Dean: Technology, campaign strategy and what went wrong

Just got back from guest lecturing at the Harris School of Public Policy at the University of Chicago. I was asked to join Pat Quinn and Kevin Conlon to talk about political leadership; Quinn, to talk about his decades of service protecting consumers in Illinois, and Conlon and I to talk about the Dean campaign.

I needed to hear Kevin’s story of how he ended up running the Dean campaign in Illinois — I hadn’t heard it before, and it reaffirmed my original enthusiasm for the campaign. (Recalling that enthusiasm takes some effort these days.) Kevin ultimately signed up not because he was angling for an ambassadorship if Dean won — no, he signed up because he and Governor Dean shared some strong feelings on key issues, and after spending a couple days with Gov. Dean last spring, he concluded that Gov. Dean was a good person who was in the race for the right reasons. As simple as that.

Between hearing Lt. Gov. Quinn’s remarkable list of accomplishments and Kevin’s decades of service to the labor movement and the political establishment, I felt more than a bit out of my league. But it was fun to tell the story to a group of graduate students who seemed unfamiliar with some of the inner workings of the campaign. Here’s a rough sketch of what I talked about:

I first contacted the Dean campaign in August of 2002. Shortly after reading this article, I called Burlington and volunteered to do whatever I could. I even recommended they start a weblog. (I’m sure I wasn’t the first, or by any means, the only one, to be suggesting that back then.)

Fast forward to the Winter. I had a handful of conversations with Bobby Clark, then Mathew Gross. A rough plan emerged, with me being a utility player to plug some holes on the weblog as they needed help. I ended up not doing much until late May, as the transition from Blogger to Movable Type was afoot. We launched the converted site — the first presidential candidate weblog to let supporters (and opponents) talk back — on June 10, 2003.

In the class tonight, I encouraged the students to read The Cluetrain Manifesto (you can buy a copy here — and talked in particular about the fourth chapter by David Weinberger and Doc Searls. David and Doc talk about markets as conversations — and I told these students that what the Dean campaign figured out was that online politics was just as much about the conversation. (That David consulted to the campaign should come as no surprise.)

Ultimately, the Dean campaign needed three things at the beginning of 2003. It needed money, people, and name recognition. You can’t have one without the other two, although you really want a balance of all three to be a successful candidate. Dean had none of the three in January. By any traditional playbook, there was really no way to emerge from the back of the pack and capture the nomination — so Joe Trippi and Governor Dean figured out that they needed to do something different.

That “something” was to decentralize the campaign — and much of the technology that emerged followed the same pattern: shift the responsibility outward to the fringes. Get the grassroots involved. Fundraising? They can raise their own money. Events? Plan their own. Posters? Print their own. Meetings? Meetup.com. And so on. Never had a campaign given up so much control at the outset — and yet it started working. When we led all candidates in Q2 with $7.6m, it was clear the strategy was working.

Recall the three goals: people, money, name recognition. By the end of the year, the campaign had raised over $40m, had 600,000 plus names on its list (its own record for a non-incumbent) and became the first non-world leader to grace the covers of all three newsweeklies in more than 30 years.

By all accounts, the campaign addressed its three goals. How? By creating an environment in which the volunteers were activated, passionate, and committed. It wasn’t just that the technology let the volunteers participate. It was that the technology put the supporters in control, recognized their contributions, encouraged healthy competition (both within the campaign and with other campaigns) and personalized the campaign.

• Control. Already covered: if you wanted to hold a “Cat lovers for Dean” party, nothing was stopping you. And a search at the Dean website would lead other feline fans to your door.


• Recognition. At every opportunity, the campaign went to great lengths to highlight who did what. If someone said something in the comments to the blog that the campaign liked, the campaign highlighted it. (Here’s one example.) DeanLink, the campaign’s first stab at a social networking service, actually posted the pictures of the top DeanLink members (i.e., those who had recruited the most new people). Jontathan K-T, the 14 year-old wunderkind from Alaska, received enormous national press for his role in the campaign. He brought more than 500 people into the campaign.


• Competition. When the campaign put up its first bat — to challenge the $250,000 fundraiser Vice President Cheney was hosting the following Monday — Gov. Dean ate a $3 turkey sandwich and challenged his supporters to outraise Bush. The result? They more than doubled Cheney’s take, proving that the Q2 fundraising haul was no fluke. Supporters were encouraged to hold house parties — and those whose parties were raising the most would get highlighted — leading to hosts working doubly hard to get onto the “leader board”.

• Personalization. This campaign had not one face, but thousands. But it’s the faces and names that emerged from Burlington that cemented the grassroots’ connection to the campaign. We got to know Mathew, and Zephyr, and Nicco, and Clay, Joe, Zack, Garrett, Michael, and so many others. Knowing who was there — not just some nameless, faceless campaign staffers — made contributing to the campaign so much easier. The technology — in particular, the weblog — simply made this simpler and more powerful.

Final lessons for a campaign:

1. Give up control. Kevin Conlon challenged me on this, as he doesn’t think that giving up all control is a good thing. Kevin has far more experience on this than I — but I think there’s a difference between discipline and control. In Iowa, Kerry (and Edwards, for that matter), had far more discipline. Dean’s campaign had spun wildly out of control by this point: rather than stay on message, Dean’s campaign had gone negative in the final week (against Kerry and Edwards), viciously fought with the Gephardt campaign, and had 3500 volunteers who were overwhelming Iowans with phone calls they’d long since grown tired of. Nevertheless, the notion of going from January, 2003 to December, 2003, for me is that they never would have been in the position to lose had it not been for the decision to cede much of the control of the campaign to the grassroots on Day 1.


2. Be transparent. Let people in, let them see how the process works. The minute we understood how to get involved — by simply posting comments — many people were hooked for good. Many had never felt that connected to a political process at any level.


3. Encourage creativity. While you always risk one of these, you more than make up for that with genuine, passionate contributions. Once people feel like their contributions are valuable, they are no longer vulnerable to other campaigns or other messages. At that point their contribution ceases to be about the candidate, it’s about them and their input.

1. Listen. After you’ve given up control and encouraged the creative juices of your supporters, you damned well better listen. When the campaign updated posters to reflect regions that weren’t included (What about Ex-Pats for Dean?!), it demonstrated that this campaign was a conversation, not a broadcast. That was not only refreshing, it was unprecedented in modern presidential politics.

So… what went wrong? I don’t think the technology was much of a factor either way. I stand by what I said right after the loss in Iowa. I think there are a few key items that will ultimately prove to have been the end of the campaign:

• Mismanagement of the Gore endorsement. Trippi was kept out of the loop on this — still not clear to me why. That’s not to say that Trippi would have handled it any different, though the fact that the campaign manager isn’t in on the biggest political endorsement in the last several elections speaks about much larger issues (see below). The point is that at this moment, the campaign tried to bridge the insurgent campaign with the front-runner campaign. And it failed. Miserably. If, as Trippi has claimed, the campaign knew that they were in trouble by December, this endorsement had to be put off or managed more closely. So what if Gore wanted back in the game? Let him into the decision-making process and help him find the right answer.


• Campaign factions. That at least two major factions developed within the Dean campaign speaks to a fundamental failure in Dean’s role as a manager to fix the systemic problem that ended up contributing to his downfall. If he’d wanted to win, Dean had to nip this thing in the bud — whether that meant easing Trippi out of his operational role much earlier (probably in November), or having a cage match between the interested parties to find a solution. Either way, this was Dean’s problem to fix, and he punted.

• Isolation. Dean’s authenticity stemmed from his lack of poll-tested bromides. He said what he felt, and it resonated with people. Yet as he got later and later into the race, he demonstrated a propensity for saying not only the wrong thing (which could be excused as a gaffe) but as saying the thing that reinforced others’ opinions of him, and made it that much easier to define him. For me, it wasn’t the Saddam Hussein “but we’re no safer” statement that killed him (though the polling numbers sure do seem to indicate that was fatal. For me, it was the moment he told the older guy in the audience to sit down and listen to him. There’s self-confident and aggressive (good qualities in a President) and then there’s annoyed self-important (neither is good in a President). He didn’t have people around him coaching him through what was expected of him by the casual voter — the passionate ones were either hooked or not — and as a result his delivery struck many as tone-deaf.

On the balance, I finally have a little bit of distance from the experience to say without reservation how proud I am of what we did. We gave the Party a voice, we showed legions of voters that they can make a big difference, and we showed the other candidates how to stand up for Democratic ideals. It remains to be seen what role Howard Dean will play down the road — but he seems committed to changing the party for the better, and that’s important. He showed thousands of people how to take responsibility for being a citizen again. He let us know that Democracy and Freedom require Action.

I’m terribly disappointed that we didn’t finish what we started. I still feel that as late as December it was ours to lose. Could it have turned out differently? Hard to say. Kerry ran a hell of a campaign. But to think of the assets we had at our disposal — and that we squandered the lead — it’s sad.

Dean accomplisehd what he accomplished while incubating a campaign that learned to use technology in ways it hadn’t been used before. For that, I’m grateful. I think we will see candidates replicate his model. It will start at the local levels — and may be another 4 or 8 years before we see another presidential candidate try and replicate Dean’s model. (If only to avoid the tag of being the “next Dean” which would be code for, “Will scream his way into oblivion.”) But I think the decentralized model — risks and all — will become a model for future campaigns. I sure hope so, because if I’m right, it means we all get a larger say in how our country is run. And that’s what it’s all about.