Orca vs. Narwhale: The Tech Gap

The Atlantic has a fascinating piece on the technology (Narwhale) behind the Obama campaign--how it was created, who built it, and how it came together. The story is basically this:

1. Team Obama brought in world-class technology leadership under the governance of super-nerd Harper Reed. This team included people with data skills, scaling know-how from the likes of Twitter, and other heavy-hitters.
2. They developed rock-solid (if, erm, sometimes late) infrastructure that delivered on several ground breaking promises: near-real-time updates, combining tons of key data to turn out even less-likely voters, and tools to make Obama's massive in-the-field army super efficient.
3. The thought-leadership group ("Digital" vs. the above's "Tech") learned things like which of several different versions of an email would change the take from around 500k to 2MM--and, because of focus testing, only sent the winners. They learned how to target voter-segments based on TV viewing habits making their ads more effective (as well as cheaper as they paid less per ad than Romney).

By contrast, Romney fielded technology by the name of Orca (the only animal, we are told, that eats Narwhales). This was a web-based strike-list which would track names of people who voted and allows volunteers to use the list to contact voters on it who haven't voted so that you can get them to the polls. It crashed pretty badly on the morning of the election. Romney got a lot of blame for the potential impacts of this--including here--but it probably wasn't warranted: even if Orca worked perfectly Romney would still have lost.

What Does This Mean?

The problem here is not that Obama ran a 21st century campaign and Romney ran a 20th century one--it's that Obama was able to get top-end heavy-hitters to take a pay cut, move to Chicago, and build his 22nd century infrastructure while Romney was, more or less, doing what campaigns have always done (run on people, leverage technology ... badly).

Unless there are some seriously ideological super-nerds lurking in the wings the 2016 GOP effort will need to hire talent ... talent which, probably, can't be bought. In other words: the gap isn't in the execution of the technical strategy ... it's in the people. There are things about scaling for one-day-a-year that simply put are not widely known. Yes, some of the top consulting companies can get people with those skills--but they come at an extreme price: there is no life after the campaign, no IPO, etc. After the election you get your last pay check and your credit card is canceled.

What Do I Think?

A lot of people made anti-Obama movies this year. There was Obama 2016, Dreams From My Real Father (which was mailed on CD to 1.5 million people. Alas, I did not get one), Hating Brietbart (not exactly Anti-Obama, but basically so) and Atlas Shrugged Part II which I am pretty sure is not ideologically fond of him.

By contrast HBO produced Game Change with real star power about the 2008 election. It's good, actually--if, perhaps, not factually (who knows? It's certainly more accurate than Dreams From My Real Father, which alleges that Obama got a nose job to look less like his real dad a Marxist black activist). 

The difference is between what Hollywood does and what, I'm afraid, amateurs do. I have not seen the above movies (give me time--I'm waiting for Netflix) but I can assure you, from what I have seen already, that while things like Obama 2016 are probably competent (I imagine I will find it watchable) and I actually quite liked Atlas Shrugged Part I (I understand they needed all new actors for Part II because they couldn't get the originals back) they are not Hollywood grade. I kinda loathe Michael Moore and I bet his efforts take on D'souza's pretty handily.

Simply put, if conservatives are waiting for the silver screen to beat Hollywood at its own game? Don't hold your breath. It appears this may be true--and far, far more devastating--in the technology field as well ... for the same reasons (the ideological bent of nerds and film makers may be pretty baked-in statistically).