Through a partnership with non-profit Democracy Works, I led a team of researchers and designers at the SAP Design & Co-Innovation Center to help address low turnout rates among America's young, first-time voters. We used a hybrid research approach to find key behavioral trends that influence whether young voters will vote or not, and we provided design recommendations to address each one, ultimately helping push more people to the polls.

Democracy Works

Team Lead, Lead Researcher and Designer


United States


Young Americans between ages 18-24 vote at rates significantly lower than other age groups. And, with 10.7 million more eligible voters in 2016 than 2012, it is clear that this segment of the population is growing in, both, size and influence. Democracy Works, a New York-based non-partisan nonprofit dedicated to improving voter turnout in the U.S., understood the unusually high stakes of the 2016 election, but lacked the research capacity to understand the challenges first-time voters face. The organization was also in need of design resources to translate insights into their flagship product, TurboVote, an online tool intended to help users navigate the voting process.


Our exploratory research was centered around meeting with young people in their natural environments – be it their homes, libraries, fraternities, or their favorite coffee shops – and allowing them to steer the discussion, a process that allows for broader exploration and discovery. When meeting with a female participant in her dorm building, for example, we observed TV rooms in which residents watched debates and discussed issues. We discovered this to be an important touchpoint to reach young voters where they already feel engaged in political discourse.

Contextual observations like stickers on participants’ laptops, or background images on their phones, also helped unearth additional insights. For instance, while most respondents of mainstream political leanings felt comfortable publicly showcasing their views to the world, participants with less prevailing opinions were more hesitant to do so. Lastly, we used semi-structured ethnographic interviews, complemented by co-creation activities and prompts, which helped uncover instances in which participants’ stated opinions were, intentionally or not, different from what they thought. For example, when asked how they would go about convincing a friend to register to vote, we noticed that respondents tended to rush and state the first thing that came to mind. When we gave them a written prompt instead, they felt that they were allowed time to think and write down their answer, without feeling the urge to promptly fill the silence.

  • gallery-image
  • gallery-image
  • gallery-image
  • gallery-image


We conducted product-level usability testing with existing and new TurboVote users, who were tasked with performing actions, such as onboarding and sign-up, online voter registration, registration address change, and sharing TurboVote with friends. This allowed us to assess how well TurboVote was designed to meet the needs of our target user group. For example, the TurboVote app filters information based on a user's location. However, our research revealed that it could be difficult for this particularly transient demographic to have a single permanent address in the span of a few years. Most participants resorted to entering their parents' address instead, which prevented them from voting in local elections that concerned them more directly. This led us to consider a system of proactive address update notifications synced with the school calendar, when students are most likely to change addresses.


By looking at other services of a similar nature, we discovered several unique attributes TurboVote could better communicate to users. Unlike other tools that mostly help users register to vote, TurboVote also follows the user around, sending notifications via text or email containing important information about elections occurring in their area. Despite the underlying value proposition of such a service, our research revealed that users did not clearly understand this aspect of TurboVote, prompting us to make these distinct offerings clearer.


We quickly realized that even the most motivated young citizens fail to vote when there are not timely triggers. We found that the strongest trigger to register and cast a vote, is when the nudge comes from someone a young voter personally knows and trusts. This led us to rethink the way Democracy Works designs their outreach programs, particularly those aimed at universities and colleges (e.g. instead of a university dean sending mass emails to students, using a more personalized peer-to-peer communications model). We also made it easier for users to share TurboVote with personal connections through social media and email-share buttons with pre-populated, yet personable content. This enabled users to recruit their friends to the platform, subsequently offloading some of Democracy Works' user acquisition work to their most active users.


Even though young people recognize its inherent bias, most political content that they consume and share exists online and on social media. This setting, where potential voters are most comfortable, where they are already having conversations about politics, where they feel like their most political or civic selves, and where they are more susceptible to political messaging, is the perfect environment to transform enthusiasts into actual voters. This led us to consider TurboVote's partnership model, and to push the organization to participate in those online conversations by strengthening partnerships with popular media entities, like Buzzfeed and Snapchat, and also by aggregating, or creating, native content on key social channels. This proved to be an effective model.


While young citizens found voter registration to be relatively easy, they often struggled with the before and after, feeling lost in the process. We worked on design changes that would give users a better sense of place and help walk them through the process with clear steps, by providing feedback along the way, and by sharing information to push users across that "last mile" to the voting booth. This meant redesigning the TurboVote sign-up and voter registration flows, adding progress indication bars, and changing the language of the end screen to help people feel that they've accomplished something important. In addition, we designed a virtual "I registered with TurboVote" badge that could be shared on social networks and passed to friends, which tapped into behavioral studies about people "voting to tell others," as demonstrated by the success of the "I voted" stickers in the physical realm.


While Democracy Works already understood the value of in-depth design research, they lacked the internal capacity to carry out this type of work on their own. We stepped in as a de facto extension to their team at a critical time leading up to the 2016 presidential election. Since the engagement, Democracy Works has continued to use our deliverables to inform their roadmap for 2017, and beyond. Additionally, our approach has helped the organization think about Turbovote in a new way: not as a product, but as a service. We challenged Democracy Works to think about where TurboVote begins and where it ends within the overall voting and political engagement process, resulting in conversations about user acquisition, service touchpoints, and how to tap into TurboVote's extensive network of partners.


The timeline of this engagement was set up to allow for the implementation of prioritized product changes immediately preceding National Voter Registration Day. Our design changes contributed to meaningful engagement results (3x higher completion rates), as well as a higher number of voters served, with over 500,000 additional TurboVote subscribers in 2016 alone. This meant that the platform served over 1 million voters during the 2016 election, compared to 150,000 four years prior—a dramatic increase in TurboVote's reach.

I would be remiss if I ended this case study without mentioning that any outcomes of this project were diminished by the overwhelming anguish and dejection I personally felt over the results of the 2016 election. While subsequent exit polls showed encouraging signs in young voter turnout, they pale in comparison to the impact the election has had, since.