One interesting class of data buyers and sellers that we have engaged with on the dmi.io journey has been participants in the political data world, who perform activities such as lobbying, political research, polling, political consulting, analysis and predictive modeling. With data now becoming a central part of any campaign strategy, I want to use this blog to share some of the learnings we have gained from the industry experts and leave you to think about whether your data has value to the campaigns.
Despite all the hoopla around the 2012 election and the success of the Obama campaign in harnessing big data, there is a long way to go until this process reaches maturity. In particular, the data is often insufficiently granular at the voting district level. Keep that in mind as you read on, as you may be pleasantly surprised at the ability of your data offering to play a role.
The challenge in an election campaign is simple to describe, but hard to achieve:
- Identify the potential voters who are most likely to be influenced by the campaign promises
- Communicate those promises to the potential voter
- Mobilize the voter to actually vote
Sounds easy! The challenge, of course, is how to do this at scale and within campaign budget. That’s where data comes in:
- Step 1 is essentially a customer (voter) segmentation exercise
- Step 2 is a channel segmentation exercise
- Step 3 is also a voter segmentation exercise but on a different axis
In this blog we are going to focus on the voter segmentation issue. Good voter segmentation is a non-trivial task and involves layering multiple segmentations to create a well defined target population. Let’s look at some of those segmentations and think about the data opportunities in them:
Geographic / Demographic – demographic profiles at voting district level provide a good starting point; much of the data is based on publicly available data such as the US census, although the data gets out of date pretty quickly.
Political affiliations – registered voters provide a valuable filter of individuals likely to be declared along party lines (so save your breath and your dollars!), although there are 22 states without party registration so there’s a data opportunity here for someone to help fill the gap. What are the second order sets of data that demonstrate party affiliation?
Issue affinity – people are strongly motivated by the issues that matter to them and identifying those affinities is a powerful segmentation tool. For example, the environment (global warming, impact of fracking more locally), healthcare, national security, and gun control are all hot button issues for certain segments of the population. Who are the groups most likely to have an affinity to those issues, and what’s the data that helps attach individuals to those issues?
Beliefs and values – An individual’s behavior is mostly strongly driven by their fundamental beliefs and values. These could be religious, social, or political and evidenced by the organizations they are associated with, the media sources they subscribe to, or the people they hang out with to name a few.
Consumption preferences – the types of houses people occupy, the cars they drive, the types of stores they frequent, their purchasing preferences – these are all valuable sources of data to build a richer profile of voters and test how they align with political leanings or issue affinity.
So with that picture in mind, the question I have for you is as follows: if you looked at your own data assets through the lens of voter segmentation, what are the ways you could slice and dice your data to make it valuable?
If that’s triggered a few ideas in your mind, drop me a note and we can talk more. Having identified the opportunity there’s then a process of getting that data to the right audience and that’s the role dmi.io can play for you.