Attribution Analysis and Campaign Efficiency – Getting More Bang for your Buck

June 16, 2009
162 Views

fireworksHere’s some breaking news – GroupM has formally announced the Media Innovation Group as a separately-branded business unit.  It happened at the 2nd annual GroupM – 24/7 Real Media Digital Summit last week.  This is the team within GroupM that delivers innovative technology solutions for the advertising industry including ZAP and B3, and while the team has been around for a few years, this restructuring sends a clear signal that technology (and dare I say – terabyte-scale data analysis) is an increasingly important source of competitive advantage for WPP.

I delivered a collaborative presentation at the event with Brian Lesser, VP & General Manager of the Media Innovation Group, describing how Netezza’s data warehouse appliance enables their Zeus Advertising Platform (ZAP).  The presentation was fun, and I got a kick out of being on the same stage as Irwin Gotlieb (it’s hard to beat having “The King of Advertising” as your moniker).  Rob Norman’s presentation was pretty cool, too.  🙂

Attribution analysis is all the rage these days.  It attempts to solve the age-old advertising problem first articulated by John Wannamaker:  “Half the money I

fireworksHere’s some breaking news – GroupM has formally announced the Media Innovation Group as a separately-branded business unit.  It happened at the 2nd annual GroupM – 24/7 Real Media Digital Summit last week.  This is the team within GroupM that delivers innovative technology solutions for the advertising industry including ZAP and B3, and while the team has been around for a few years, this restructuring sends a clear signal that technology (and dare I say – terabyte-scale data analysis) is an increasingly important source of competitive advantage for WPP.

I delivered a collaborative presentation at the event with Brian Lesser, VP & General Manager of the Media Innovation Group, describing how Netezza’s data warehouse appliance enables their Zeus Advertising Platform (ZAP).  The presentation was fun, and I got a kick out of being on the same stage as Irwin Gotlieb (it’s hard to beat having “The King of Advertising” as your moniker).  Rob Norman’s presentation was pretty cool, too.  🙂

Attribution analysis is all the rage these days.  It attempts to solve the age-old advertising problem first articulated by John Wannamaker:  “Half the money I spend on advertising is wasted; the trouble is I don’t know which half.”  Solving this problem involves tracking and analyzing every exposure that every single consumer experiences across every relevant campaign generating those exposures.   The traditional attribution approach has involved simply giving the last click before conversion all the credit – but that’s just wrong.  In fact, the Atlas Institute reports in “The Long Road to Conversion: The Digital Purchase Funnel” that half of the total ad engagements with consumers who convert occur more than a week prior to actual conversion (pdf), and finds in “Engagement Mapping: A new measurement standard is emerging for advertisers” that the “last click” model ignores between 93% and 95% of audience engagements with online advertising that occur prior to conversion (pdf).  Attribution analysis enables a more accurate understanding of what leads to conversions, and these insights can lead to dramatic improvements in campaign lift and ROI.

The attribution analysis problem has two important characteristics that make it difficult to solve with anything other than the latest and greatest technology for terabyte-scale data analysis:

  1. The data volumes are very large – requiring tracking and analysis of every advertising exposure for every single user in every campaign.
  2. The ad hoc nature of the analysis means that it is often impossible to anticipate the types of questions that will be asked of the data – so traditional approaches to delivering fast query performance (e.g. indexing) won’t work very well, because those approaches require advance knowledge of exactly how the data will be queried.

So speed-of-thought analysis of terabyte-scale data sets is once again the business challenge du jour…and the media industry is once again getting rocked by the tech geeks.  🙂  While consensus standards on attribution modeling won’t arrive for a few more years – every advertiser, publisher, and agency needs a credible solution for quantifying attribution beyond the “last click” – right now.

And importantly, solving this problem is proven to create measurable business value.  There is some compelling content explaining Microsoft’s perspective on “Engagement Mapping” here.  And Havas has also published a nice paper on “Attribution Weighting” (pdf).  But my favorite concrete example is cited in the hot-off-the-press Netezza case study on “Conversion Attribution Modeling Across the Purchase Funnel” (pdf) – in which one large banking client was able to maintain a 25% conversion rate with 33% lower CPA on their campaign due to the attribution analysis delivered by the Netezza-powered Zeus Advertising Platform.

Photo credit:  Joe Penniston