Commentary
The Rise Of The Humans
- by Joe Mandese @mp_joemandese, April 23, 2018
Maybe it’s because I’m utterly confused by the second season opener of “Westworld” Sunday night, but I’m thinking a lot about humans this morning, especially the notion that the more we become a machine-oriented as species, the more we need human beings to govern those machines to make sure they’re serving the right masters and optimizing the correct outputs.
But the real reason I’m thinking this way is a couple of news announcements about some human talent plays. Definitely comScore’s announcement this morning that it has hired digital wunderkind (and 360i founder) Bryan Wiener as its new CEO, not to mention adding digital agency pioneer (and former GroupM Chief Digital Officer) Rob Norman to its board.
But there was another announcement late last week -- which likely went more unnoticed -- that I’d like to highlight now, because it also speaks to the role humans need to play in an increasingly algorithmic world of data outputs.
On Friday, New York-based Truesight Consulting and London-based Marketscience Consulting announced they were pooling their human talent. While they are not yet calling it an actual merger, the “strategic partnership” partners two teams of the marketing world’s leading strategic thinkers who help big marketers solve the vexing problem of modeling systems.
Chances are, you haven’t heard of either firm, but they are run by some of the smartest minds on modeling, especially marketing mix, media mix, and econometric models that have become de rigueur for most modern-day marketing organizations.
Truesight was founded last year by two Madison Avenue vets -- David Dixon and Sebastian Shapiro -- who were running Publicis’ Ninah Consulting before striking out on their own. Prior to that, they were key players inside Interpublic units.
Marketscience Consulting was founded in 2012 by Peter Cain, a highly regarded marketing analytics expert who has been advising leading marketers in financial, consumer goods, retail, tech and pharma categories how to leverage marketing-mix models.
The partnership comes at a time of increased clutter and confusion in the modeling world, which has grown diverse and fragmented with the rise of data and digital platforms. Mix modeling, which gained prominence among big packaged goods marketers in the mid-1990s with the development of regression modes that could tie media inputs like TV advertising campaigns to incremental sales gains, has exploded with the availability of data and media options in the past 10 years into an array of so-called “attribution models” that seek to attribute results in more real-time.
A recent white paper created by Sequent Partners for CIMM (Coalition for Innovative Media Measurement) lists 25 best-in-breed modelers alone.
Many of those are positioned as turnkey solutions that come in and essentially outsource a marketing organization’s modeling needs based on that modeler’s solution. The Truesight/Marketscience partnership seeks to help marketers bring that process in-house -- for a number of reasons, including the fact that some of the data might be too sensitive to outsource, but also because the precise method of modeling might require nuances unique to that marketer.
Joe Mandese is the Editor in Chief of MediaPost. You can reach Joe at joe@mediapost.com.
Good article, Joe. I would add that despite the availability of so much more "data" these days, not all of it has a bearing on determining the ROI for many branding campaigns, as opposed to other types---notably DR. The reason for this is the fact that the orientation of current "data" sources is essentially far too granular---like trying to track what every individual consumer "saw" ad-wise and correlationg this with actual behaviour.
Your typical branding advertiser----and such companies spend over $200 billion annually on ads for their branding efforts in the U.S. per our latest "TV Dimensions Alert" estimates-----is still mainly concerned with attaining brand awareness and getting its positioning message across in the belief that this will drive overall sales gains once the advertising has done its job. Sadly, many of the modelling evaluations done for such advertisers lack crucial information---like awareness tracking relative to competitive brands, realistic audience tonnage data by key demos for all competing brands, information on the impact of non-advertising variables such as word-of-mouth , sales promotional activities, distribution, etc. Instead, many still rely on ad spending figures as a surrogate for audience and not much more than sales results. In other words, some of the new turnkey systems will find that the advertiser may welcome such an approaach but even so the crucial data that needs to be plugged in is simply not there. Accordingly, the new breed of modelers may have to develop ways----or paid services---- to fill this void as well as using what information is available.