Article as it first appeared in Campaigns and Elections

Update: 9/7/2016

Data is not the answer? I know we’re not suppose to say that. Hear me out.

Data is commodity, like oil, lead, or pinto beans. It’s an object, not a strategy or solution. Data doesn’t solve problems, people solve problems.

This article was written for Campaigns and Elections a few years back. Two years is a long time and technology has obviously changed since this was posted. What struck me after rereading it, was that the idea that data can be a solution hasn’t really changed.

Please take a look. I’d love to get your feedback. Reach out to me directly at justin@votertrove.com with your thoughts.


WHY THE BEST DATA IN THE WORLD WON’T WIN YOU AN ELECTION

By Justin Gargiulo, Founder of VoterTrove

To extract the most value, campaigns will have to move beyond the idea of data as a solution.

Nowadays we talk about data as a solution to the problem of winning elections. It isn’t. Data is an object, not a technology, and objects by themselves do not solve complex problems. Technology does.

Me playing golf with any clubs

What do you think would happen if Joran Spieth gave me his personal set of custom golf clubs to try to qualify for the U.S. Open? I wouldn’t even come close.

What if I tried to qualify for the tournament with a 30-year-old set of clubs that was just run over in the parking lot? It would be the same exact result. No matter the caliber, golf clubs are still just objects. The golfer, though, isn’t.

The best data in the world won’t win you any elections just sitting on a hard drive, the same way that Tiger Woods’ golf clubs won’t help me win the U.S. Open. Thinking data is a perfect tool is a common misconception in the campaign world. Moreover, big data can in fact put a smaller campaign on a fast track to diminishing returns.

At a certain level, data can just become noise. From what we’ve seen, the available volume of cheap data impacts smaller campaigns in a few different ways. When these campaigns get their hands on it, they believe they have an advantage because they feel their opponent may not have the same data. Based on this perceived advantage, they’ll spend lots of time building strategy around it, often too much time.

Learning that there are seven non-primary voters in Omaha’s 7th Congressional District who read Field and Stream is not going to help a candidate get elected. If you have the resources to execute a programmatic microtargeting strategy at this level, then reaching out to those voters might just make sense. If you’re the other 99.9 percent of campaigns out there, it doesn’t.

Now, it may sound as if I’m bashing data. I’m not. I have been a data evangelist since 1998, when I was handed a Connecticut statewide voter file in Microsoft Access on three separate CDs and told to “figure it out.”

It’s incredible how much data is now available to campaigns. But my concern is with how it is, or isn’t, being used.

There are three parts to the data-as-a-solution process: Data is purchased and gathered (object), decisions are made (strategy and technology), and then action is taken. Then it starts all over again. Those CDs I got back in Connecticut were not going to solve any problems on their own; they were just step one in the process.

Data Informs Decisions

It’s easy to understand why Big Data in politics is sexy. The granularity of data available beyond the voter file is incredible. Put this together with data gathered by the campaign through phone banks, canvassing, autodials, and online, and it’s not inconceivable to have over 30 data enhancements on a single voter.

But after you have the data, what then? Putting these data points to use is the hard part. You need to ask yourself what pieces of data are going to inform your decisions. Of those pieces, what’s the hierarchical structure—which pieces have more influence? What is the universe size? Is it a narrow, persuadable universe, or is it a broader segment?

Data lays the foundation, and technology helps you build these targeted universes, but when it comes down to it, phase two is driven by real people, making real decisions on whom to talk to and what to say.

Where technology and data create real advantages is in the iterating that takes place on the backside of the feedback loop. How quickly can the campaign shift its assumptions based on the inbound, real-time flow of data?

What impact, if any, does the most recent data have on the campaign’s predictive models? This is where data and technology really begin to create huge advantages. At the beginning of phase three, technology becomes the solution, putting decisions into action and enabling fragmented datasets to be more easily digested by the campaign.

A Complex Mosaic

A mosaic is integration at its finest. Mosaics give meaning to disparate objects by weaving them together into a larger narrative. I’m not an artist, but if I were tasked with building a mosaic, I’m pretty sure I’d want to organize my pieces. Sorting all the available pieces together in buckets by color and size would enable me to more clearly evaluate the objects I have to work with, and it would greatly increase efficiency when I begin to create my mosaic.

Campaigns face similar challenges in informing decisions and executing strategy. For most, it is the task of effectively assembling data in a way to help them build a narrative. Failure to organize and standardize data will lead to mosaics that are either incomplete, inaccurate, or both.

This problem is not solved by more data. In fact, our work with campaigns has shown us that more data only worsens the problem. The problem is solved by integration, and enabled by technology.

Data is the Gas. Integration is the Engine

A system with a smaller amount of integrated data is more valuable than a larger amount of siloed, fragmented data. Integration gives data exponential value. The whole is greater than the sum of its parts.

There is no clearer example of the benefits of data integration than in GOTV programs. Most recently, our clients in the June primaries leveraged our platform to consolidate positive ID data via CSV imports, web forms, phone banks, canvassing, Facebook, and more.

Not only does this integration generate a more complete mosaic, it also greatly increases operational efficiencies. Many political directors we work with estimate that a consolidated data platform saves them anywhere from five to 15 hours a week in time typically spent cobbling together fragmented spreadsheets.

When campaigns begin to implement their strategy, the integration of outbound messaging capabilities into a central data management system provides an added layer of value. As action is taken, data is fed back into the centralized platform, enabling the campaign to adjust its strategy as needed. Just like data, disparate technology platforms that can’t talk to each other can be a drag on a campaign’s resources.

Just as data integration saves time, so does platform integration. Learning one platform or interface that does 10 things is much easier than learning 10 that do 10 things. For the 99 percent of campaigns that need to operate lean and scrappy, platform integration can add enormous value.

Rapid response is a great example of this. After listening to her candidate’s opponent grossly distort the voting record of her candidate, the campaign manager on one of our campaigns acted. Within five minutes, she targeted and sent a “press-one” call to all of the campaign’s supporters urging them to connect to the radio station to defend her candidate.

There’s More to Come

Just as with the explosion of cheap data, the emergence of APIs, or application programming interfaces, has enabled technology companies to build products that do more for less. With each passing cycle, campaigns will have a growing selection of integrated technology platforms to choose from. To extract the most value from these advances, campaigns will need to move beyond the idea of data as a solution.

Leave a Reply

Your email address will not be published.Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>