Guest Columnist Dean Barber: Big Data on the Campaign Trail
07 Jan, 2013
Editor’s Note: Dean Barber, principal of Barber Business Advisors, LLC,originally published the following blog excerpt on November 11, 2012 at Barberbiz. He was kind enough to let us share this extremely relevant post with our readers in regard to big data and how algorithms reined supreme in the 2012 presidential election.
We believe the content will give your company useful insights in regard to big data collection and analysis and what it can mean to your business.
Big Data on the Campaign Trail; One Toke Over the Line
Last week, I spent entirely too much time warning about the hazards of exploding consultants, eventually getting around to a far more important point that big data will increasingly determine how business will be conducted in the future.
Little did I know that big data would play such a huge role in the re-election of a president. I say that because, judging on the reports that I have been reading, Team Obama was plowing new ground, doing things that had never been done before in a political campaign in data collection and data interpretation.
This was far more than the selling of a president. Rather, this was war room science where algorithms reined supreme.
What is now apparent is that Mitt Romney ran a conventional if not a retro campaign, relying on media consultants and pollsters who wanted to fashion a brand that would hopefully move voters to the polls. The message was often a bit muddled, but the point is the Romney campaign employed tried and true tactics to unseat Barack Obama.
But the campaign to re-elect the president was a post-modern campaign that depended heavily on organized gathered data and computational formulas designed to identify and motivate those who might be inclined to contribute money and vote for the president. This was math and social science with a purpose.
Come Tuesday morning, the Romney team was understandably on edge, while journalists covering the campaign described the Obama camp as exhibiting a certain calmness, so confident were they as to the end result.
Now the purpose of this blog is to not to revisit the campaign rhetoric or the debate on which candidate might have the better choice. The woulda, shoulda, coulda is over. The American people have spoken. That train has left the station.
Rather, I am exploring how a political campaign embraced big data into a winning formula. Just how did they do that? The answers or rather the clues might provide businesses with valuable insight on just how analytics can be leveraged in providing for an edge over the competition.
And prevailing over the competition is the name of the game, in business, politics and war.
At the very start, Obama campaign Jim Messina had promised a totally different, metric-driven kind of campaign. “We are going to measure every single thing in this campaign,” he told Time magazine after taking the job.
Messina hired an analytics department five times as large as that of the 2008 operation, headed by a man who had crunched huge data sets to maximize the efficiency of supermarket sales promotions. Dozens of “scientists” conducted data-mining experiments, which were given mysterious code names such as Narwhal and Dreamcatcher.
According to an excellent Time magazine report, this massive data effort helped Obama raise $1 billion, remake the process of targeting television ads and created detailed models of swing-state voters that could be used to increase the effectiveness of everything from phone calls and door knocks to direct mailings and social media.
Running Tests and Developing Models
A new mega file was created by consolidating data bases, merging get-out-the-vote lists with fund-raising lists. This went far beyond finding potential voters and getting their attention. Rather, the data crunchers were running tests and developing models to predict which types of people would be persuaded by what kind of appeals.
About 75 percent of the determining factors were basics like age, sex, race, neighborhood and voting record. Consumer data also helped round out the voter profiles.
Data collection and analysis took center stage in a metric-driven e-mail campaign in which dozens of fundraising appeals went out each day. The analytics team built a detailed picture of voters in key states, allowing for deep dives into exactly where each demographic and regional group was trending at any given moment. They processed and reprocessed the data nightly to account for all imaginable scenarios.
“We ran the election 66,000 times every night,” a senior official with the Obama campaign told Time. “And every morning we got the spit-out — here are your chances of winning these states. And that is how we allocated resources.”
So what is the takeaway from all this? Businesses, too, are learning that knowing more — much more — about your customers will play an ever-increasing role in motivating them toward action. And that is where big data can make the difference.
A few weeks ago, an editor with a business newsletter from the Northwest asked me, among other things, about my perceptions of the business climate of his part of the country. I was a bit reticent to answer.
I told him that as a site selection consultant, I advise companies to go where they need to be based on many factors. I said it was absolutely conceivable that a client company should need to have operations in the Northwest for reasons of proximity to customers and suppliers.
But he pressed me on the business climate thing, so I finally relented. I said something like this: “Ok, look, I believe that you have some issues regarding permitting, especially environmental permitting, that can slow down a project to the point that it can take you out of the running. That’s not good.”
Naturally, when the story came out, the editor quoted some unnamed source saying that I was full of it. Fine. I can live with that. No big deal.
But now, lo and behold, voters in the states of Washington and Colorado, have essentially legalized the use of marijuana for recreational purposes. That’s right, recreational purposes, not medicinal.
Now I have to speculate, and admittedly it is speculation at this point, that such moves will not be helpful to the business climate in those two western states. I can foresee a huge conflict in the workplace, something that employers and any prospective employers do not want to face.
Will the Feds Act?
How and if the federal government will act is going to be key. I don’t think legalization for recreational purposes will or can stand. And for the record, I would not be offended if the feds stepped and in effect said, “No, you can’t do that. We got our own laws, too.”
There is a time and a place for states’ rights to be damned. States have proved to be fully capable of doing the wrong things, witness the Civil Rights movement in the South in the 1950s and 1960s, in which state governments openly embraced and upheld white supremacist views in restricting the voting rights of African Americans.
And while Washington and Colorado are not trying to restrict rights but do, in effect, just the opposite, I cannot envision how this move will be allowed to stand. Again, I have to think that this will play hell on the business climate in both states.
Simply put, if I am a manufacturer in Washington or Colorado, I am not too pleased with the notion of my products being potentially compromised with quality issues because I have stoners on the shop floor. I would want to monitor and measure not only product defects, but my health care costs and rates of absenteeism, all of which I would suspect would increase as a result of legalization.
At Issue is Productivity
Now it is likely true that legalization of marijuana may save law enforcement agencies substantial amounts of money in terms of the cost of enforcing drug laws. But it does nothing to help or contribute to advancing innovation and efficiencies on the shop floor of a production operation.
Ultimately, this is issue about productivity at the workplace, pure and simple. And while the notion of employers intruding into the private lives of their employees gives me great pause, if what employees are doing adversely affects their ability to perform their job, then it becomes a workplace issue.
Hey, I am a live and let live kind of guy. But I also think there is no issue here that drug abuse (and yes that absolutely includes alcohol) costs our economy billions of dollars a year in lost time, increased health costs, quality control, decreased production, a whole host of factors. And employers have the right to take reasonable measures to protect themselves.
In site selection, we try to the best of our ability to get a handle on the quality and the quantity of labor in any given place. People are the lifeblood a company and for a community, too.
Whenever possible, I want to measure the measurable, but I also want to get a feel for the place. This usually entails some gumshoe reporting (I was a newspaperman in my prior life) and behind-closed-door interviews with existing employers. If a rural county is covered up in meth labs, well, I want to know about it because it could pose a threat to my corporate client’s investment should this location be a place of interest.
Now how do you think most employers, particularly manufacturers in Washington and Colorado, feel about the now sanctioned use of marijuana for recreational purposes within the ranks of their employees? Do you really believe that it will not affect productivity on the shop floor? I can only imagine what they would be telling me and none of it is good.
Please understand that I am not saying that Washington or Colorado would automatically be stricken from consideration during the site selection process. That would be going to extremes.
But what I am saying is that I have questions. Plenty of them.
Dean Barber is the principal of Barber Business Advisors, LLC, a site selection and economic development consulting firm based in Plano, Texas. He can be reached at 972-767-9518 or at firstname.lastname@example.org. Please visit our website.