The missing factor in attribution

A recent column on LinkedIn by Gary Angel of Ernst & Young has given a great overview of current methods for attributing conversions to digital marketing campaigns.

The headline he chose contained the provocative phrase “Why Sophisticated Modeling of Attribution is Mostly a Waste of Time”.

Likewise, a Vox article he linked to had the attention-grabbing headline of “Has Google sold billions of dollars in ads that don’t work?”

However, Betteridge’s law of headlines is clearly in effect here. By the end of the Vox article, we’ve been walked back to a statement that “If you’re a small startup looking to grow your customer base, … results suggest that search ads will work for you.” Not exactly the end of digital advertising as we know it!

Similarly, Angel of EY isn’t against attribution, algorithms, or models. His main point is that the current attribution algorithms don’t account for the fact that certain customers are going to convert, regardless of ad campaigns. In his words:

“[T]he first and most fundamental step is to understand what the default conversion likelihood is for any given type of customer. That’s the credit that isn’t up for grabs. Then build the model of which campaigns, in which order, and in which frequency drive incremental lift. To do this right, you’re very much in the world of algorithmic attribution – weighting touches to resolve credit.”

It’s not hard to imagine an advertiser doing market research to determine probabilities of conversion among various demographics, and then factoring those probabilities into their attribution models. So, if a certain type of customer is determined to be already 65% likely to purchase a product, then a campaign for that product can only receive 35% attribution at most for conversions from that type of customer.

Angel’s point is that by taking these propensities into account, advertisers won’t overestimate the value of digital campaigns. Otherwise, advertisers end up in a state of reminiscent of Conductor’s skit about addiction to paid media…