The Role of Attribution Management's Descriptive Analytics

Volume 1, Issue 8 - September, 2011

Dr. Payman Sadegh, Director of Research & Development, Marketing Analytics, Visual IQ

The goal of marketing attribution is to trace the origins of marketing performance back to the activities that produced that performance. With the explosive growth of data and the availability of advanced analytics tools, attribution management has become an increasingly scientific and objective discipline rather than a subjective art.

Advanced analytics can help address several key questions:

  1. Predictive: How will marketing spends across multiple channels affect performance toward your marketing goals?
  2. Prescriptive: What is your most efficient portfolio of marketing spends?
  3. Descriptive: How have past spends on various channels contributed to your marketing performance across multiple channels?

The latter – descriptive analytics – offers several key insights. First, it maps the quantitative output of cross channel models into qualitative relations that can be easily understood and acted upon by marketers. Second it allows discovery of cross channel interactions, i.e. how channels cannibalize and drive the performance of each other. Finally, it offers marketers a diagnostics tool by which they can track changes in channel performance and take action as needed.

For Example

Descriptive analysis of a cross channel attribution model developed for one marketer reveals several interesting insights about how channels interact. A cursory read of the data shows a relatively low return on ad spend for display advertising while showing an extremely high ROI for email marketing. Featuring high media cost, the marketer uses display primarily for retargeting previous website visitors. On the other hand, the insignificant per volume cost of email allows the marketer to reach millions of customers and prospects at very little cost. A deeper analysis of interactions provided by a cross channel analytical models reveals the following about this organization’s marketing portfolio:

  • Display drives sales that convert through other channels – most notably, organic search. In fact, the largest revenue from display is realized indirectly through other channels, at least tripling the effective (true) ROI of this channel compared to the “last click” or “last act” attribution model.
  • Despite the appearance of being inexpensive, the largest cost of email marketing is not realized through its volume, but by way of the discounts offered to prospects reached through this channel. Including the effective cost of sales attributed to this channel reduces its observed ROI by several orders of magnitude.

Qualitative + Quantitative Input

Analyses like this are critical to making sound marketing decisions even when predictive or prescriptive models are also available. This is because mathematical models may not incorporate constraints in the decision process as effectively as expert human judgment can. A proper understanding of channel interactions as measured by true performance metrics allows the combination of qualitative judgment with quantitative rigor provided by mathematical models to balance the art and science of today’s marketing.

Subscribe to the IQ Advisor

Click here to subscribe to The IQ Advisor monthly e-newsletter.