• by Tom Hespos , September 9, 2002

It was 10 PM on a gorgeous summer night and I was hanging out in my mom’s driveway with one of my best friends from high school. “Sam” was considering leaving his career and running for political office. He had come to me to seek advice on how he might take his message to the people in his town.

Sam knew more about marketing than I was prepared to give him credit for. Campaigning for local office in this part of the country is down to a science. You ignore the other party’s loyalists. To an extent, you ignore the people who vote with your party as well. The real battle is fought on a much smaller landscape that consists of a handful of swing voters who pay close attention to the issues.

“Tom,” he asked, “how can the Internet help me?”

I started with a line of reasoning that is anything but new. “Sam, the Internet is a great medium for presenting a message with depth,” I said. “You can generate awareness, and for the people who have a legitimate interest in who you are and what you stand for, there’s always a website.”

Sam was intrigued, but he wanted to know more.

“Why don’t I see other candidates using the Internet to get the word out? What can the Internet do to get me closer to the swing voters?”

“That’s a toughie,” I said. We talked about the characteristics of the audience for a while. They vote, they don’t always vote for their party’s candidate, but they definitely get out and vote.

“We could gather a list of swing voters by looking at voting records,” I said. “Then, we could reverse-append email addresses to those names and have a working list of folks to send email to.”

Problem is, such a campaign could be damaging. People are becoming more and more sensitive to unsolicited email, and someone could just as easily make a decision to vote against Sam as they could to vote for him, influenced solely by their perception of Sam as a spammer.

“Then again, maybe the answer is to run geo-targeted web ads to a list of ZIP codes,” I said. “And you definitely should have a website. The people who are politically conscious enough to care about local elections will definitely expect you to have one. That could help to swing a few votes your way.”

Sam left that night and we promised to talk about using the Internet to help him with his campaign at some later date. No doubt there will be a small “Vote for Sam” Internet campaign at some point, but the whole conversation got me thinking…

At this point in the game, I should have had a lot more to offer Sam. The Internet promises to be a medium in which we can build mass against niche audiences. The problem is, how do we identify niche audiences when they’re so incredibly obscure?

Content doesn’t always do the job. I’ve always been a fan of advertising products to the low-hanging fruit by running on sites that are tied into the product category or the brand in some way. But the truth is, people only have enough bandwidth to visit sites pertaining to only a handful of interests. For instance, I’m into tropical fish as a hobby, but I visit tropical fish-related websites only a couple of times a year. How would a marketing company reach me with an ad for fish food? With the current state of ad technology, no one would be able to identify me as a part of that audience.

This got me thinking about my buddy Dave Morgan and his company, Tacoda Systems. Tacoda has built a system that makes it much easier for publishers to sell ads against niche audiences. Of course, Tacoda can do much, much more for publishers, but it is the promise of being able to reach specific niches that truly intrigues me.

The concept is simple – define an audience with a set of rules. For Sam’s campaign, this might be the people in his particular town who voted in the last few elections, but didn’t always stick with their political party. Pull the data to define this audience from several sources – content visited, behavioral data, declared data, external databases. Then, export this audience to a publisher’s ad server and see how many people can be reached through that publisher’s website, newsletters and other Internet communications.

If every publisher had something like Tacoda, I could be calling sites and asking them to profile the audience for me now. “I want swing voters in this small township on Long Island,” I could say to them. “Here’s how I want to define this audience. Start looking at your internal data now and let me know how many people I can reach that fit these criteria.”

Folks, I’m really looking forward to an Internet that has the potential mass reach of broadcast media, the self-selection of print media and the specificity of direct mail. That is the promise. And we’re getting there, to a degree, but we need more tools like Tacoda to help us leverage everything publishers know about their audience. To fail to capitalize on this strength would be a big mistake.

Gone are the days when publishers collected a ZIP code from users during their site’s registration process, but failed to be able to deliver targeted inventory against that ZIP code. Declared and observed data need to be leveraged at every opportunity. It is this data and the resultant intelligence about a publisher’s audience that can take this medium to the next level.