I promised more regarding the future standard of response campaign tracking via multiple attribution so here is the next, albeit short, installment. This post focuses a bit on the devil’s advocate position on the arbitrary nature of current multiple attribution protocol reporting.
I had a chance to speak with Doubleclick last week, who reminded me that their “Exposure to Conversion” custom reports are in fact multiple attribution reports, which actually have been in-market longer than the Atlas Engagement Mapping reporting. The format is very different from that of Atlas, who developed a slick interface to go along with what is essentially a very complex process. Doubleclick provides the data in a more raw format to allow the agencies to manage the data sets manually. (using excel, access or a data mining or dashboard tool for example).
Ultimately Atlas and Doubleclick offer different means to reach similar multiple attribution reporting ends.
Of course, there is no “magic button” to make the multivariate process of attribution simpler. Whether a slick interface or raw data is provided, there are too many variables for any current solution to be deemed accurate. However, I still believe that any multiple attribution reporting has to be better than the last ad standard. The reality is that accuracy lies somewhere in between the two.
Weighting of criteria such as format, ad size and even recency, for example, are very subjective and lack the support of empirical data to direct the weighting. Over time the ad servers will be crunching aggregate level data to establish more standards in this regard. However, none of the current models take environment into account. In theory, relevancy of placement (another arbitrary and subjective variable) would have to be included in the equation somehow. For example – if a consumer is exposed to a rich media unit in a non-contextually relevant environment yesterday and a standard ad in a contextually relevant environment today, how would the inter-relationship between relevancy, format and recency effect influence? In other words, how would these variables be weighted in the multiple attribution protocol equation?
As you can see – there are no simple answers. One of the issues this poses is the level of adoption among an industry struggling to keep up with the data intensity needs of today, let alone those of tomorrow. To a degree, agencies and marketers are indeed seeking the “magic button” that doesn’t exist.
I still proclaim this to be part of the future standard, even if it takes a couple of more years to get there. I look forward to a bright future of multiple attribution reporting, not only for response campaigns, but also for brand based campaigns where multiple touch points create aggregate influence. We currently lack the necessary insight to prove where that influence is coming from. Roll over John Wanamaker!