We shaped our experience around three personas who use our marketplace. Using surveys, interviews, and market data we learned:

Both buyers and sellers used multiple services simultaneously to expedite reaching their goals. We interviewed our users about which services they use as well as what they liked & disliked most from each. We had them evaluate the search & discovery experience, available search filters, as well as data available on a listing.

We learned that sellers are eager to provide information when prompted that will help sell a car faster. Buyers appreciated having a feature rich search to narrow to the best car for them and to have car listings be as in-depth and detailed as possible.

We also conducted a stack ranking exercise with buyers and sellers that included different autos data attributes. This would help us prioritize what search filters to develop as well as what information is displayed on listings.

As a perk for participating in the Dealer program, dealers seamlessly provided car listing data through a feed. In order to not bias filtered search results towards dealers, we added car data inputs to the Item Upload flow so private sellers could add the same attributes.

Our initial hypothesis was that using a VIN or license plate number would provide a detailed data set for the car without much effort from the seller. However, in interviews sellers were apprehensive about providing it for perceived privacy issues.

We then explored adding manual inputs to the Item Upload flow. Because adding additional fields to the flow could impact completion, we mitigated this this risk by making most fields beyond Year, Make, and Model optional. During usability, most sellers commented that they expected to see the additional fields and liked that it prompted them for data they may not have entered otherwise in the description. After launching this feature, we saw strong entry rates with negligible impact to completion rates.

One goal for the buyer experience was to make autos filters relevant, discoverable, and easy to use. We considered adding to the existing filter sheet but we were concerned it would be unclear that autos specific filters were available. To make them sufficiently discoverable, I revisited our header system holistically.

We intended to test two versions: one using the existing filter interaction and one with an exposed filter treatment. I drafted some low fidelity explorations that surfaced autos filters, provided easier access to common tools like sort and price, as well as revisited the organization of existing elements. The “Invite” action got very little engagement and customers annecdotally told us they were much more likely to introduce OfferUp to friends through word of mouth. This increased our confidence to deprioritize it on focused search pages.

During usability testing, buyers immediately recognized the autos filters and appreciated the results updating instantly after applying them. When asked to update their location radius, customers confidently tapped on the map marker icon.

Our existing 3-column search layout displayed tiny thumbnails with no useful information about the vehicle. As part of our investment in the search experience, I explored new layouts that included at-a-glance info that customers had defined as critical: price, year/make/model, and mileage.

The new 2-column layout increases the photo size. During my explorations, I discovered that setting the tiles to a consistent height helps improve scannability and visual uniformity.

Buyers mentioned during usability sessions that having the vehicle data displayed on the search page is useful and helps them decide which listings to investigate further. However, due to engineering complexity, we weren't able to release this filter as it would have required a rebuild of our layout technology. Given customer interest, we kept the feature on our roadmap for a future release.

In addition to improving the private buyer and seller experiences, we also focused on our dealership customers. Because we had a significant share of used vehicle traffic and dealers already had budgets for marketing their vehicles, we developed a monetization strategy centered around dealers. Focusing on dealers also allowed us to maintain free listings for private sellers, which would not inhibit our marketplace inventory.

Using surveys, interviews, and dealership visits, we prioritized tools and features that help dealers sell. The subscription program included Verified Dealer branding to increase customer trust, special dealership profile pages, promoted results in search, and a dashboard to view the performance of their listings.

Dealer listings also included unlimited photos, click-to-call functionality for quick access to invested leads, and "estimated payment" information to help support dealers' financing businesses.

In just the first few months of the Dealer program, Autos became the primary revenue driver for the company. We continued to survey participating dealers and integrated their feedback and feature requests. One ongoing challenge for dealers was lead quality, which informed the next phase of our roadmap.

On the consumer side, we observed strong filter engagement, higher contact volume, and increased listing views in the Autos category. Additionally, we tracked improvements in customer satisfaction in our monthly survey.