Why it’s ‘impossible’ to solve the measurement issue
Marketers and publishers seeking a single source of truth in their analytics are ‘doomed’, Nunn Media’s managing director Chris Walton has argued.
Speaking at the final session of Mumbrella’s Publish conference, Walton was trying to answer the question of whether it’s time to start again with metrics.
Mal Dale, Chris Walton, Annabel Hodges, Michelle Levine and moderator Zoe Samios at the Publish conference
Walton said his view was the existing metrics, however flawed, are fine as practitioners understand their limitations.
This only matters because its used to settle price. If you don’t have a contract which defines the error bars, then your price variance against what you can measure, vs what Nielsen can measure is kind of your own problem.
it gets complicated with arbitrage. in web ads, via the ubiquitous company, I bid low cents on eyeballs for research reasons, and get huge eyeball counts. As it happens I got 3PAS and I know from independent tracking pixels what I see distinct from what google says I see. I understand the variances. I don’t have to like them but I can at least say “yes, I know why this differs” which isn’t that far from using VISA card absolute ground truth stats on purchasing to verify the market research question.
I guess if people wind up having to pay a third party to occasionally audit and check Nielsen figures, or if some industry accepted error bar is out there, its really ok. Because in the end, this is only useful because it helps set price. Mind you, if there is a huge upset and somebody calls it into question, the value proposition could dive to zero pretty quickly. Which is what PVR did to the original handbook method, after all.