There's one other interesting component to the campaign's structure. And that's the presence of two big tech vendors interfacing with the various teams--Blue State Digital and NGP Van. The most obvious is the firm that Rospars, Jascha Franklin-Hodge, and Clay Johnson cofounded, Blue State Digital. They're the preeminent progressive digital agency, and a decent chunk--maybe 30 percent--of their business comes from providing technology to campaigns. Of course, BSD's biggest client was the Obama campaign and has been for some time. BSD and Obama for America were and are so deeply enmeshed, it would be difficult to say where one ended and the other began. After all, both Goff and Rospars, the company's principals, were paid staffers of the Obama campaign. And yet between 2008 and 2012, BSD was purchased by WPP, one of the largest ad agencies in the world. What had been an obviously progressive organization was now owned by a huge conglomerate and had clients that weren't other Democratic politicians.
One other thing to know about Rospars, specifically: "He's the Karl Rove of the Internet," someone who knows him very well told me. What Rove was to direct mail--the undisputed king of the medium--Rospars is to e-mail. He and Goff are the brains behind Obama's unprecedented online fundraising efforts. They know what they were doing and had proven that time and again.
The complex relationship between BSD and the Obama campaign adds another dimension to the introduction of an inside team of technologists. If all campaigns started bringing their technology in house, perhaps BSD's tech business would begin to seem less attractive, particularly if many of the tools that such an inside team created were developed as open source products.
So, perhaps the tech team was bound to butt heads with Rospars's digital squad. Slaby would note, too, that the organizational styles of the two operations were very different. "Campaigns aren't traditionally that collaborative," he said. "Departments tend to be freestanding. They are organized kind of like disaster response--siloed and super hierarchical so that things can move very quickly."
Looking at it all from the outside, both the digital and tech teams had really good, mission-oriented reasons for wanting their way to carry the day. The tech team could say, "Hey, we've done this kind of tech before at larger scale and with more stability than you've ever had. Let us do this." And the digital team could say, "Yeah, well, we elected the president and we know how to win, regardless of the technology stack. Just make what we ask for."
The way that the conflict played out was over things like the user experience on the website. Jason Kunesh was the director of UX for the tech team. He had many years of consulting under his belt for big and small companies like Microsoft and LeapFrog. He, too, from an industry perspective knew what he was doing. So, he ran some user interrupt tests on the website to determine how people were experiencing www.barackobama.com. What he found was that the website wasn't even trying to make a go at persuading voters. Rather, everyone got funneled into the fundraising "trap." When he raised that issue with Goff and Rospars, he got a response that I imagine was something like, "Duh. Now STFU," but perhaps in more words. And from the Goff/Rospars perspective, think about it: the system they'd developed could raise $3 million *from a single email.* The sorts of moves they had learned how to make had made a difference of tens, if not hundreds of millions of dollars. Why was this Kunesh guy coming around trying to tell them how to run a campaign?
From Kunesh's perspective, though, there was no reason to think that you had to run this campaign the same as you did the last one. The outsider status that the team both adopted and had applied to them gave them the right to question previous practices.
Tech sometimes had difficulty building what the Field team, a hallowed group within the campaign's world, wanted. Most of that related to the way that they launched Dashboard, the online outreach tool. If you look at Dashboard at the end of the campaign, you see a beautifully polished product that let you volunteer any way you wanted. It's secure and intuitive and had tremendously good uptime as the campaign drew to a close.
But that wasn't how the first version of Dashboard looked.
The tech team's plan was to roll out version 1 with a limited feature set, iterate, roll out version 2, iterate, and so on and so forth until the software was complete and bulletproof. Per Kunesh's telling, the Field people were used to software that looked complete but that was unreliable under the hood. It looked as if you could the things you needed to do, but the software would keep falling down and getting patched, falling down and getting patched, all the way through a campaign. The tech team did not want that. They might be slower, but they were going to build solid products.