finally time to lay single-device, last-click attribution modeling to rest, and adopt multi-touch, cross-channel models.