// my role
product designer
UX researcher
// client
Auction.com
// summary
We performed a four-day design sprint with key experts to redesign our homepage.
// problem statement
Currently, Auction.com’s homepage doesn’t offer much value to users. We wanted to increase value to both logged in and logged out users by serving our logged in users curated lists, and by giving logged out users reasons to make an account as well as a better understanding of who Auction is and our value propositions.
// project goals
Meet with expert stakeholders from multiple departments to better understand their needs and goals for the homepage redesign
Using design exercises, decide what features should go on the homepage and which are most important
Rapidly prototype the first draft to capture the main idea and general flow that was decided from the sprint
Test prototype and iterate based on real feedback from early users to ensure we’re on track
// ux methods
Since the homepage is the anchor of our website, we wanted to get stakeholders from multiple departments and perform a 4-day design sprint to begin iterating on how might we redesign the homepage. We gathered stakeholders from Engineering, Marketing, Product Development, UX, Content, Operations, and Asset Management.
Day 1: We let our team frame the question by performing How Might We’s. From there we sorted them. Then we asked, “Can We?” which noted which constraints we needed to work around. After, we did an exercise called In 2 Years Time, which let us start to imagine what we could do in the next two years. At each stage, we voted on the best and found the ideas that resonated with the entire team. We looked at some Lightning Demos to see how other websites are organizing their homepages (even if they weren’t in real estate.) We ended the day with a customer journey map and asked each team member to sketch out their vision for how the homepage could look.
Day 2: We reviewed each sketch and marked which elements really resonated with us. We talked about the elements that people were drawn to. Then we voted on our favorites, the ones that everyone agreed were viable. From there, we did a quick card sorting exercise to figure out the hierarchy of what order the elements should be laid out in (for example, many people had testimonials on their sketches, but we varied on how far down it should go.) Finally, we decided by committee how to test the brand new design.
Day 3: Headphones on, nose to keyboard, rapid prototyping performed by me.
Day 4: UX Researcher, Content Writer and I get together to write the usability testing scenario. We also decide who to test: the logged out experience needed users who had never heard of Auction.com before, and the logged-in experience needed users who had used Auction.com at least a little. We put the tests up on UserTesting.com, and the results start coming in. We organize the results by each category and report findings back to the committee.
// key focus areas
How might we educate new investors about Auction.com?
How might we better collect and showcase the data that investors find relevant?
How might we classify properties so they are easier to find and more inviting?
How can we use AI to enhance the user’s experience?
How can we serve a curated product-based homepage that entices investors to buy?
// ux solutions
Unique experience based on user’s GOALS
We realized during our customer journey that the user’s goals on a logged out page are quite different from the logged in page. When logged out, the user ultimately wants to log in if they’ve made an account already, , make an account, or learn more about Auction.com to decide if it’s worth it.
The logged in user has more specific goals — they want to browse properties, find out information about a specific property, or follow up on past properties they’ve already bid on or even won. Since we know more information about them, we can serve them curated recommendations based on what they've already showed interest in.
SHOW value catered to the user’s goals
Since we cater to everyone from the first-time investor to dedicated professional investors that bid on behalf of companies, I pushed for us to split our page into dedicated personas. By letting users self-identify, we can serve them a much more catered experience in terms of what education they need and what properties they might be interested in. This also lets us show more of our value without overwhelming users with text.
LET THE USER CONTROL WHAT’s served to them
On a site like Amazon, your recommendations are often curated based on what you’ve already browsed. We liked that idea, but since our inventory is filled with unique properties, it’s possible you haven’t found a property that meets everything you’re interested in, and properties come and leave our site often.
Instead, I recommended a chip-based tag system. This would let users describe all the tags they like, and their curated list would give them results that match as many of these as possible, to help them find the “right” property as soon as it pops up.
//outcome
The early UserTesting results have been mostly positive, but the users have noticed some flaws in our design that we missed during the sprint.
One user noted that she didn’t want to make an account because we failed to explain our cost-structure on the homepage, and several wondered aloud if the account was free or how much it cost.
Our design moved the search bar up into the navigation menu on top, and as a result many users did not notice it at all. They assumed that they needed to make an account before they could search, which wasn’t true.
When we asked users to speculate what would happen if they selected a persona, we got mixed answers. Some assumed it would take them to make an account, while others thought it would take them to a page that had curated education based on that persona. A few users even skeptically guessed that all three links take you to the same page, and that the personas didn’t actually do anything.
We have organized the findings from the usability tests into an Affinity Diagram, and these results will shape future versions of the page as we continue to iterate on the design. I’m confident this design will come to fruition by the end of Q1, 2020.