Howard Schultz considers running for President
Howard Schultz is the former CEO of Starbucks and was considering a run for President. I joined his team at 1560 LLC as the Lead UX Researcher and Designer. Because his potential run would be as a third-party candidate, we were limited in our tools that most candidates acquire in a two-party system. My focus was building out those internal tools. The biggest project I worked on was leading the research and design for a volunteer app that would help us collect data to create voter models.
Tools & Methods
User journey mapping
Design thinking workshops
Although this product was used mostly on a phone, it was designed as a responsive web app, volunteer leads could help volunteers set up their “turfs.”
When I first got to 1560 LLC I ran into an issue. This was an organization that didn’t have current users, and did not fully know who they wanted to target.
With that in mind, I knew I had to figure out who our users were, or at least come up with a hypothesis to test. The team couldn’t build a strategy until we knew who we were targeting and what they wanted.
My background in UX didn’t make me the strongest candidate for this task. To me this sounded more like a marketing campaign, because we were trying to acquire an audience – in this case a huge percentage of the American population.
My experience in my work has mostly revolved around increasing engagement or loyalty with products. Instead, this was attracting people to a product and then selling them on it. There was also an extra twist. The product was a person and the sell wasn’t only money, it was a vote.
To get an idea of where we should start, I thought we should look at what had been working in the short time the company has been around before I joined.
I started by digging into who had visited our website and given us their emails.
We had access to L2 voter file data, which had been broken down into Mosaic groups (household-based consumer lifestyle segmentations) by Experian.
I scanned through all of the groups and found resources online that broke down the characteristics of each group. Then I started putting together data for our top groups who had given us their emails.
For example, Silver Sophisticates were 66-75 earning $100,000 – $125,000. I took the biggest sections and put together a summary.
Our top 4 segments were older individuals who had money, but weren’t very tech savvy.
Then I looked at our social media analytics and realized that there was an issue with our strategy. We had spent a large amount of ad money on 18-34 year olds, but were not able engage with them.
Our biggest group, once again, was the 55+ crowd.
But how could we get the youth on board?
First step, talk to them.
To understand the younger demographic, I started to interview people who would be voting for the first time in the 2020 election. I conducted one interview before we had a larger team meeting where I could present what I had found during my research. In addition to presenting the information about the current traffic and social media engagement, I decided to share some insights from the singular interview, to find people to help me recruit.
The quote that highlighted our lack of success in advertising on Facebook was “There are not very many people on Facebook, college students use it some.”
Other insights centered around the younger audiences being on Snapchat and Instagram. They were mostly on snapchat for their news, because “You don’t have to follow a certain account, but you can still see stories of news sites.”
There was also a concern around data being gathered. “It makes it harder to see a variety of view points because it uses my data.”
This was in addition to the added benefit of not having to go to a different site when you swiped up to see more details to a story. The load time was a big issue according to my interviewee.
Around the same time that I was recruiting young people to talk to, I was told that we could not target people on age range. We needed to target a variety of age ranges, because we couldn’t win the election without it. But I still had an issue as a User Experience Designer, how was I going to make design choices if I couldn’t rely on commonalities within my user group?
I dropped setting up more interviews and switched my time over to another approach – one that might be better suited for a potential presidential election.
I gathered political groups from firms like Resonate and Simmons and I asked our political team to group them.
Not being political myself and lacking the detailed understanding of our political ideology as a country. I also asked them to order the groups on a scale from very liberal to very conservative.
I wanted to use their expertise of being in the political arena and on the road to validate the groups that two different firms had put together. I allowed the workshop to be open ended and gave them the ability to add information that was lacking too.
I put all of this together into an interactive PDF to send around the office. I presented it to a smaller group, showing them insights from pulled articles such as the New York Times on why we should not be discouraged by the negative social media comments we had seen thus far. The people who were currently interacting about politics on social media were the far left and far right. Most of the country in the middle, we we were actually looking to target, would not be involved in the political conversation until later.
And then I watched my first video of polling data and everything changed…
The way that politics works is somewhat opposite from how UX works. When it comes to user experience, you find your group of users and listen to them and then build solutions specifically for that group. Many successful start ups have been made this way, where an unknown problem for a group of people was uncovered accidentally.
Politics is different. Instead, politicians test different phrases, word choices, and positions with voters, and then see which ideas or words get the highest or lowest ratings from their party, and then build their platform around what people like.
There is also so much data tied to being a republican and democrat that you can target people based on this one sole factor. For example, republicans are more likely to drink dark liquors vs. democrats drinking clear ones.
Unfortunately, the same set of data is not available for 3rd party candidates. This put the political strategy team in a position where they couldn’t use their usual tactics. Instead, they had to wait to see how moderate democrats such as Biden were polling to know where and how to place Schultz in the line up.
During this waiting game, work load and travel for the larger team slowed, giving me a chance to collaborate with the political field team.
Because of the lack of data on 3rd party voters as well as leading political tooling organizations not wanting to work with a 3rd party candidate, my team decided that the best use of our time was to create an app to gather information about this group. It would let volunteers know where a household was in our conversation with them and include a survey platform to engage with voters.
We wanted to give Schultz an opportunity to make decisions based on current opinions of voters. We also wanted to empower all volunteer types with the app to engage with households they were visiting and to feel motivated by messages from Schultz along their journey.
This survey was also sold across the organization to show that it could work as a landing page to a bigger marketing campaign to help us understand what potential voters wanted.
To start this app, we first had to understand how the field team had used similar apps in the past. I started with individual interviews. I needed to know what the ideal amount of touch points they wanted with a potential voter. I also needed to find out what they had successfully done in the past. The developers were ready and waiting to build, so from these interviews, I started sketching a basic idea of the app. I brought the field team over to the whiteboard to get their input and then started creating a basic paper prototype.
I created three different ways to go between the map and house views and then went back to each person I had interviewed one on one to get their favorite and any other thoughts they had.
Then I made updates to the paper prototype based on the feedback. I soon realized that I needed more detailed information about the three types of house visits, so I came up with a workshop to fill in the blanks. The workshop included people within the organization that had experience going door to door and/or managing volunteers. I brought the updated paper prototype and unfinished flows to the workshop and gave the group a task. They has to use index cards to build out the flow for each type of house visit. They could discuss as a team the order and what to include.
Something that came up in the workshop that I hadn’t planned for was the fact that volunteers needed to tag houses in the system based on people not being home, if they had moved, and other circumstances. Surprise finding from workshops are my favorite reason to conduct qualitative research. I put together a quick sketch of scenarios and white boarded a flow to alert my dev team of the extra complexity.
It was finally time to test the app with real world house data. Each person on my team, including developers accompanied two political team members to watch them use the app in a neighborhood.
We all reconvened as a group once we got back to the office to discuss any updates that were needed for the app before Schultz announced his run. There were a few updates, but the political team was mostly really impressed compared to the software they had previously used on both the democratic and republican side. We had successfully created an app in-house for the field which was needed since other companies wouldn’t work with us out of fear of being shunned from the party candidates that they represented.