Virginia Dept of Transportation
Helping citizens report road problems and get easy assistance
I observed a number of calls that the customer service team fielded. I also placed a couple of legitimate road repair requests to experience both sides of the conversation. In addition, I listened to calls and surveys collected previously by the research team. I identified the data needed for various types of requests and created a conversational script that we used to define the flow of the form.
I sketched out several ideas for a design of the form, including responsive and mobile versions. Once we settled on a solution that felt right on paper, I built a working responsive HTML prototype that we could iterate on quickly to improve flow and copy. Throughout the project, we continued to analyze data and tweak the architecture of the form so that the most common requests were easy to address.
We tested the prototype with people unfamiliar with VDOT’s process, both in the office and outside. We ran through the prototype with the VDOT customer service team on a regular basis to verify that our findings would work with their process. Finally, we ran a usability lab at the local university where we had over a dozen students reporting various common road problems through the mobile web app to identify usability issues.
A conversational approach
From early in the project, it was clear that the information needed to successfully respond to a request depended a great deal upon the type of request being made. Rather than ask for everything the customer service rep could possibly need (and thus overwhelm the citizen), we decided to make the form conversational. How the citizen answers each question affects what further questions they are asked and even how the common questions were phrased.
The result was a simpler form for citizens to fill out. They felt like it was listening to them, and it only asked them for details relevant to their problem.
Making the form feel conversational was both challenging and rewarding. It required a lot of technical work and copy tweaking to show different questions and prompts based on the request. But as we ran through it with people, we heard multiple times that they felt like the form was listening to them. It also helped VDOT because requests came in with more relevant details.
State agencies are often adverse to being at the leading edge of technology trends. Some of the techniques I advocated for (like password-less accounts) didn’t end up in the system because they were still relatively foreign concepts. Starting with the leading edge, though, allowed us to compromise closer to a user-friendly solution (like using Hide/Show password).