DESIGN FOR THE RIGHT PROBLEM
By: Katie Olson
Acting Director
How many times have you or your organization been presented with an incorrect problem statement that leads you down the path to the wrong solution? Katie Olson talks about the importance of design thinking to determine the correct problem and find the right solution.
August 2021, after the U.S. began formal withdrawal from Afghanistan, federal agencies were flooded with thousands of requests from our Afghan allies seeking asylum status. In response, the State Department stood up the Afghan Special Immigrant Visa (SIV) program for Afghans who were employed by the U.S. government, allowing them to seek refuge for themselves and their families in the U.S.
Defense Digital Service, where I serve as Director, was asked to assist. The stated problem was sorting through 20 years’ worth of employment data to verify SIV applicants claiming to have worked for the Department of Defense (DOD) or a contractor of the DOD. But, we immediately recognized the organization of data amassed over the previous two decades would not produce the quickest path to Afghan employee verification. We needed a way to focus on just those people who came forward as an applicant, and to help them get the documentation they needed as quickly as possible. After meeting with the State Department externally and DOD’s Acquisition and Policy teams internally, my team determined the need to create a mechanism for employers with DOD contracts during the Afghanistan mission to streamline an otherwise convoluted process and connect an Afghan applicant back to their original employer for verification. Thus, we created Project Rabbit, which allows employers to upload and then verify their previous or existing Afghan employees seeking asylum in the U.S.
I have been presented with an incorrect problem statement time and time again over the last 15 years of my civic tech career. Earlier in my tenure at DDS, I was asked to support the Joint Pathology Center at the Defense Health Agency with its effort to digitize the world’s largest collection of human tissue samples. The digitization effort had been languishing, and DDS was asked to address the backlog of digitizing millions of physical tissue slides, which were stored in basement boxes looking like the final scene in Raiders of the Lost Ark. Once again, my team recognized that we needed to redefine the problem space. We landed on first addressing the process of digitizing new slides to avoid adding to the backlog, and then digitizing a more focused subset of the slides centered around one key opportunity area, prostate cancer. Today, we’re digitizing nearly 1M slides per year because of the focus on process.
I cannot emphasize enough the importance of design thinking and getting the problem (way before even thinking about the solution) correct. Almost every project I’ve been asked to put a team on, whether at the City of Chicago, the Department of Defense, or even as a Peace Corps Volunteer in Malawi, the problem presented is almost always not the core issue.
One more example to prove the point. Prior to DDS, I started the City Tech Collaborative, a public-private partnership with the City of Chicago. One of our first projects was digitizing the city’s maps of underground utilities in order for the City and utility companies to plan and deconflict construction projects. At the time, the City had about 30 different utilities (electric, telecommunications, transportation, even old telegram companies) taking up underground space. Each utility possessed their own paper-based maps of their underground assets. The City knew the process of overlaying individual companies’ and its own maps was inefficient and often led to inaccuracies. But the City also believed that the right approach was to digitize individual maps into a composite map.
After backing up and spending time with individual utilities and construction companies, my team at City Tech realized the actual problem was keeping the maps updated. Every day, there is a construction incident in the City of Chicago that instantly renders a map (paper or digital) obsolete. What the City actually needed was a system and technology to keep maps constantly updated. Thus, instead of spending years digitizing maps with little to no result in outcome, City Tech invested in a technology out of the University of Illinois-Chicago that could take iPhone pictures and render them into CAD maps. Hence, a construction worker could photograph their work, upload it, and update the map instantly for future reference. The actual problem was very different from the stated problem, and therefore so was the right solution.
Right before the holidays, the White House issued an Executive Order calling for enhanced service delivery. In addition to defining terms like “customer experience” and “human-centered design”, the EO calls for specific reforms to the way U.S. residents receive things like passport services. This is a step in the right direction, but I wonder how many of these reforms will fall prey to the same trap I saw with city maps, the Joint Pathology Center, or the Afghan SIV process.
One lesson is never to digitize all at once, but that’s not the point. You would never take that approach anyway if you understood the core problem, and that the ultimate goal is 9 times out of 10 to create a sustainable, forward-looking process. It’s not about the maps, or the slides, or 20 years’ worth of Afghanistan data. It’s not about the data at all. It’s about the ability to create a repeatable way to maintain an always-current, composite sketch of the city, or to verify employment as applicants emerge, or create a picture of digital pathology as specific collections become relevant. Data is a means, not the end.
As the U.S. government starts to embrace product delivery by, for, and of the people, let’s make sure to keep the end in sight. There’s beauty in organization, but peoples’ lives are literally on the line, as is the case with the Afghan SIV process. We don’t have time to correct the past. Focus on the end state and the rest will follow - including the data.
