Online Political Ads Find Lonely, Isolated Voters—And Barrage Them

The boss of a leading voter-tech firm explains why we all hate each other

Voters line up to cast their ballots on Super Tuesday March 1, 2016 in Fort Worth, Texas. (Photo by Ron Jenkins/Getty Images)

Voters line up to cast their ballots on Super Tuesday March 1, 2016 in Fort Worth, Texas. (Photo by Ron Jenkins/Getty Images) Ron Jenkins/Getty Images

My firm’s ad targeting technology has powered thousands of campaigns, won two patents, and has become one of the most widely used voter targeted digital ad platforms in the country. We win by targeting the smallest of segments with the hardest hitting ads. And it’s exposed a weakness in our system that probably can’t stand the test of time.

Broadcast media is in irreversible decline, and incumbents at all levels are safer than ever before, and the stuff holding us together—communities, leagues, clubs, and associations—isn’t what it used to be. Bowling Alone, the Robert D. Putnam book from 2000 about our withering community ties, was written before Facebook, before Pandora, and before broadband addressability. If anything, it underestimated the speed at which our connective tissue would rot. For all the talk of “community” on the web, we are increasingly communities of one, hitting Like to substitute for showing up.

Now more than ever, the conditions exist for us to scream past each other, lack empathy, and think the other guy possibly hates us. We all bear responsibility for contributing to this; the most successful of us leverage this trend to win. But I will never disarm and neither will my competitors. At the very least, the political and public affairs community should acknowledge there’s a problem because … you live here too, guys.

In the near future I will tell my children that there was once an indisputable truth based on an agreed-upon set of facts. There may have been some debate a century ago about who sunk the Maine; but somehow arguing the facts out in a daily newspaper seems less virulent than what goes on today. With consolidation of ownership, but fragmentation of outlets today, we have hundreds of angry little William Randolph Hearsts, all screaming different truths in real time directly onto our iPhones. No filters. No delay. And no worries about being embarrassed at the absurdity of their argument.

Digital ad spending on campaigns, about $22M in 2008, is projected by to be $3.3B by 2020.

We will be the last generation to experience broadcast media in any recognizable form. The network news audience is one quarter the size it was in the 1980’s. Brian Williams is half the man Tom Brokaw was, and it’s not all his fault. Half of all broadcast radio stations will be gone within a decade. The AM/FM stations in my four-year-old Jeep are still on their factory settings. Traditional TV viewership has plummeted for everyone under 50. The high water mark for television ownership passed around the time that Extreme Makeover: Home Edition was cancelled. (There may never again be a Ty Pennington!)

You can argue when and how it’s going to happen, but broadcast media has entered its terminal phase.

I don’t care so much for the broadcast industry, but I do care that it held us together as a nation. All that wasted broadcast advertising had an unintended benefit: it kept the really nasty language in check. The more addressable our media, the less we worry about an errant message saying Barack Obama is from Kenya or that George W. Bush hates black people.

What happened? Technology, money, and the law happened.

Thirty years ago, computing processor speed passed the point where it could become a useful tool. Moore’s Law, which says that computers are improving at an exponential pace, allowed for data processing at a scale where voter registration and participation data could be sorted and distributed to printers and telemarketers. (Your iPhone has 6000x more memory than the Voyager 1 spacecraft.) This was such a novel development that in 1982 the cover of Campaigns & Elections magazine featured “Microcomputers: Their Use in Small-Scale Political Campaigning.” The article featured a room-sized computer spitting out data that today we’d consider totally worthless. But it was notable enough even then to warrant heavy coverage in the magazine that I later ran from 2006 to 2010.

Without exponentially-growing computing power, we could have never have had the national voter file. First came vote history data from the states and counties, processed to allow message segmentation. Candidates and causes could target just primary voters instead of everyone in promising zip codes. Political direct mail and telemarketing experts got rich by stealing market share from broadcast and newspapers, because they could find the subset of Americans who made candidates their party nominee.

We began targeting primary voters because they voted in past primaries, and began the feedback loop that has a lot to do with why we all hate each other. With uncompetitive general elections – the vast majority of elections in America – moderate voters often don’t have much value to someone like me. We spend less time talking to them about big problems likes ISIS and infrastructure and instead want to legislate where Caitlyn Jenner can pee.

Broadcast television has dipped under half of all campaign spending in many races. Newspapers accelerated their dive. Things that could be targeted, even nominally like cable, grew in market share and even more so real dollars. In the years between just 2008 and 2016, Borrell Associates says that addressable media (cable, telemarketing, direct mail, and Internet) went from about 17.5% market share to over 25% market share of all political spending. Addressable media will account for more than 40% of all political spending by the late 2020’s. This has all happened in the vacuum of virtually zero regulations, as the FEC still acts like it is 1985.

The Help America Vote Act sought to ensure that hanging chads would never determine an election, but also required secretaries of state to get their act together when it comes to keeping up voter lists. The national voter file, once sold by three small companies, is now available at a fraction of the cost from more than dozen companies, each competing to be the lowest bidder. The bullets got cheaper while spending blew up. Digital ad spending on campaigns, about $22M in 2008, is projected by to be $3.3B by 2020.

The last time polarization was at this level, we had just given black men (but certainly not women) the right to vote. According to Pew, in 1994, 17% of Republicans thought that Democrats were a threat to the nation’s well being. By 2014, that was 36%. Since 1992, the number of self declared moderates in America has dropped from 40% to 34%. Religious intermarriage might be up, but conservatives marrying liberals… hell no. One overlooked fact about all this money pouring into the system is that because we are so good at gerrymandering, the funds are aimed at just thirty or forty competitive congressional races.

My firm, Audience Partners, began deploying this technology around 2009 and later won two patents for voter targeting off the voter file. That allowed advertisers to disaggregate the audience from the website they were visiting. The math involved with saving clients’ money by only delivering a message to their intended audience is compelling. But that doesn’t exactly mean it’s good for civil discourse. We talk to voters like they’re in a soundproof room rather than crowded party. And by being the first company in what’s now a industry billing hundreds of millions of dollars per year, we opened the door to what’s next, well beyond connected devices.

Tack on thousands of additional data points, and now my ability to target is greater than your ability to develop different creative versions.

My point is that this narrowcasting make it harder to live as a nation of communities. That’s the phrase George Bush used to describe us in his Thousand Points of Light speech, accepting the Republican nomination in 1988.

It’s rare when I’m targeting more than 4% of any population for a candidate or cause. 96% of you will never hear the outrageous thing I just said to your neighbor, best friend, or coworker. Tack on thousands of additional data points, and now my ability to target is greater than your ability to develop different creative versions. Do this over and over, trillions of times over a decade or two, and you should not expect us to come back together for a sane school board meeting or care about a bowling league.

The end result?

We’ve created some very, very angry people. Targeting primary voters created a feedback loop where we only talk to engaged voters, and people are arguably engaged because they are hit repeatedly to draw out the most vitriolic messengers and messages. And just like Paris Hilton is famous for being famous, engaged voters are engaged because we engage them.

We contributed to this. And we can help contribute to a solution. That starts with recognizing that our more powerful targeting tools, decline of all things broadcast, and rotting connective tissue mean that competitive elections are more important than ever. And just maybe we want to draw political lines that look like normal shapes we learned in kindergarten, help the FEC recognize that things like text messaging, email, and the Internet exist.

Jordan Lieberman is president of CampaignGrid, a division of Audience Partners, which developed the first platform to deliver data-driven online advertising for political, advocacy and public affairs campaigns.  He is the former publisher of Campaigns & Elections magazine.

Online Political Ads Find Lonely, Isolated Voters—And Barrage Them