Software is eating the world. As Iowa’s failed election app shows, that can sometimes be terrifying. The mobile app was supposed to make reporting results smoother. Instead, the app failed on multiple levels, resulting in “inconsistencies” and widespread fears of election hacking.
At Fox News, the schadenfreude is real. For those hoping the first Democratic primary election in Iowa would be a step towards restoring faith in American democracy, the failure of this little piece of enterprise software is an absolute nightmare.
No matter where you fall on the political spectrum, the Iowa app offers case study in how not to launch enterprise software. And for those of us who work with alumni software, the parallels and lessons are painfully real.
The Iowa app was launched to around 1,600 users (precinct leaders). Alumni software providers like Firsthand, iModules, or Blackbaud operate at similar levels of active users per school. The demographics of these precinct leaders are also similar to those targeted by alumni software: college educated, volunteers, and mostly not in their 20s or 30s. The Iowa app’s failure therefore offers key lessons for engaging alumni with software:
1. Failed software erodes trust
This one is obvious. Heads will roll after the Iowa debacle. Those who built the app, championed it, supported it, and consented to it have lost face and will struggle to promote another solution like this again. Similarly, users who were burnt once, will be shy about trying software like this again.
Lesson for alumni leaders: Build experience and trust on a smaller scale before launching something to your entire alumni community.
2. Keep it simple stupid
Whenever you introduce technology, you are asking people to change their behavior. And if you ask for too much of a change, it’s not going to happen. The Iowa app required that all precinct leaders find and download the app on their mobile phones. Organizers then asked them to test it with a test pin, then enter a real pin on election day and then enter the results manually into the app while also entering a unique set of security codes. Reports indicate that before the app even had a chance to confuse people with its user interface or crash, most precinct leaders didn’t even install the app.
Lesson for alumni leaders: Look for software that is human: easy to adopt and easy to use. Mobile apps sound great, but can be a barrier to wide-spread adoption.
3. Communicate your launch early, clearly, and often
It’s not clear everyone got the memo on the Iowa app or if they even read it. Those who did, reported getting a “final app instructions” email just hours before the election that with a confusing set of bullet points. According to the NY Times one bullet point advised users that the “test pin you’ve been practicing with is no longer available.” Another guided users on what to do “If the app stalls/freezes/locks up.” “Close out of the app and log back in with your PIN,” the email read. “The app should save where you were.”
Lesson for alumni leaders: You can’t rush adoption or skimp on the frequency or clarity of instructions when asking alumni to try something new.
4. Train your admins
It’s clear that precinct leaders were largely unaware of how to work the app.They hadn’t received proper training and they hadn’t been forced to test it. Based on this alone, the failure of the app was preordained.
Lesson for alumni leaders: Train your admins. Figure out who needs to learn what and ensure that they learn it.
5. Test it with end users
“Inconsistencies” in results brought many people up in arms about whether the app was sufficiently tested for cybersecurity threats. What’s abundantly clear however is that the app never went through proper end-user testing. Reportedly, users struggled with both the user interface and the app crashing.
Lesson for alumni leaders: Get a panel of actual end-users to test your solution and identify usability, reliability and functionality issues before launch.
6. Make sure it works at scale
Anyone and their uncle can build a mobile app today. While an app may seem quick and responsive in a demo, having that app perform with even one hundred concurrent users requires considerably more sophisticated engineering. The Iowa app needed to handle over 1,600 concurrent users. Reports of it crashing, when so many hadn’t even downloaded the app, tells us it was nowhere near handling that kind of load.
Lesson for alumni leaders: It’s hard as a client to stress test your software at scale, so look to providers and apps that have a track-record of delivering software at scale.
7. Be flexible, provide support
The fallback plan for the app failing was that precinct leaders would do it the old way and phone in the results. But the phone lines didn’t work like they used to and couldn’t handle the volume. Multiple callers dialing in for one precinct could be on hold for an hour before their call dropped.
Lesson for alumni leaders: Give alumni an old way and a new way of doing things. Whenever possible, give them a human to speak with.
The Iowa voting app failed epically, putting an enduring stain on the Iowa Caucuses and weakening people’s trust in bringing technology to elections. The stakes in launching apps and initiatives to hundreds and thousands of alumni are also high, but if we can learn the right lessons, it doesn’t have to be a game of Russian roulette.