Maren Hesla is a principal at Mission Control, a Democratic direct mail firm. Ed Peavy is the founder of Mission Control. Aaron Leibowitz is a principal at Targeted Creative Communications, a Republican direct mail firm.

For 20 years, people have been writing the obituary of political mail. But richer targeting and smarter creative, coupled with the increasing difficulty of reaching people through traditional broadcast media, have made mail more indispensable than ever when it comes to reaching voters.

In all forms of communication, the trend is toward individualization and personalization. But while we await the promise of TV ads delivered to specific cable boxes and cookie-matching that allows saturation-level online communication, mail is already reaching specific voters with targeted messages.

Think about that again—with today’s technology only one medium is delivering the right message to the voters who can be most moved by it with 100 percent penetration and a high level of confidence: direct mail. In this changing campaign landscape, that makes us in the direct mail world feel much better. What follows is why we’re convinced the state of the political mail industry remains strong.

Targeting

Just 10 years ago, mail targeting was largely limited to age, gender and precinct-level geographic data. Party registration and race were available in limited states. But better voter files, bigger computers and the availability of commercial data have led to the widespread availability of modeling, which targets voters at an individual level. All contested House and Senate races offer race-specific models that allow us to target on partisanship and turnout. In highly partisan precincts we can cherry-pick those individuals who are more open to our candidates.

Additionally, we’re finding increasingly accurate issue-specific models that allow us to target voters based on their attitudes toward abortion, guns, environmental concerns and other issues. We can frequently hone in on very specific voters.

In Elizabeth Warren’s Senate campaign in Massachusetts, voters who supported President Obama but were not yet in support of Warren were a key target group. It wasn’t that this group was necessarily supporting Obama and Republican Scott Brown—more often than not they simply didn’t know Warren well enough to support her. And since their partisanship was weak, this group did not automatically endorse the Democratic candidate. These “Obama-not-Warren” voters were successfully modeled and they became an important mail target.

The Rodney Davis campaign in the 13th Congressional District in Illinois had a similar problem, only compounded. Due to the dynamics of Davis’ late entry into the race, Republicans weren’t rallying as quickly as the campaign needed. Many seniors in the district also liked that Davis’ opponent David Gill was a doctor. During the final month of the campaign, a model of “GOP/Romney-not-Davis” voters was created and used in the areas of the district covered by St. Louis broadcast television to supplement the relatively thin electronic message buy. A group of nonpartisan, high probability senior voters who disliked cuts to Medicare more than they liked the Affordable Care Act were also targeted uniquely through the mail.

This sort of targeting will only become richer and more precise in future election cycles. The lesson from the corporate world is that multiple iterations provide the best models, and as we become more adept at using low-cost IVR surveys, at capturing and incorporating individual-level response data from field programs, and in refining our use of commercial data, the models will become more predictive.

Data Quality

A less heralded benefit of the advancement of modeling is improved data quality. Mailing to old addresses, deceased voters and undeliverable addresses has wasted untold amounts of money. But improved NCOA (National Change of Address) technology from the postal service, the availability of commercial Change of Address lists and an improved ability to purge deceased voters from files has lead to a significantly higher rate of delivery. We do still suffer from the “garbage in, garbage-out” axiom in data management, but our tools to reduce the garbage have improved.

Testing

Perhaps no area of direct mail has shown as much growth as the tools we use to test and refine mail. For years, our best test of mail was in focus groups. But that unnatural setting produced flawed feedback. Sitting around a table full of strangers, people inevitably bemoaned negativity and asked for straightforward, comparative, well-sourced information. They read each piece cover-to-cover, questioned its veracity and offered suggestions on graphic design. In the real world, voters glance at a piece of mail for the amount of time it takes them to travel from the mailbox to the trash can.