back to notes

Embracing a new playbook: how data-driven decisions can change the conversation in local races

January-February 2013 |Campaigns & Elections pg. 26| Stefan Hankin.

In the post-analysis of the 2012 election there has been a lot of credit given to the Obama campaign for its use of capturing and mining data to help increase the turnout operation. Additionally, the new take on where and when to buy ads has been heralded as a new model for the future.

All of this credit is well deserved and should be discussed, looked at and adapted. However, what is getting less attention is how changes are happening in the campaign world on races where budget decisions are a lot tougher, personnel resources are less abundant and you actually have to work to get any earned media.

The industry has been changing at light speed over the last decade thanks to the growth of social media, the fragmentation of media consumption, the nationalization of politics and the advent of Super PACs. As a result, new ways of thinking are emerging both by necessity and by ingenuity. Campaigns that are ahead of these changes are showing a high level of success, while those who are running the same playbook from the 90s are more often than not sitting on the sidelines now and wondering what happened in November.

In the not-too-distant past, targeting on most campaigns was a vague concept that focused on large groups of supporters--women in districts A, B, D and F, for example. When it came to reaching these groups the discussion centered on how many gross rating points a campaign could have on the airwaves in the week before the election. If you are talking in those terms today, you are probably dealing with an outdated campaign.

The focus should be on concepts such as probabilities, volatility, ceilings and floors. If your messaging focuses on a "Message Box," take a good look around the room and realize you are looking at the equivalent of a Norman Rockwell painting. If the conversation is focused on message paths, game theory and messaging terrain, you are likely on a modern campaign.

An Unconventional Plan

My firm had the pleasure of working on a forward-thinking campaign this cycle in Ohio--Lou Gentile for state Senate. Gentile had been appointed to the office and was running for a full term. His district, Ohio SD-30, is in the southeast part of the state and includes 10 counties, 5 media markets and over 350,000 registered voters.

In 2012, it also happened to be one of the few, if not the only, truly competitive state Senate districts in Ohio. With a heavily contested presidential election and a competitive U.S. Senate race to contend with, we knew getting voters to pay attention was going to be tough.

From the start, Gentile made it clear that his expectation was that he was going to put together a team he felt comfortable with and one that he knew would tell him what he needed to hear, not what he wanted to hear. Thankfully, we had a candidate who was putting together a team that would figure out how to win, not just how to spend the campaign's money.

The biggest challenge was that we didn't have enough money to be up on every media market, and the large number of voters in the district made it impossible to just rely on multiple mail pieces in areas that were not seeing our television ads.

Some decisions were easy. We weren't going to run ads in the Cleveland and Columbus markets given the high cost and the fact that they each only hit one county in our Senate district. After that, the decisions became much tougher.

We took the findings of our baseline poll and began to overlay past election results to determine what the most likely scenario was going to be county by county. After multiple conference calls and dozens of mail and media plans, the team, led by general consultant Joe Rettof; media consultant Tierney Hunt, and mail consultant Karen Petel, came together and brought Gentile and Campaign Manager Ryan Braglin our recommendations.

At the beginning of the process each team member brought to the table their take on what a "conventional plan" would look like. From the media side this meant starting with an even number of ads (approximately 2,000 points) in both the Wheeling and Parkersburg media markets. While both are in West Virginia, their coverage area hits about 70 percent of Gentile's district. The additional 30 percent of the district, according to the baseline plan, would be receiving the bulk of the mail budget.

We decided that while this would be the safe decision, it was not going to be the winning strategy. So instead we formulated a plan based on probability of movement, volatility in the electorate and good old-fashioned math. The plan removed all points from the Parkersburg market and put the majority of the money into Wheeling. Additionally, we added targeted but low-level cable buys in the counties that were not covered by the Wheeling market and adjusted the mail plan to hit very specific demographic groups that focused more on weak-leaning Democrats and Republicans.

This less orthodox approach was met with some skepticism, to say the least. The Parkersburg media market represents 20 percent of Gentile's district, and we were telling the candidate he shouldn't spend in any serious way on TV ads there. When the candidate's friends and confidants were shown the plan, they were far from convinced.

As any good candidate should, Gentile questioned the logic of "going dark" in a good portion of the district. The answer from the team was simple--this is the right approach because the numbers say so. Sure, we could run ads in both media markets and run a conventional campaign and we still might win. But if that's the approach we took, we wouldn't have changed the probability of winning.

The numbers showed that regardless of how much time, money and effort was put into Parkersburg, our chances of growing our level of support above 40 percent were minimal at best. At the same time, the odds of us getting less than 36 percent there were incredibly low. Meanwhile, in Wheeling, the electorate was much more volatile so we could substantially outperform our needed target numbers, thereby changing the probability of success.

After a few conversations and discussions, Gentile approved the plan and with close monitoring throughout the last few months of the election through polling, phone IDs and other measurements such as early vote tracking, we were able to track the plan's performance and make small adjustments as needed.

In the end, Gentile won the race with 53 percent of the vote. Even more impressive was the fact that Gentile outperformed President Obama and Sen. Sherrod Brown (D) in eight of the counties while the three candidates ran equal in the remaining two.

It's just one example of how data-driven decisions are changing the face of campaigns, and I'd argue it's a reason why more candidates and consultants need to be drawing up their plan of attack in similar ways.

The New Landscape

It's always hard to know exactly when dynamics change, but 201.2 was clearly a new landscape. At the top of the ticket, both campaigns raised over Si billion. If you add in the outside money, the number more than doubles. The concentration of these funds and the attention heaped on 10 battleground states had a profound effect. If you were in a targeted state, just about every ad on TV from Labor Day until Election Day was political, households were receiving multiple phone calls a day for voter ID, persuasion and targeting, and turnout operations were being built and perfected on the precinct level for the top of the ticket.

If you were a state that received no attention, the turnout operation wasn't being run from the top and voters were just getting news about what was happening in states like Ohio and Virginia. In either case, a non-presidential or U.S. Senate candidate was almost assured an inability to break through. Campaigns that understood the dynamic they faced adjusted their plans and embraced a new way of looking at how to allocate the campaign's time and resources.

From a statistical standpoint, campaigns come down to how you can increase the probability that you will receive more votes than your opponent. Too often in the political campaign environment we equate spending money to increasing the chances of winning, but the discussion should actually be about changing likely outcomes. On smaller races, funds are always going to be lower than you want and tough decisions will need to be made when it comes to how to allocate those funds. In the past, the question tended to focus on what is the most cost-effective way to talk to the largest number of voters. The question should actually be: how do we increase the probability of success?

An additional challenge is getting past the notion that there is little incentive to take risks on campaigns. If consultants take a risk or try something different and the campaign goes down in flames, blame is most likely to be heaped on the consultant. The sad truth is that it's easier for consultants to play it safe, cash the checks and hope for the best. The risk of doing something different could be felt on the bottom line of a firm and that's seen as a bad business decision.

Thankfully this dynamic is changing as a new generation of operatives move up in the ranks. Safe and typical are less likely to be viewed as the best place to start, and less forward-thinking settings are viewed as completely outdated.

Stefan Hankin is founder and president of Lincoln Park Strategies, a Washington D.C.-based public opinion firm.

Hankin, Stefan