LONDON — How many marketers are prepared to uncover the best ad-buying strategies, by stopping buying ads? That’s what Vodafone has done, and it claims to have boosted efficiency by 10%, leading to millions of pounds in spend being returned.
In this Beet.TV panel interview, DataXu CEO Mike Baker describes the project as: “A series of continuous experiments where media volumes and types of investments are continuously varied – perfectly randomized so that one can measure sales and understand which of these investments in which medium and territories has a causal relationship with the sale.”
So, what happened on the ground, and how did it work out? Vodafone head of brand strategy David Still says: “We take parts of the country, we differ our levels of spend across different media types and then we compare that to other areas where we’ve kept spend constant.”
Or, in plain English, as Still puts it: “If you want to prove something really works, the best way is to take it away.”
DataXu’s Baker says it’s a technique called “experimental design” that’s as old as the hills. And Still says – for figuring out where to put money in a fast-moving multi-channel universe – it works better than other performance measurements, which he sees as “fundamentally flawed”.
“Media performs in a completely different way from econometric models, last-attributable-click… anything we’ve seen before…,” he adds. “We’ve made some big quite big changes to the media we use and the amount of media.”
The pair say the strategy is usable and scaleable in categories and companies besides Vodafone’s telco segment.
They were interviewed by MTM London founding director Jon Watts.
This video was produced in London as part of our Addressable & Advanced TV Summit hosted by Sky Media and presented by FreeWheel and Invidi. Please visit this page for additional segments from the event.