MIAMI \u2013 Put several media industry professionals on a panel and ask them about the challenges of television and video audience measurement and you\u2019re going to get lots of lively conversation. Along with some very blunt observations about just what\u2019s holding back the progress of \u201cadvanced TV.\u201d<\/p>\n
So it was at the recent Beet Retreat Miami 2017<\/a> when representatives of Tru Optik, Inscape, Nielsen Catalina Solutions and Team Arrow Partners gathered on stage with moderator Joanna O\u2019Connell<\/a>, VP and Principal Analyst at Forrester Research. While there was overall cordiality, there was no mincing of words when discussing the business and technical challenges of audience measurement.<\/p>\n They ranged from agency compensation that accentuates cheap CPM\u2019s over business outcomes to the speed with which campaigns can be analyzed for return on investment. After summarizing the four core ways to measure TV\u2014panel, set-top box data, automatic content recognition (ACR) data and in-app\u2014Tru Optik<\/a> CEO Andre Swanston offered these observations on device validation and audience de-duplification:<\/p>\n \u201cIf I spent a million dollars with Hulu and a million dollars with NBC and a million dollars with Fox and a million dollars with Crackle, that\u2019s great that I had 100 million impressions. How many households did I reach?\u201d De-duplicating across multiple publishers and platforms is \u201cnot anything super sexy or exciting, but it\u2019s some of the basic things that people come to expect across linear and digital.\u201d<\/p>\n Matt O\u2019Grady, CEO of Nielsen Catalina Solutions<\/a>, cited speed to market as a big challenge. \u201cCampaigns are running longer and longer, or somebody is always on in the marketplace, so our traditional methodology of test and control is just too slow quite honestly. So we\u2019re investing very heavily in new models that are based upon artificial intelligence and multiple models running simultaneously so that we can get faster results and so that we can get true in-flight reads that somewhat mimic what the end result is.\u201d<\/p>\n Then there\u2019s the issue of whose data are deemed to be the most useful, according to Jodie McAfee, SVP, Marketing & Business Development at Inscape<\/a>. \u201cWhat we hear a lot of is ‘we think that the legacy data sets and specifically Nielsen data is flawed.’ And then people will look at our data and they\u2019ll go ‘well this doesn\u2019t match up with Nielsen data.’ The same people that believe that those legacy data sets are flawed have business systems and operations historically built around those datasets that you literally would have to practically blow up the entire market just to get everybody to change.\u201d<\/p>\n