To subscribe to the monthly C&E email newsletter and event announcements click here.

Presidential elections sometimes create seismic waves in the campaign industry and last cycle is proving to be the Big One. Assessing the damage, it appears Big Data has taken a hit.

Consider broadcaster Chuck Todd’s recent piece headlined, “How Big Data Broke American Politics,” in which he argues that federal pols believe they “don't need centrist or swing voters to win.” As a result, we get congressional gridlock. Or more specifically in the case of the Democratic Party, you get electoral defeat.

Before Todd offered his theory, Democratic consultant Dave Gold wrote a piece in February headlined, “Data-Driven Campaigns are Killing the Democratic Party,” in which he notes that “microtargeting leads to microthinking.”

As the CEO of a national, non-partisan political data and technology firm, I mostly agree — with a few tsunami-sized caveats.

Recently, Democrats, have gone down the rabbit hole of politics by addition, while the Republicans have been using multiplication. The Democratic vision of a “whole” created from the aggregation of dozens of fiercely self-identifying individual groups hinders a campaign’s effort to create an overarching emotional theme that connects all voters.  

Extreme microtargeting pushes candidates to create a campaign of bullet points tailored to each subset of the electorate. While that might have worked well in 19th century American politics when communication wasn’t instant, today it results in a 24/7 cable news machine that shows a candidate who appears to lack a central message and core. 

None of this is to say that Democrats are uniquely susceptible to this problem. Republicans have certainly had their own campaign failures. “Division” is the flip side of “multiplication.”

President Trump’s “Make America Great Again” worked because it was a call to action that could, at least by supporters, be seen as cutting across many interest group lines whereas Hillary Clinton’s “Stronger Together” feels like a warm embrace with no particular goal in mind and reminds us of our divisions.

By now we’re all aware that Clinton won around 3 million more votes than Trump. But she’d be sitting in the White House today had she followed her husband Bill’s advice and spoken with a meaningful message to the large swaths of formerly Democratic-leaning voters in places like western Pennsylvania, Michigan, Wisconsin and Ohio.  

Her failure as a candidate reminds me of the famously self-destructive television interview that Ted Kennedy had with Roger Mudd. With Kennedy frozen like a deer in the headlights, Mudd asked the obvious and softball question, “Why do you want to be President?”  Kennedy’s answer was a political disaster — halting, muddied and vague. 

That response, the reaction to it and his failure in the primary made it crystal clear that it’s not enough to have a masterful political strategy for winning the election and certainly not enough to feel that it’s your turn. If you can’t tell people why you want their vote and what you will do with it if they give it to you, you’re as lost as Ted Kennedy was when Mudd asked him that simple and direct question. Bottom line? No core message, no victory.

Where I disagree a bit with Gold is that the problem isn’t the data, it’s the way data are used. The siren song of data analysis has lured many candidates to their destruction on the unforgiving electoral rocks.

But that’s because they have learned the wrong lessons from those analyses. The analyses should have helped them understand how to talk to specific subsets of the voters — not what to say.  It’s about the techniques of communication — not the message itself.

No campaign can win without the right message. That said, no campaign can win without the right data and analyses as well.

In the political data technology business we like to think of ourselves as providing the means for candidates to communicate with the voters. Without accurately processed and enhanced data and the right platform to use it, that communication becomes inefficient, expensive and misdirected.  

Undeliverable addresses, incorrect telephone numbers, bad digital ID matching and other problems waste precious campaign dollars and these are just the basics. Enhancements to the national voter file with sophisticated predictive modeled data on issues and attitudes and commercial attribute enhancements supplemented by private data gathered directly by the campaign makes it possible to understand the passions and motivations of the voters in each state.  

The most successful candidates will use those data and targeting derived from them to convey their core campaign message in emotional language tailored to each subset of voters. It’s the same message — it must be the same message — but it’s delivered in different ways to touch each voter where he or she lives. An argument for a stronger economy is a conversation about bread on the table for one voter and freedom from government interference for another. National defense is an issue of patriotism for some and a concern about safety for others.  

Gold is right in pointing to the need for a campaign narrative and framework, which is something that can’t be developed by committee. It must flow from the souls of the candidates themselves and must inform the very rationale for their decision to run. But without the data necessary to target that message and guide the candidate in communicating that passion, it will all be for naught.

Is the existence of Big Data killing the Democratic Party?

Absolutely not.

Are consultants sometimes using those data in such a way as to crowd out and confuse what should be a passionate campaign mission statement?


As Democrats look toward the midterms, their biggest challenge from a data perspective is the monopoly they’ve created both on the national and state levels. For some time now, central control has limited access for Democratic candidates to a small data universe and limited DNC-approved technology choices.  Privately-collected data can and should always be kept secure and in-house, but there’s nothing inherently partisan about richly-enhanced and highly-processed national voter files. Election 2018 will give Democrats a chance to reset both their data programs and their core message.

Bruce Willsie is the CEO of L2, a political data processing and technology firm.