Framework for Radar User Research
A recap of our 2018 user research
One of the biggest challenges facing the dApp (decentralized application) ecosystem is user experience (UX). Like most dApps, Relay has made UX improvement a focus for 2019. To help the ecosystem, we’re sharing some of the strategies and tactics we use for UX upgrades.
Improving strategies and tactics for UX is complex, but our approach is quite simple. We research and design a feature, gather initial user feedback, then evaluate and evolve our user interface based on that feedback. Along the way, when we notice a recurring support topic, we make an educational site. For example, we created sites like wETH and Token Allowances to help new users and saw related support tickets plummet.
The blockchain industry mandates new user actions for both dApps and users. With Relay, our most common user issues are related to the platforms we’re leveraging. For example, wallet set-up and integration, the Ethereum network in general, and the nascent 0x protocol technology. As a dApp, we’ve had to learn how to do support for each piece of the user flow, even the ones we don’t control. Meanwhile, users are faced with brand new, often unintuitive user actions. We needed a method to provide effective UX in the new technical environment.
2018 User Research Overview
Talking to users is the most effective way to improve UX. Just this past year, we selected 65 users to conduct in depth UX interviews with. These users were split 50:50 between new and power users who regularly trade on Relay. We included new users to see how they interact with the Relay interface for the first time, and what their initial hurdles are within the app. We included power users to test our own assumptions around Relay usage, and learn about what features they’d like to see in the app.
All interviews were divided into two parts. First, a conversation about trading habits and their understanding of blockchain technology. Second, we asked users to share their screen, so we could record them using Relay and engaging in the specific tasks we provided.
The prompts on the screen interviews were varied based on answers provided in the initial interviews, but the overall theme was the same for all users. While providing a general range of prompts such as “Buy X quantity of XYZ tokens with $20 of ETH,” users inevitably faced 0x protocol barriers such as wrapping ETH and enabling tokens. We refrained from intervention as much as possible because observing how users overcome obstacles is just as important as identifying those obstacles in the first place.
After conducting all interviews we had a substantial amount of clean user data to analyze. Thanks to those users, we updated many aspects of the Relay UX! Below, we will go through three of the most common problems we found and describe our solutions.
User Issue #1 — Navigation Bar
The navigation bar is a critical user interface for any product, and the Relay navigation bar is no exception. Users look to the navigation bar to complete tasks quickly, like connecting their wallet or moving to other pages.
Problem: Users found the minimalistic icons and simplified buttons to be confusing, and requested a more obvious and direct experience on the navigation bar.
“top-right side icons are hard to understand,” “screen mode has too large of a portion,” and “I am confused with WiFi icon.”
Solution: We linked directly to the setup guide, reduced the weight on the icons, and used words instead of symbols. This ensures we are providing easy access to user’s most desired app functions while placing lower priority on less demanded (or naturally intuitive) functions like the support page and screen mode selector.
User Issue #2 — Market Selector
Using the market selector is a core user action that must be simple. Users spend much of their time in the market selector deciding whether to (and when to) buy or sell tokens.
Problem: Users were unaware of token searching and were getting frustrated with the dropdown menu. Further, they asked to have more actionable data in the market selector, as this is where they made trading decisions.
“I would like to see price movement here,” “since it is only two marketplaces, it would be better to show both markets,” and “I do not want to see tokens alphabetically.”
Solution: We redesigned the market selector to include search functionality and market data at a glance, like the 24hr price change and volume. Switching from wETH to DAI quote pair is also more intuitive.
Issue #3 — Order Book and Order Handler
The order book and order handler received the most feedback during user interviews.
Problem: Users wanted more flexibility in setting their order duration, identifying where their order was going to be placed, and more compatible input fields.
“I am confused which part of order book is for selling wETH,” “I want to copy the price from the order book,” “Adjustable time period for orders, set your own time,” and “Shorter time frames for orders will increase liquidity.”
Solution: To overcome confusion around the order book, we created an indicator line that hovers on the order book to show where the order will be placed on the book. In addition, we added more flexibility in the order expiration period, giving users the option to change order time from five minutes to a maximum of nine months.
Conclusion
Like the early days of the Internet, dApps have many new and complex user actions. We all must lean in, talk to users, and create new design patterns. We’re sharing our UX framework and examples from our user interviews in hopes that we can help other dApps navigate these complex new user actions. We’re committed to continue learning and sharing our findings with the community.
If you’re a dApp builder and have suggestions on UX process, we’d love to hear from you.
If you’re a user of any dApp, don’t hesitate to reach out to the project you’re using and give them critical feedback.
If you’re interested in participating in Radar’s 2019 user research, contact us here: outreach@radar.tech
