A new predictive analytics model is providing a more accurate way to attribute which elements of a media mix yield results. Designed for Jeep by Toronto-based digital communications agency Organic, the new capability has already predicted the automaker’s 2008 retail sales within 1% of actual numbers, resulting in 15% savings for the brand, which, when implemented, returned $17 million in marketing value.
Jeep and Organic worked together to create a series of proprietary algorithms to predict business success based on multiple sales variables spanning data sources from web clicks to macroeconomic factors such as the value of the dollar, gasoline prices and inflation.
The new capability evaluates info from a broader set of inputs like demand history and media plans, and a wider set of sources (real-time, dealer, corporate, offline, online, environmental and batch feeds) into one consolidated data model to help optimize the marketing and media spend.
‘We are moving out of the era of ‘the last click wins,” Steve Kerho, VP analytics, media and marketing optimization at Organic tells MiC, ‘and toward an era of better visibility into each aspect of the media spend. Sadly, to date marketers haven’t had the insight needed to navigate these tough economic waters. This tool allows marketers to look out of the front windshield, versus the rear view mirror.’
Kerho points out that, given the new model’s ‘data-hungry’ format, a lot of different agencies and entities have to come together and share information for this to work.
Jeep and other Chrysler brands plan to implement Organic’s analytic model into upcoming strategic operations planning, as well as other clients.
