Last-Click Attribution Models Give Marketers Incomplete Picture
First- and last-click attribution methods least effective at measuring multichannel influence
Marketers are used to having to prove digital advertising’s worth, but findings from Econsultancy and Google Analytics suggest the majority of marketers worldwide failed to use attribution models that properly depict the influence of each ad format and marketing channel on the consumer’s purchase path.
Instead, they justified and planned digital marketing investment using last-click attribution models.
The last-click attribution model was the most common method used by marketers and agencies worldwide. Under this model, the last click receives full credit for any revenue generated.
Some last-click models factor in both marketing and paid advertising influences, but under the paid search last-click model, only paid ad formats such as search and display are credited. Social media or organic search clicks do not receive credit, which makes it difficult for marketers to justify investment and spend on non-advertising marketing tactics or to know from where else they are driving purchases.
Lesser used methods of attribution—including first-click, the direct counter to last-click; linear, a model that assigns equal weighting to all interactions; and customized by channel, a type of linear model in which marketers assign their own custom value to each touchpoint—were used by a greater number of agencies than marketers.
First-click and last-click attribution models are easiest to measure, but their use can over- or under-credit an ad format’s influence on conversion activity. For instance, February 2012 data from Adobe measuring revenue per visitor to US websites, broken down by attribution model, showed that search generated 38% more revenue when measured via first-click attribution than last-click. Social’s first-click slant was even more dramatic: 88%.
Additional Q1 2012 findings from digital marketing firm RKG showed similar differences in measured revenue contribution from key online advertising and marketing touchpoints when moving away from a first-touch to a last-touch attribution model.
In addition to paid search’s decline in revenue when measured using last-click attribution, when factoring in other touchpoints, organic search and comparison shopping sites—which are often used to vet and compare products and services—also saw declines in average revenue per conversion. As the last-click model measures influences farther down the funnel, channels are credited with a changing amount of revenue. The differing results offer great evidence that a one-touchpoint experience is atypical for internet users and not a good measure of digital effectiveness.
Not surprisingly, Google Analytics and Econsultancy found both first- and last-click attribution models the least effective. Linear models and models that customize by channel were considered most effective by client-side marketers.
Moving away from a first- or last-click model, marketers must also look to incorporate metrics beyond standard click counts into their modeling mix to better reflect top-funnel touchpoints like display, where clicks on banner ads often account for less than 1% of all views, and viewthroughs are a more meaningful indicator.