Eric Cantor’s loss may be Patrick Ruffini and Kristen Soltis Anderson’s gain. The young digital strategist and pollster launched Echelon Insights in the wake of the former House majority leader’s primary loss. Their pitch: data integration can help GOP campaigns spot pending changes in public opinion, or corporations identify emerging trends.
It’s all about a “more integrated decision-making capability,” according to Ruffini. They want Republican campaigns to get to the point where “decision makers are basing their decisions on data as opposed to what the highest paid person in the room says.”
“Campaigns, in general, are a very siloed enterprise between the mail folks, the TV folks, the digital folks and I think a model that fits inherently challenges people to share data across silos to create a more integrated and rational decision making capacity,” says Ruffini. “To us, that’s what the focus is: integrating a lot of these pieces.”
C&E: Was it luck or planning to launch the firm after Cantor’s loss?
Ruffini: We’ve been working on this for a few months, and it just happened to occur before the intended launch of the firm. It certainly raised a lot of issues. I think the broader issue is, what is the general state of measurement on campaigns? How do we more effectively measure political efforts and bring other forms of data into the process, whether that’s analytics or increasing data from the online world, like search volume. There’s certainly a need for us to think differently about how we actually measure how we’re doing in these efforts.
C&E: Do you think Cantor’s polling was of because respondents knew it was his campaign behind the call?
Ruffini: There have been a lot of questions raised. Were they polling a broad enough sub-sample of the population, for instance? To me the broader issue is that very little polling was done. So in a very late-breaking race, which a lot of primaries tend to be, I don’t think there was a poll done late enough to actually capture [Brat’s] momentum. And looking at, frankly, other data sources like search trends that could provide hints as to what’s been going on [didn’t happen]. Our theory is that you can better prepare for these types of situations if all your data is integrated in a single place, in a holistic manner.
C&E: So to really get an accurate picture you need all these other data sources?
Ruffini: Our point isn’t to say, ‘social media is more accurate than a poll.’ There’s clearly still a lot of validity to polling. It gets it right far more often than it gets it wrong. But in certain situations like this one, it was a snap shot in time. But in order to spot the canary in the coal mine, to predict an impending movement in public opinion, I think you would look at a variety of factors—from donor information to social media information to search volume.
There’s a lot of competitive intelligence you can do on your opponent in terms of their operational capability. So our goal is to integrate all that in a single place. Maybe a trend is not going to be immediately apparent in a poll, but it will be in all of these other indicators that are giving you more of a complete snapshot of what’s going on.
C&E: Are you setting up an office?
Ruffini: There will be a real office, most likely Alexandria, but we’re still feeling it out. There will be resources. This will be our primary project. We’re very excited about where this can go. We’re looking to hire probably a half-dozen [staff] in pretty short order and then go from there.
C&E: What’s your focus for 2014?
Ruffini: We’re looking ahead. We don’t have huge aspirations to come in at the last minute on a lot of campaigns. But we are looking to collect [data] for a lot of people who were frankly surprised by the types of results we saw in Virginia 7. Are there better ways we can help predict what’s going to happen? To a large degree, our work will probably be more with corporations and a little less dependent on the cycle.
C&E: How is polling evolving?
Ruffini: In the last couple years we’ve seen a huge proliferation of the ability to measure things directly as opposed to simply from a survey. We certainly have survey capability. But that is increasingly being complemented by broad data feeds from online that are being made available based on user behavior. That is going to be the future of measurement in the marketplace. I think we need to be focused more on how to facilitate more fundamental, broad efficiencies. I think that starts here.