overview
Designing the On-Demand Delivery Experience at Favor
Favor’s mission (one of Texas’s largest on-demand delivery providers) was to improve the lives of consumers by connecting them with one of our delivery drivers for personalized delivery service of nearly anything they need. I led Product Design & UX Research. I managed a team of 5 designers and 2 researchers and was responsible for hiring, leveling, and growing the team.
As one of the original 4 members of the Product team at Favor and the second design hire, I had a hand in each hire for the Product team and designed or directed nearly every major release. At the time of my departure, I had helped build a diverse Product team of 18 PMs, Product Designers, and UX Researchers. I helped create and scale many of Favor's products and led our efforts to craft our first product vision. I also built Favor’s UX Research function from the ground up.
As a designer, I designed experiences for our consumers placing orders through the Favor platform, our delivery drivers (we call them Runners), our merchant partners, and our internal teams at Favor. I consistently worked with engineers and product managers to envision, test, and implement new features for our products.
On this page, you'll find high-level overviews of a few of my favorite projects from the last few years. I’m happy to walk through case studies of any of these projects in more detail one-on-one.
To comply with my non-disclosure agreement, I have omitted and obfuscated confidential information. Although all of these projects are live and currently being used, some of the images I show are pre-production designs.
Roles
Head of Product Design & Research
Design & UX Strategic Leadership
June 2019 - Present
Product Designer & Team Lead
August 2017 - June 2019
selected projects
Redesigning the Favor consumer mobile app ratings experience to gather better experience data
Problem
Favor had no shortage of ratings data to tell us what customers thought of their orders, but without a lot of manual parsing, it was incredibly difficult to understand what this data actually meant or how trends could be related to product releases and marketplace dynamics. We needed to standardize this ratings data and quickly be able to understand what was driving it.
Process
I led a full evaluation of existing ratings data and learned what attributes tended to drive positive and negative ratings. I also discovered that people have really surprisingly inconsistent ideas of what a “one-star” vs. a “five-star” experience is–some users actually used one-star to mean “an outstanding experience.”
Result
My team designed and released (in early 2020) an updated ratings experience. Highlights of the experience:
Star explanations: We now explain to consumers a standardized idea of what we mean by 1⭐️ or 5⭐️s to ensure we are getting consistent, relevant experience data
Quick attributes: We ask the consumer to indicate what factors contributed to their experience. These were labels I created from my analysis of existing ratings data.
Consolidated fields: We ask for one, holistic experience score (since users had trouble understanding the nuances of rating their driver vs. their order in the previous experience). To them, it was all one experience.
This experience has increased user speed and comprehension, made users more likely to rate their experience, and increased our ability to action on ratings data.
Crafting a custom chat tool for Support Agents
Problem
Our support teams were using 4 different tools to do their day-to-day job of communicating with our customers–one very inefficient (and expensive) chat agent, one easy-to-ignore phone tool, one aged and inefficient internal tool to proactively discover problems and reach out to users, and Zendesk for ticketing and email communication. This made their response times incredibly slow as they searched for context and actions in different platforms and was hurting us as our business scaled fast.
Process
I led all aspects of design and research (and at one point, product management) on this project. I started by watching agents work at their desks, asking questions and understanding their workflows. After extensive contextual inquiry, I designed a support tool from the ground up to help agents chat with our Runners and take quick actions on their account and orders all in one place (built on top of Twilio’s pre-Flex Task Router technology). After many wireframes, iterations, and live tests with agents, we released version 1 of a new support tool in early 2019
Result
We were able to boost agent productivity by ~40 percent for the agents who moved to our new tool. Meanwhile, I learned a ton from continuing to watch agents use the tool, and made several UI and user efficiency improvements in our Version 2 release in late 2019 (about 6 months after version 1). This tool is still in use today, and at the time that I departed Favor, was being rebuilt on top of Twilio’s Flex platform.
Balancing Runner supply with consumer demand through neighborhoods
Problem
As Favor grew, the area we covered in each city expanded, but our ability to guide our Runners to the right places at the right times didn’t scale with our coverage. This made it very difficult to balance supply and demand in our markets.
Process
We designed these experiences alongside top Runners, showing them wireframes and proposed experiences along the way. We explored many different ways to display the incredibly dense information we needed to show Runners, paring down to the simplest information needed to find the best place to work at any given moment. Our explorations ended in a full revamp of our Runner sign-in and scheduling experiences to better distribute our Runners in areas we need them and ensure that they stay busy.
Please pardon the extremely dirty whiteboards in our old office.
Result
This was one of the first experiences I designed at Favor in 2017 and remains one of the most impactful, saving Favor millions of dollars in Runner top-ups (basically, guarantees we pay to Runners who weren’t able to remain busy enough).