Data analytics was the big winner in the 2012 US Presidential race. In fact, 11:17 PM (US ET) November 6th was the moment data analytics went mainstream. This was when Ohio was officially projected to go to Obama. It was the ultimate validation for Nate Silver and his data analytics approach to election forecasting. To much fanfare he accurately predicted the results of the election in all 50 states without doing any of his own polling. He used sophisticated analytic models based on data from as many third party polls he could find. To this he added the secret sauce of data analytics – a keen understanding of how different types of data from different sources relate to one another in context.
- Standardizing records: Unifying the customer (voter) database
- Widening perspective: Combining diverse data types: demographics; buying/voting history; response by media; donation/activity by trigger (celebrity dinner), model (contest) and method (mobile); group/church membership, social networking activity (Reddit), etc.
- Judicious targeting: Carefully identifying the potential for influencing voters that could influence the election. Not worth targeting easily influenced voters if they don’t live in a county that can help swing a state. Not worth targeting difficult to influence voters even if they live in a critical county. This is essential for achieving impact and ROI.
- Media mix modeling: which media channels have the greatest impact on which kinds of voters?
- Action oriented outreach: Understanding the specifics of why and how certain people act and designing multiple outreach experiments (progressive offers, channel mix, social references, etc.) based on that.
- Openness to innovation: data driven models may point to approaches that are counter intuitive for some decision makers. They can seem risky and mysterious. They will not be right all the time. Controlled risk is part of the evolutionary process to effectiveness. Without a tolerance for experimentation however, you will not develop a data driven culture, you will in fact kill it.
- $200M EU lift based on a sophisticated solutions recommendation engine
- 45% more subscription revenue with no increase in a multi-million dollar marketing budget
- Tens of millions of dollars in revenue uplift from simple web behavioral changes