1:07 p.m. October 26, 2012

You Didn't Code That

From radio between the wars, to television in the 60s, and video discs in the 80s, the political campaign that can figure out how best to use the day's technology has always enjoyed a significant advantage.

No one knows in advance when a particular technology's time has come – Howard Dean was ahead of the curve in online engagement in 2004, but social media was critical to the Obama campaign just four years later. In 2012 both campaigns are more than happy to pour money in to try to find that elusive technological edge. But how to spend that money?

Romney Director of Digital, Zac Moffatt, is banking on big companies like IBM providing "the best of the best", doing lots of that outsourcing that big companies find so sexy, and ridicules Obama for America as thinking "that they have the only people that understand big data and social media". To hear Moffatt speak, the Chicago campaign is a buzzing hive of hubris.

The real difference is that, while Romney relies on traditional technology players, Obama prides itself on a small, scrappy team, such as has spawned many Silicon Valley darlings such as Facebook, Twitter and Amazon. In those circles, in fact, hubris is often considered a virtue (along with laziness and impatience, but try as I might I can't turn them into a political metaphor).

Certainly Team Obama is owning the hubris: much of their digital strategy being a continuing evolution of the ideas that didn't help Dean in 2004. Further, the team's digital attitude – developers, data analysts, Uncle Tom Cobley and all working under one roof – comes across very much like a cocky Californian startup, complete with amusing error pages on their website and a hipster CTO who ironically asserts that he's "probably one of the coolest guys ever". Or maybe it's post-ironic; I've never really understood hipsters.

Sign up online as a volunteer for the Obama campaign and you'll see their vision of grassroots politics online, known as Dashboard. It's basically a web haven for Obama nerds, helping volunteers join local teams, find events to participate in and giving access to online training videos, grassroots funding tools and a nifty phone banking system that allows the central campaign organizers to mold their volunteers to fight the important battles. If your neighborhood is already pro-Obama, no point in calling everyone to persuade them of the evils of Paul Ryan's budget–but maybe you'll be asked to phone into a battleground state instead.

Dashboard is a startup in almost every detail, from the occasional bugs balanced with regular updates, and an apologetic message when things go wrong showing the president talking to a pirate. It's even got stats porn, telling you how many phone calls you've made or how much money banked for the campaign. And yet it's still a curiously one-way street, relatively easy to find things to do for the campaign, but lacking any particularly social feeling–if you want that then go to a real life event and hang out with other Democrats seems to be the message.

The Romney campaign provides a lot of the same tools, of course, but many more are off-the-shelf with some (usually limited) customization. The one difference that struck me most was their phone banking tool, a platform called FLS Connect. All the functions of the Obama tool are present, but where Dashboard's call sheet feels crisp and integrated -- the kind of thing you might expect to come out of Google -- FLS Connect is cookie-cutter, with phrases like “the campaign” suggesting its use in hundreds of races at once.

Even their introductory video reminded me of nothing so much as air safety films from the 1990s, a generic screencast that showed how to use the system without betraying either humanity or any association with Mitt Romney. (Admittedly a problem with many Republicans these days.) By contrast, the Obama campaign has training videos hosted by a perky and mediagenic campaign staffer green screened over a bustling campaign office.

Maybe it'll turn out that a chirpy web interface with funky neighborhood team names will shift the balance of the election against a morass of different tools that feel more like an office fire drill combined with emergency root canal surgery. I doubt it, though. Anyway, the most important work in both campaigns' digital work is below the surface.

Remember Zac Moffatt's line about how the Obama campaign thinks they alone understand “big data”? You just knew I'd come back to that, right?

When I first read that quote, my finger started itching to flick through my Rolodex of Obscure Industries for someone who could help me make sense of it all. At the time I was having dinner with NSFWCORP Editor-in-Chief Paul Carr, which would've made that awkward (staff Rolodex are outlawed at the NSFWCORP dinner table, mostly due to the fact that Mark Ames' Rolodex spans twelve volumes), but as soon as I could get away I dialled up Alex Steer, from Big Data consultancy Fabric Worldwide.

"We advise people to own their own data", he begins, rather than just rely on stuff that anyone can rent in, advocating what is termed a "first party" data management platform. Both the campaigns seem to have this covered, from the data that Romney started building up in his 2004 bid and which Slate identified as a key driver in his public commitment to win the Iowa caucus, to the complex and overlapping data sources the Obama campaign is apparently wrangling under the codename "Narwhal".

Mother Jones has an infographic that tries to explain (although the cartoon supporter distractingly makes me think of Daphne Blake from Scooby Doo, prematurely aged by bringing up all Fred's kids while Shaggy lives in the garage).

It's difficult to be certain, because the campaigns aren't giving much away about their big data technologies, but it seems that here again the Obama team has built it, while the Romney campaign has bought it – probably from CMDI, who have billed them for more than $1.5 million so far having previously provided tools around donor and voter lists for the campaigns of three Republican presidents, as well as congressional and statewide races.

As reported by the AP, Team Romney has also been working with at least one specialized agency, Buxton, to mine public data sets for potential donors – something that the Obama campaign, which has raised over half its money in small (less than $200) donations, can presumably better take care of via its grassroots efforts.

All this data allows the campaigns to do some pretty impressive things – keeping track of solid Romney voters in Iowa, figuring out which issues are important to a particular Obama supporter in Ohio – and, increasingly, what to say to an independent voter to change their mind about a candidate. None of this is particularly new as an idea – marketing companies have been doing it for years, and the techniques have been used in political campaigns going back at least as far as the 90s – but the scale is something new, and for many something worrying. The danger is that this starts to feel creepy – the vast majority of people are weirded out by highly targeted political messaging, according to a study earlier this year, far more so than for commercial advertising.

In the year that Charles Duhigg revealed that Target can figure out if you're pregnant without asking the question directly, I was a little surprised that political targeting is considered more intrusive – but it all makes perfect sense says Steer, noting wryly that "nobody thinks that your preference in socks is part of a public conversation".

At least for now politics is still that: public – but with the trends toward micro-targeting of emails, posted mail outs and volunteer door stepping I wonder how long that will last. In 20 years it may even be considered less acceptable to be shown a generic political campaign message: "that means they don't care about me!".

In the meantime, a data mining backlash is a real concern, particularly for a big government party like the Democrats – and even if better targeting means less irrelevant content being shoved at you, as true in politics as it is in advertising. "Just because people don't understand what you're doing doesn't mean their concerns aren't important" points out Steer, "or that they can't hurt you", a view backed up by Google's ongoing wrangling with the European Union over privacy laws; and with increasing public concern over government data privacy (my favorite being the UK government's blunder in gathering sensitive and identifying information about attitudes to online pornography – and then leaking them to others who took the survey) it would only take one widely-publicized mistake to turn a creepy tactic that could give one campaign a much-needed edge into a vilified one that leaves the candidate lying bleeding on the sidewalk.

There will never be a real way of knowing, even after the results are in, whether the lean-and-scruffy, build it here approach is better than the rich-and-happy, buy it in one for any particular problem. But whichever way they choose to go, the intention is the same for all campaigns: find and manage donors better, maintain the base and get out the vote, and seek out issues that can flip key voters and then speak as directly to them as possible.

It may take a few years, but don't be surprised if some time in the 2024 campaign you're watching a political advert when the candidate looks you right in the eye and berates you for not helping out more at your kid's school. Where's your social responsibility? Vote United American Party to restore our sanitation budgets!

If nothing else it gives me hope: with nearly 25 million registered independents, in 10 years' time political campaigns may be the only places still hiring writers.