First published in Research World May 2009
Joel Benenson, pollster for the Obama campaign, talks to Simon Chadwick about how to use research to focus on what makes the most impact.
You were with the Obama campaign right from the beginning in January 2007. Did it seem like a long shot?
I think a lot of people thought it was a long shot. One of the strengths of our campaign was that none of us ever succumbed to the conventional wisdom.
Would you agree that in many ways the success of the campaign revolved around establishing a clear and strong Brand Obama and did research help to define and refine that brand?
When you talk about a brand in a political context, what was clear from the beginning was that President Obama believed in certain things very strongly. He believed in a certain kind of politics that he was going to practice that was going to be less partisan. He believed that the American people were often ahead of the politicians in understanding what’s going on, and that he was going to speak to them frankly and candidly. And even where there were things that didn’t seem popular, he was still going to say them. And so the word ‘brand’, it’s easy to make it sound like it’s something that’s manufactured; in this case it was just who he was and how he wanted to run his campaign.
The research didn’t help develop any of that but it did help to elucidate certain points. We knew we were running in a ‘change’ election and that Obama was going to represent change and talk about change, but we were able to use the research to sharpen the vision of change. There were lots of things that could represent change, so we used the research to focus in on the aspects of change that would have most impact.
And so the research informed not only the message but the feel of the campaign as well?
Yes. But President Obama had such a clear sense of what he believed, and where he wanted to take the country that the things that we researched grew out of his ideas and approaches, and we constructed research that simply tried to strengthen what he was saying. Some candidates are at a loss for what issues they want to talk about and where they should put their focus. He never was, he was always spot-on.
There are a lot of politicians who are obsessed with polling data. He’s not. During the campaign, he wanted the broad picture more than the details. He wanted a sense of what was happening out there. He would pay attention to the trends and want to know whether things were changing, or if attitudes were shifting. I also remember a conversation during the campaign with Vice-President Biden, who would come back and say, “Here’s what I’m feeling out there from people, in this state or that state,” or “Missouri, I think it’s going to be close.” And he was right, we almost did win Missouri. So that’s an asset that some candidates have, a good sense of what’s going on.
A key difference between the Obama and the McCain campaigns was that we were very strategic all the way through and they were very tactical.
Were you also able to define McCain in the minds of the public?
One of the things that the research told us early on was that this image of John McCain as an independent or a maverick of sorts didn’t exist outside Washington. In fact when he started describing himself as a maverick we immediately tested ‘maverick’ as an attribute in the poll on each candidate, and the fact of the matter was that by two points Senator Obama was seen as more of a maverick than McCain.
How did your polling managed to decipher the real meaning of experience?
We had a very strong team doing polling in focus groups. Experience can mean different things for people, and so when focus groups were conducted we would ask what people were looking for when they talked about experience. Rather than using superficial words, we were able to create attributes that were important characteristics that voters really wanted in a candidate, that people really valued as part of the bigger issues. For example, ‘steady in crisis’ was a quality of experience that they liked. If you ask people if someone ‘has the right experience to be President’, what you’re basically saying is ‘evaluate their resumé’. And one of the things we were able to use the research for, early on, was to identify that it was about leadership qualities, not about resumé credentials.
What people wanted in terms of a president was someone who would show strong leadership, who would tell people not just what they wanted to hear, who was steady in a crisis. Those were all things that Barack Obama was building his campaign on anyway, but we were able to show his good judgement. That was very important, as a component of his experience. We used our research to drill deeper, and learned that we were very competitive on these qualities that people valued. It enabled us to resist a conventional debate on the word ‘experience’.
You are also partners with iModerate. How did the inclusion of instant messaging IDIs in quant interviews give the campaign an advantage?
We did the bulk of our ad testing in dial groups and focus groups, and we only used iModerate once or twice early on. We had a hypothesis that the power of Obama’s voice was extraordinary, that when people watched him speak, they were not just moved but they immediately sensed that this was someone who was speaking with great conviction, who was very committed to what he believed in. And so when we tested some ads early on, we did one internet test and we used iModerate, and it confirmed what we had heard in focus groups, that ads that had him in them had a disproportionate power, that people reacted very positively – and not just positively to the ad, but about those characteristics they saw. So based on that testing, we made a strategic decision that we were always going to have an ad on the air that had him in it speaking to people.
In the ’92 election a lot was made of the War Room for rapid reaction. Was there something similar here?
When you do research in a political campaign, we explore ways we can defend ourselves and respond. So that by the time the attack comes, you’re almost never taken by surprise. We didn’t set up a room but there were daily calls at both ends of the day, but it’s really in the preparations, in thinking strategically about how is the other side going to come at us, and how do we answer it? In July we did an attack and response poll of everything we thought the McCain campaign would throw at us, and tested potential responses to it.
By 2012 it’s likely that ‘mobile-only’ households will be a sizeable proportion of the population. How did you overcome it in this election, when the young voter was so important to you?
When you poll, you don’t have a list of people with mobile phones but you can ask them if they’re answering a cell phone, and there are a lot of places where the area codes are the same, so you have some sense of whether you’re getting them.
We knew what the incidence of cell-phone-only usage was, among young people, and we ran a check on that, but we also did a parallel poll to see if demographically, attitudinally, politically, there were significant differences between people between the ages of 18 and 30 who had a land line and people of that age group who were mobile only. We learned that attitudinally the populations were the same, so we didn’t inject mobile polling or mobile phones into our sampling. But when we created our sample plan, we generated more numbers in that age-range, to get enough respondents.
We’ll have to wait and see if this will hold water in four years’ time. I suspect that there’ll be more lists emerging, and that obviously poses one of the challenges that we’ll all have to think through by 2012.
You regard going through the data as almost investigative detective work. Is it all about how you analyse the data?
Look, if you write a good survey, you have written a survey to produce data that’s going to be impactful in terms of your analysis and in terms of the decisions that you’re trying to make. You have some hypotheses when you draft the survey, and you develop the best way to test those hypotheses. I view the questionnaire as the art form. That’s where you really lay down the architecture of what your data could look like, and if you’ve done it well you’re going to have a very rich data set.
The point is that there are going to be clues buried in there. I sometimes see people who look at cross tabs and they’ve got ten different data points and sub-groups that they’re looking at. We typically have 56, we set 28 on a page, but frequently we create 84 different groups, because we don’t just create demographic groups, we often use multiple variables, attitudinal variables that are able to create very creative groups, because we use a lot of techniques to figure out what might be driving their particular responses. We really spend a lot of time looking at what underlying attitudes and values or baggage people bring to the decision-making table that affects how they react to information and make decisions.
It’s not just about what we say; it’s about what they’re bringing to the table as well. And we designed surveys that allowed us to look at both ends of that.
What lessons do you take out of all of this, for your corporate consulting?
I’ve had the luxury of working with some great corporate clients. I think that too many corporations are locked in to doing what they know and what they’re comfortable with and repeating it, and they like the consistency of always doing something the same way. I do too, but the reality is that dynamics change daily and monthly.
Competitors come at you from different ways. If you’re in the record business and you sell CDs, what do you do about digital downloads? Being very attuned to what changes are happening in people’s lives and how your situation might change is vital. Even if you’re not in a competitive situation, we often help our corporate clients understand layers and textures that are at work in the decision-making process. There’s a competitive framework and mindset that we bring to everything we do.
Joel Benenson is CEO of the Benenson Group