THE PROBLEM

Hawkfish’s proprietary data showed early signs of what its CEO coined the Red Mirage: a scenario in which Donald Trump would appear to lead on the night of the 2020 U.S. presidential election, only to be overtaken by Joe Biden after a certain threshold of mail-in ballots were counted.

Public understanding of such a scenario was vital in encouraging those voters planning on voting in physical booths to not stay home assuming that Trump would be victorious based on developing election night polls. Without that understanding, the country could be swayed to believe that Trump was the true winner, even after it was confirmed that Biden tallied more electoral votes.

Further, it was projected to be days after election night, if not weeks, before enough mail-in ballots could be counted, and a winner was called.

In an already stressful climate fueled by a year-long pandemic, rising unemployment, and partisan hysteria, honest analysis of incoming data leading up to an official announcement of a winner would be critical in keeping the country stable.

It was clear, then, that there was a pressing need for a predictive tool that incorporated the pivotal mail-in ballot data in its models to show voting results in their proper context.

The tool would need to function as both a primer to help normalize the Red Mirage concept before it happened on election night, and a comprehensive data source highlighting key areas to focus attention on after election night as votes continued to be tallied.

Hawkfish, with access to exclusive data models incorporating mail-in voting expectations, would attempt to meet those demands as two related, but distinct, dashboards, which could be leveraged by media members to help keep the public at large informed about what was predicted to be an unprecedented election.

THE TEAM

As the lead UI & interaction designer, and key UX designer under the Director of UX, my role on the project was to:

1. Create visual guidelines, wireframes, and core components supporting all the necessary interactivity required by the dashboards.

2. Build prototypes to demo product features for stakeholders and developers.

3. Work with the development team after handoff to facilitate accurate translations of the final designs, and collaborate on changes when pivoting was necessary, all under severe deadline restrictions.

THE RESEARCH

Competing dashboards allowed users to choose from curated election night scenarios, then tweak Democrat/Republican electoral votes to discover how outcomes might swing.

A key observation from the competitive analysis was that none of the available dashboards factored in mail-in ballots as a parameter or interactive component, bolstering the conviction that our tool would add value in this space.

In addition, the breadth of data in competing dashboards’ existing scenarios was limited to the state level, and user interaction was restricted to toggling a state’s level of partisan support to either hard, leaning, or neutral.

The REQUIREMENTS

A key business decision was to shift the focus of the dashboard on election night by introducing new feature sets and removing old ones. While there would be a visual overlap, a retooled interface with different UX considerations would be necessary.

The pre-election dashboard would:

1. Focus on trending election night scenarios, and how they compared with Hawkfish’s own final outcome predictions.

2. Feature advanced user interaction with more customizable parameters than those of competing tools.

The post-election dashboard would:

1. Focus on narratives which would highlight key areas of the country to keep an eye on as the vote count continued to update.

2. Feature comprehensive voting data down to the county level.

The DESIGN DIRECTION (Pre-election)

An audience made up of politicos in the media had been carefully primed to use the dashboard upon release. Catering to that user base steered the direction toward a desktop-first design.

Although the tool would be functional at any size, this provided opportunities for moments of visual impact more difficult to establish on a mobile device.

With a key factor being the ability to measure differences between election scenarios, a compelling way to present those comparison maps was needed.

An early option I explored was a single map area with an overlaid slider that revealed/hid the final state of the race on one side of the slider and the current state on the other. A more impactful option was a single map area with a switch to toggle between the current and final states of the race, highlighting those states that flipped to expose the Red Mirage.

While a large dedicated area for a single map allowed for a clean design, internal testing with users uncovered a negative side effect:

The default view showed Donald Trump in the lead on election night, and hid the important outcome of states flipping to Joe Biden behind a user interaction.

Another important piece of user feedback was gleaned: as this dashboard was to be used by media personnel who understood that our models incorporated the critical vote-by-mail data, the potential of a screenshot showing an out-of-context map where Trump was leading could, if propagated, perpetuate the false narrative of a Trump victory, and devalue Hawkfish to boot.

To address this, my final design arranged two maps side by side, ensuring that the predicted Biden victory map would always be visible.

A second key feature of the pre-election dashboard, the ability to fine-tune scenario parameters, needed to be tucked away to keep the basic interactivity of the dashboard free from distraction, but also be easily accessible for advanced users.

A floating action button that, when pressed, revealed scenario inputs at the state and national levels (and also showed the selected scenario’s data sources), was my eventual solution based on its simplicity while fulfilling requirements.

The DESIGN DIRECTION (Post-election)

The post-election dashboard would be pushed live on election night, replacing the original dashboard.

This version would be a highly curated experience, with interaction limited to selecting states to reveal detailed voting data down to the county level, and selecting up-to-the-minute narratives to uncover Hawkfish’s interpretation of salient data points around counties to keep an eye on.

Immediately shown to users would be the current state of the race, and key data points to be aware of on a national level.

As it was also important to show that the data was up-to-date, the dashboard would exhibit notifications of the latest called states, and refresh countdown timers.

Thoughtfully crafted narratives would keep users focused on the not-yet-called counties most impactful to the race, along with their historical significance in prior elections.

A meaningful team effort (and initial brainchild of the project’s UX Director) that underwent numerous iterations was a data visualization chart that distilled the current voting data around Trump and Biden into a compact snapshot, incorporating into it a comparison of how Hilary Clinton fared in 2016.

The chart would be applied to national, state, and county data.

Surfacing comprehensive voting data at the county level would be the killer feature for advanced users of the post-election dashboard, as they would be able to quickly see relative county population, level of partisanship, demographic breakdowns, and more.

The Final Designs

In terms of UX impact, the final product addressed feedback from earlier iterations to optimize user needs.

In terms of business metrics, the effects of the dashboards in media coverage was difficult to track, but the Red Mirage became an oft-spoken term during the election cycle, and, internally, the dashboards represented the first time Hawkfish teams across data, tech, product, and design would coordinate closely on a single project.