Project description:

“Which of these 2020 Democrats agrees with you most?” is an interactive quiz to match readers to the Democratic primary candidates that they most closely align with on key issues.

This project sprung from a database of 86 questions and nearly 2,000 stances that we built over the course of 2019. We reached out to all of the Democratic campaigns, the biggest field in history, in order to compile this original reporting.

From this database, we pulled 10 questions – later expanded to 20 – that we thought revealed the most interesting differences between the candidates.

Impact reached:

The reader response to this project was extraordinary. Millions completed the quiz, and thousands took to social media to express their delight and surprise at the results. The feedback made one thing very clear: The project was getting readers to engage with tough issues, and to take a step back and think hard about whether their preferred candidates actually aligned with their worldviews.

This is a major achievement. Understanding the policies that a presidential candidate would enact is a critical part of the election process.

Every election, readers tell us that they are interested in reading more substantial policy coverage and less “horse race” coverage. In practice, this request is often aspirational. Any political reporter knows that lots of policy coverage is ignored by readers, even if heavily promoted. This quiz proved that you can get people to engage with this coverage – even coverage that is fairly in the weeds – if you are thorough with your reporting but also creative with your framing and design.

The project made an impact on our readers, but it also made an impact on our publication. The quiz got several thousand people to make a strong commitment to journalism and subscribe to The Post. It was the paper’s most successful project of 2019 in this regard.

Techniques/technologies used:

Technically speaking, this project was not that complex. The database of candidate stances was a sprawling Google Sheet. The layout of the quiz was done in React, but that is pretty common for our pages nowadays.

What really made the project shine was the design of the user experience. Policy deep dives can scare away readers that don’t consider themselves wonky. The design of the page – the question wording, the font sizing, the illustrations, the button shadows, the sticky bar, the explosion of confetti at the end if you answer everything – was in service of making the information and results as approachable as possible. There is lots of detailed information on the page, but it is only available once the reader has already made a decision to engage.

Most of these design decisions were subtle. We spent a good bit of time, for example, on deciding what the feedback for the buttons should be – how quickly the results and background information should fade in. But good design is greater than the sum of its parts, and without these interventions the page would not have worked nearly as well.

The most important tool we used, beyond design, was the old-fashioned reporting that got us the candidate stances in the first place. That was also the hardest part of this project, described in the next section.

What was the hardest part of this project?

The most difficult part of the project was compiling our database of policy stances. From March through December, we published policy pages covering nine broad areas: health care, immigration, voting, climate, education, foreign policy, the economy, guns and criminal justice.

This was a major undertaking. We asked the candidates 86 questions. Given how much the Democrat field swelled over the year, this ended up working out to a final total of around 1,800 candidate stances.

Getting these stances was not as simple as sending along a questionnaire. For one, the questions needed to be meticulously crafted to be as clear to readers as possible without being too broad so that the candidates all sounded about the same. It’s often not in a campaign’s best interest to give straight answers on controversial subjects, and getting these campaigns pinned down on a tough stance was a difficult act of reporting that we had to repeat over and over again.

When campaigns passed on giving us answers, we researched their previous statements and legislation to get a sense of where we thought they stood. We kept the database constantly updated even as the candidate’s own stances shifted over the course of the year. In addition, we researched and wrote background information on all 86 questions to help explain to readers the nuance and importance of the policies we chose to highlight.

All and all, the quiz was just a capstone of a year of diligent campaign reporting that put us in position to create the most comprehensive interactive possible.

What can others learn from this project?

As mentioned above, I think that readers could and did learn a lot from this project. For practitioners, I think there were four big lessons.

One: People who work on graphics teams are reporters like anyone else in the newsroom. Kevin Uhrmacher dreamed up our policy database, but was aware that the campaign reach outs it would require would be a big change for him. With some guidance from other parts of the newsroom, this project transformed him into a diligent and effective campaign reporter.

Two: You shouldn’t be afraid to be flexible with your plans and your teams. Kevin Schaul, Kevin Uhrmacher and I worked on the policy pages for a lot of 2019. The database by itself was, frankly, not connecting with readers that well.  We were a bit stuck, “too deep” in our own reporting. I brought in Brittany Mayes to help design the quiz, and it was her recommendations for framing and page design – coming from someone with an interest but not an obsession with policy – that really elevated it into something that would connect with millions of readers.

Three: Interactivity works, if well-designed and used appropriately. This has been an ongoing debate in newsroom graphics: Will readers actually click on your buttons? Or is it safer, especially in a mobile environment, to ditch that kind of interaction and lean into the scroll. We are pleased to report that over three quarters of readers who answered one question in the quiz answered every question, even after it was expanded to 20 questions.

Fourth: Policy coverage can work, and is worthwhile. But you need to be creative about your framing and presentation if you want to reach a broad audience.

Project links:

www.washingtonpost.com/graphics/politics/policy-2020/quiz-which-candidate-agrees-with-me/

www.washingtonpost.com/graphics/politics/policy-2020/