Improving Dealer Workflow with Agile UX

The Project

“There’s too many clicks.”

“There’s…there’s what?” - Me

My new Product Manager (PM) tilted her head and gave me an inscrutable look. I had just joined her team 3 weeks ago.

“Too many clicks. Our Dealers keep telling me it takes too long to find what they need in the management queue. That’s something you can fix, right?” - PM

“Oh! Well, to be honest I’m not entirely sure, but I can definitely try!” - Me

This was my first request for work on the Dealer Service Management Tool (DSMT)*, a complex legacy app meant to help dealerships manage their service departments. And I was ready to go: a new design challenge, a new team, and a chance to make people's lives just a little better.

Goal 1: Reduce the number of clicks Dealers must take to accomplish their tasks.

Context: Putting the “User” Back in UX

John Deere had recently undergone a digital transformation and was adding UX designers to product teams. Pre-transformation, “UX” generally meant mockups the design team might send every 1-3 years, so in addition to solving for “too many clicks” I was also intent on driving a more collaborative and human-centric approach to product development among my team mates.

After that first conversation, I approached my PM:

“So, I know you’re used to getting mockups for something like this, but I was wondering how you’d feel if we talked to the Dealers first? I’d like to know what’s most important before we design something to support their needs. I can give you a proposal on what I’m thinking about by the end of the week if that works for you?”

And to my surprise, she agreed! I soon began planning discovery interviews that I'd be facilitating with 5 of our biggest dealerships.

Goal 2: Understand our Dealers’ key priorities and pain points when using DSMT.

Goal 3: Help the DSMT team be more collaborative & user-centered during product development.

Why Users Were Unhappy

During discovery with Dealers, we learned that the most important pieces of information on the DSMT management queue were hidden in multiple, stacked accordion menus, each of which had to be expanded before you could view or open the next one.

Well, I’d probably be annoyed too. No wonder we’d been getting complaints.

At least we had the core problem identified and a prioritized list of the most important data points in hand to take us into our next phase: Ideation & Design.

Goal 4: Redesign the DSMT management queue to elevate the most important data points.

Approach: Design & Learn Together

As mentioned, my developers and PM had traditionally been excluded from the design process; similarly our users had also not been privy to see work-in-progress out of concern that the lack of polish might reflect poorly on our company.

I understood why John Deere had approached software development this way. It mirrored the slow, careful approach with which we’d always designed and built our high-quality core products: tractors and other large machinery.

But for digital products, this kind of early feedback was vital to ensure our design was A) something people actually wanted and B) something our developers could actually build. So in addition to my standard design work like sketching and building mockups, I also led the team through a range of collaborative design activities, including:

• Co-designing possible solutions with my PM in Figma
• Holding design feedback sessions with Dealers for in-progress work
• Whiteboarding & post-up sessions with engineers for critique & work planning
• Iterating on the design during development to accommodate technical constraints

This collaboration resulted in a better design (see Outcomes below), and set the tone for how I hoped to work with my team as the UX Designer on future projects.

Outcomes: Better, Faster, Stronger 🤖

Goal 1: Reduce the number of clicks Dealers must take to accomplish their tasks.
Outcome: We reduced the number of clicks required to see key data points from 2 clicks down to 0.

Goal 2: Understand our Dealers’ key priorities and pain points when using DSMT.
Outcome: With the input from our Dealers, I prioritized 15 possible data points in the management queue by both importance and frequency. The key pain point was that some of the most important data points were needlessly buried in a complex UI.

Goal 3: Help the DSMT team be more collaborative & user-centered during product development.
Outcome: Because both the PM and our engineers were involved throughout the process, our team was able to move quickly into development and I gained credibility as a trustworthy UX Designer. Additionally, every backlog item we tackled that sprint was tied directly to the user value, a first for the team.

Goal 4: Redesign the DSMT management queue to elevate the most important data points.
Outcome: We measured the new workflow to be 4x faster for Dealers using the management queue, and we also stopped receiving Dealer complaints about there being “too many clicks” on DSMT.

*Note: Names, branding, and other information have been changed to protect intellectual property.