For NO PARTICULAR REASON, I want to share a story about Democratic Party technology succeeding. The failures tend to be loud, but the successes tend to be quiet, so I’d encourage you to think about what you might be missing
Here’s a great blog post about some successes from 2018! medium.com/democratictech… One of the things it mentions is Voter Protection. Here’s a story about the tool used to run the programs mentioned in that post
This is obvious, but preventing voter suppression is super important! One big way you do that is by sending out polling place observers. They watch for voter intimidation and they also make sure someone knows a bear is invading a polling place (this actually happened in Virginia)
Picture tens of thousands of volunteers out in the field for a national election dealing with issues in real time, making sure votes get counted and also collecting info and archiving it to be used in recounts and lawsuits about voter suppression
This is so important that by 2016 the party had built a tool to manage such programs three times and every time it went down hard on election day. Okay, even this success story has some cockups in it
The Hillary for America (HFA) tech team heard this and said, we are going to build a version of this that does not go down on election day. And they engineered the shit out of it and it stayed up on election day
Actually, first they asked, can we get away with not building something? Can we use google forms? Can we use zendesk? But there were requirements that didn’t fit. And the pricing for most tools makes no sense for something with a lot of users that runs about once a year
Anyway, when the campaign ended, the team disbanded and that app got shut down along with all the other HFA tools. Also a key vendor it depended on went kaboom. Those things would have happened win or lose
The code itself made its way to the DNC along with everything else built at the campaign. In case you do not know, a giant pile of non-running code dependent on services that no longer exist and none of the engineers who worked on it is not a glorious bounty
A lot of people have this idea that the DNC is a big huge powerful organization that could take in something like that. It is not. I have worked on single feature teams that were larger than the entire DNC engineering team in 2017. The team is bigger now, but not that much
I joined the DNC mid-2017. A few months later, the Voter Protection team came to us and said they wanted to run a program in VA and were going to hire someone to replace that useless app from 2016. We said: Over our dead bodies! We have something that stays up on election day!
And we also said, why is it useless? It turned out there were some very solvable usability things in some workflows. Not reason to toss it out, but makes sense to think you should if you mostly deal with vendors and aren’t used to thinking about software as an iterative thing
So we invested in it: got it back up and running, replaced the vanished vendor dependencies, learned all sorts of details about how the thing even worked and how to run it, removed all the hard coded references to 2016
Then we ran it in VA in 2017. That election came down to a literal coin toss for control of the VA house. We lost the coin toss, but it was only even that close because of votes we picked up running a voter protection program with that tool
There were a few things that we had trouble with in VA, notably our login system wasn’t simple enough. So we made it simpler, improved a few more things and ran it again a month later in Alabama. Then we did some more infrastructural work for 2018. It worked!
This story is kind of boring. It’s supposed to be kind of boring. We built some software following best practices around horizontal scaling and we improved it by listening to our users. You don’t hear much about the stuff that works because it’s not news
And yeah, politics is a hard environment to work in and there are a lot of problems software won’t solve. A lot of the incentives are bad and the way election cycles work screws up the economics. But success is possible when we think long term and invest in our institutions
PS do you have a progressive tech success story? I would love to share it!
