Breaking Bad – The Iowa Democratic Party’s failure to transform mobile voting
What happened during the Iowa Caucus? How did trying to create a simple app for voting create so many issues for the Iowa Democratic Party?
Sometimes when digital transformation fails, it can fly under the radar with companies hiding behind the veil of vague press releases while the company’s employees, suppliers and partners suffer the consequences of a failed transformation. However, other times when digital transformation fails, it fails big, with a loud public thud and people wondering, how did this fail so hard?
A recent public example of a failure in digital transformation occurred during the Iowa Democratic caucus on February 3, 2020. The Iowa Democratic Party, in its attempt to modernize their tabulating and reporting process, asked voters to download a new app and submit their preferences via mobile. Unfortunately, several technical issues occurred during the voting process leading to frustrated users, challenges in reporting data and ultimately a delay in and confusion in the results. More broadly, many people also began to question the party’s ability to use digital technology effectively for the upcoming 2020 election. While not a traditional corporation, I believe that the Democratic National Party (DNP) and in close association, the Iowa Democratic Party, faced many of the same pressures that organizations encounter today when looking to modernize their processes in a competitive market.
First, the motivation behind why an organization is pushing digital transformation is important as it feeds into how the organization will decide its priorities, allocate resources, and unify the rest of the organization for change. According to the NYTimes, after the 2016 election, there was a broader push by Democrats to invest money in digital to match the Republican party’s efforts in digital advertising and campaigning.1 Said another way, the Democrats realized they were far behind their competitor (the Republican Party) in using technology to reach voters, advertise their campaigns, and record data and needed to “catch up”. While I believe it’s necessary for organizations to monitor and respond to what their competitors are doing, it is also critical to ensure that the organization is strategic around how and what it’s planning to transform. In this case, the DNP recognized they needed to “get smart” on digital, but were not focused on what projects to invest in and which firms to hire for these projects. When the DNP hired Shadow, they brought many different projects including building a centralized data base, texting technology, and digital advertising consulting. While its admirable that the DNP identified the many areas in which technology could help improve its current process, it failed in prioritizing which ones were most important given the timing and the capital available. As a result, none of the technology was well built, and Shadow even lost some of its Democratic clients such as Joe Biden who stated the texting technology “did not pass our cybersecurity checklist.”1
As alluded to above, a major issue that the DNP faced was hiring a firm that lacked the organizational capabilities to do the job required. Shadow, previously known as Groundbase, was founded by former Clinton 2016 digital campaign staffers. While it’s easy to speculate why Shadow was chosen for the job, I believe that many organizations undergoing digital transformation face the issue of undeveloped capabilities and lack of engineering resources. First, there is the difficulty of getting top quality talent to “non-tech” firms. According to the NYTimes,
“Shadow was also handicapped by its own lack of coding know-how, according to people familiar with the company. Few of its employees had worked on major tech projects, and many of its engineers were relatively inexperienced.”
For a project as important as building out an application for voting purposes, Shadow was not equipped with the appropriate individuals to complete the job and the DNP did not have the right individuals to manage the projects well. Even if Shadow had been well-resourced, the task of building, testing, and securing the application in two months would have already been a major feat. Given the poor resources and rushed timing, the application development period was squeezed so tightly that there was no time to get the app approved by the Apple Store. Voters who wanted to download the app were required to bypass the phone’s security setting and received multiple errors and warnings the app may not be secure.
While it’s easy to point fingers at Shadow and the app developers, I believe that the Iowa Democratic Party and the DNP more broadly are to blame for this failure. In their efforts to digitally transform and move away from the phone based caucus results, they lost their top priorities which should have been data accuracy and security from potential bad actors and were unable to pull the plug when there were signs that the application was not ready for use.
Given the rushed timeline, Shadow was still working on the app the weekend before the caucus and unable to perform the appropriate amount of testing which may have revealed the coding issue. Before the caucus, precinct chairs also reported having problems with downloading the app and received warnings that the app might not be secure as it did not come from the traditional app store (as noted above). This initial feedback should have tipped of the party that perhaps the app would have issues on the day of the caucus (scheduled for a days later). The party could have moved to the back-up phone system given the importance of the accuracy of the results. Furthermore, with security being a key focus in voting, an assessment of the app done by a third party cybersecurity company post caucus revealed that the app had “several problems within the code including links to people’s personal websites” and that the app was “easy to obtain which means anyone could access the infrastructure supporting it and cause damage.” Given the heightened background of cybersecurity in the United States, the party’s priority should have been to ensure the security and legitimacy of the app. While at this point, no security breach has been found, the failure to consider the major risks in security and accuracy of the app – the most important priorities for voting for the caucus – reveal a deviation from the organization’s priorities driven by the thirst for digital innovation.
Changing User/ Customer Behavior
In looking to integrate a new technology into an organization, it’s important for the users to be appropriately trained. Often, this requires changing user behavior to work alongside the digital technology or developing the digital technology so that it easily integrates into a user’s life. At the Iowa caucus, volunteers at the Iowa precincts were not trained in how to use the app and were only provided complicated written instructions for how to download the app. Furthermore, the downloading process required two-step authentication and PIN passcodes which further complicated the process. Considering that many of the caucus chairs were older and albeit less tech-savvy, it would have been better if the Iowa Democratic Party had provided some training or seminar on how to use the app or even a technical support hotline for the chairs to use when they faced errors and issues. There was a failure in understanding the user base and in providing appropriate resources for the chairs to enable and empower them to use the app correctly.
Overall, the DNP still has many opportunities to explore digital transformation as developing the right digital platforms and technologies could prove to be highly valuable for the organization. However, by falling into the pitfalls of not defining a clear digital strategy, lacking the right organizational capabilities, failing to quantify risks, and not spending time to learn about or educate their users, the Iowa Democratic Party destroyed value in one of the more important preliminary events for the 2020 presidential election.
Student comments on Breaking Bad – The Iowa Democratic Party’s failure to transform mobile voting
Really good insights into what and why things went wrong here. As someone coming from tech, I’m completely shocked at the DNP’s willingness to hire an untested firm; it makes me wonder if the process was based more on political connections (Clinton’s 2016 digital campaign) as is common in politics rather than merit/past success as is common in tech. I understand the pressure the DNP feels to catch up – but isn’t it worse to rush things that will end badly than to invest the time and energy to do it right? How could the pressure have mounted to the this fevered “do anything to catch up” state? I also wonder, did it have to be THIS app? Couldn’t there have been another way to integrate technology that would have shown the DNP’s willingness to engage with users digitally? Also, where was the plan B?? So many questions that I would love to have the DNP answer.
It’s incredible how lack of a clear digital strategy AND capabilities can lead to such a massive failure, a failure that as you pointed out could have been quiet had it not been for this being one of the most important preliminary events of the 2020 election.
I also see this as a capability question. The DNP may have had the money, but they should have recognized that they did not have the capability to hire a fire with the capability to build this platform in such a short time. They should have hired a firm to hire a firm for them. First of all, such a firm does not exist- and secondly, if the firm existed, why did they not have evidence of having built something similarly impressive in the past.
CEO’s or organization leaders have to time digital transformation so that the cost of failure is minimized. In this case, DNP went for a time of maximized cost of failure. Couldn’t this have waited for a few more years until the stakes were not as high?
The last point that I would like to agree with is the lack of a clear protocol on how to pull the plug. When things are going south, many people know it- but it’s never clear who will bell the cat. Without pre-defined ways to respond to failures of different scales, everyone is left to improvise- and people resort to the instinct to fingerpoint, as opposed to building solutions.
Awesome topic! While it’s indisputable that the DNC was a “loser” in this situation, I do think its push toward a more-efficient voting procedures is a step in the right direction. The questions in my mind after reading your post are if and how digital technology (especially mobile) can be used to make voting more accessible – and if it is possible, who should be leading the charge?
To the last questions, it seems like it should be a government-wide initiative vs. one done by individual parties. I think this would ensure more rigorous testing, increase trust, and enable a country-wide roll out with sufficient training / advertising provided to all citizens (e.g. the HealthCare.gov rollout). As to whether mobile voting of some sort is even possible – it seems like a long way off for the business case and data privacy concerns to be addressed but maybe this is something that should begin to be piloted at a state level?