Democrats Won't Necessarily Have A Tech Lead in 2016, Says Former Obama Campaign CTO
BY Sarah Lai Stirland | Tuesday, March 26 2013
Presidential candidates on both sides of the partisan aisle will face similarly difficult challenges when building their technology infrastructure in the 2016 election cycle, despite the advantages that President Barack Obama's campaign enjoyed in 2012, predicted Harper Reed, the Obama 2012 campaign's chief technology officer.
"I actually worry about this for both the Democrats and Republicans," Reed said at Dent 2013, a new conference on the process of innovation organized by two Seattle entrepreneurs and held Monday in Sun Valley, Idaho. "They are going to be in the primaries. They are going to have the same layout as the Republicans. I don’t think there is going to be much of a difference between the two parties in that regard."
On stage at the conference, I asked Reed whether he thinks Republicans will be able to catch up on the tech front in 2016 after the GOP resolved, in a recently released, 100-page "autopsy" report, that the best things to do in 2016 are all of the things Obama's did in 2012.
As CTO for the campaign, Reed said, he took the philosophy that campaigns need to "start from scratch" on the technology front and work within the technology environment of their present rather than stay in the frameworks of the past.
That means that despite all the infrastructure that was built between 2011 and 2012, the Democrats will have to begin again in 2016 — just like the Republicans. And because Democrats won't have an incumbent, the eventual nominee will have to face the same time constraints that bedeviled Republican contender Mitt Romney, which, Reed said, is a problem for a team hoping to build reliable tools.
"My concern with all of this is that this requires an investment of technology that starts at zero, and the Obama for America campaign committed to that investment," he said. "I was hired at the end of April 2011. The campaign was founded in the beginning of April, and so I had 18 months to do my job."
"A couple of things could happen: an outside organization could start building, but that hasn’t necessarily worked in the past," he said. "And these products are so specific to the candidate. For instance, Barack Obama is very big on a specific kind of field organizing, which [former National Field Director] Jeremy Bird and [former battleground states director] Mitch Stewart do, so our software represented that. You can’t use that software if you don’t believe in that. We didn’t win because our software was good. That’s true. If people think that, then they’re obviously not looking at the fact that we got more people to vote.”
Speaking directly of the Republican Party, whose spokespeople have announced a GOP search for a chief technology officer, possibly a Silicon Valley political outsider, Reed said:
I think the idea that you need a technology CTO, preferably someone with a beard who is going to come in and disrupt things, which is the thread that is coming out of a lot of these conversations, is wrong. I think it should be – you need to invest in these technologies very early, and in a very global way. And I wonder how that is going to manifest itself in the primaries, and I’m very worried that that won’t work out.
And implying that the Republicans simply don't understand what's needed for success, he joked: "I’m really worried that they’ll just end up looking at the transition between [former Chief Innovation and Integration Officer Michael] Slaby and myself and they’ll just hire a homeless person."
Dent is a conference designed to examine the processes of how innovators achieve the results that they do. Reed, ironically, insisted that the Obama campaign didn't innovate — something other Chicago tech staffers told us last year.
“It’s not technology innovation. We just executed,” Reed said.
But that is perhaps what the breakthrough was: Reed and his team's iterative process to keep shipping software out as quickly as possible, testing for weaknesses, improving the process, and creating backup plans and procedures to deal with potential failures.
Reed said that the way the campaign architected its online contribution system is a good example of that approach. Other members of the campaign's team have said in the past that at peak moments, the campaign saw $3 million flow into its coffers an hour.
“So a great example is our contribution system, which a lot of people gave money through, didn’t require a database connection. So if the database went down, it was fine. And that was something that’s not necessarily obvious at first when you’re building something like this. You had to get to that point," Reed said.
This philosophy of making everything foolproof also explained why the campaign built so much of its infrastructure in-house.
"We looked at trying to buy every single application we possibly could find, because we didn’t have time to build it," Reed said. "We just didn’t find anything that could stand up to the traffic we were getting. An example is our call tool. Our call tool got about 10,000 users a week for about 17 months, and then for the last four days, that the call tool did hundreds and hundreds of thousands of simultaneous users, and then in the last day, did about a million total calls," he said. "All the vendors said, yeah, we can do that, but I didn’t believe that."
Reed also said that the campaign built its own tools, such as its Facebook app, because he didn't trust outside bipartisan vendors, such as Votizen.
"Specifically with Votizen, there was an interesting part of Votizen where they chose to say that they were non partisan, which means that the campaign can’t trust them," he said.
Reed said that he worried over the security of the campaign's data in Votizen's databases, and also about benefitting Republicans with any lessons learned that Votizen might share with Republicans based on the Obama campaign's usage of the app.
"Why would you trust them? They could give your data to someone else," he said.
Reed also dismissed the idea of open-sourcing the campaign's software from the 2012 cycle, saying that even with tweaks, it wouldn't be relevant in 2016 since technology evolves so fast.
"A great example of this is Parse, a startup in Silicon Valley," he said. "They’ve created a backbone for mobile apps. They’re much more proven now than in 2011, and so maybe we’d use them now rather than when we passed in 2011. So as more of these things become mature, maybe we’d start using them.
"The thread that ‘why don’t we just open source everything, it saves everyone’ is false because the people who are saying that just don’t know about software."