What was the problem you were solving?

At Opendoor, we’re creating a radically new and simple way to buy and sell your home. One of the ways we do that is by providing each customer an online dashboard to digitally track their selling process, which they don’t get when selling traditionally with an agent. We call this dashboard the Closing Dashboard.

After the seller decides to accept Opendoor’s offer on their home, they begin the closing process, which historically has been a complex process. Traditionally, customers have a long list of tasks, including setting up an inspection, performing repairs, coordinating title and escrow… the list goes on.

For customers who use Opendoor, we wanted to make the closing process frictionless so they feel in control every step of the way.

Digging in

Based on research, we created a kick-off deck to define the highest priority problems with the current dashboard, summary of the design goals, customer pain points and a heuristic evaluation of the current information architecture.

Failing early

I then started out diving into existing research and looking at inspiration. Companies like Shift, Wealthfront, and Better Mortgage were all similar in that they guide customers through complex transactions.

We then brainstormed ideas in a sprint and in tactical workshops with PM and Research. We focused on ways to reduce friction (high anxiety moments that we learned in research).

We also explored ideas like adding a summary page and grouping similar tasks together. By role playing out these ideas as the user we found that grouping similar tasks together failed and would lead to customer confusion and navigation difficulties.

Prototyping & Research

We developed two concepts to test in UX research, the first was just iterating on our existing dashboard by adding a simple pizza tracker at the top and the second was with a new landing summary page with the cleaned up IA.

The second concept with a summary out performed the existing dashboard because users loved the new summary page with a clear call-to-action on next steps and the ease of navigation. This added a new screen and step in the experience and we were excited to get a clear signal that this was a good path to go down.

Getting into the details

We first built a closing dashboard for our customers 4 years ago. Early versions of the Closing Dashboard were based on capacity and need, functioning primarily as a supplementary tool for a CX(Customer Success) team member to walk a customer through each step of the closing process.

As we grew, so did our vision for the product. The idea to make the Closing Dashboard more self service would allow us to scale and grow our capabilities faster for our customers. Last year we launched 10 new cities, and it was imperative that we streamline the closing process with product.

We started fleshing out all the flows and spent a good amount of time iterating on the new summary page.

We partnered with the Head of Customer Experience (CX) to gain insight and requirements into the new summary page and also the detailed flows. They had wanted for years to make changes to the closing dashboard like adding the summary page and fixing issues in the detailed flows that were disconnected with the processes that the CX team followed and walked the customer though.

For example, the customer now had to sign their repair addendum before making their own repairs but that wasn’t represented in the current experience and led to a lot of headaches CX and the customer.

Building out

During the build phase, I worked closely with my engineers to QA the final product. We even made time to add some delight into the MVP — transition animations, loaders and fun illustrations.

We fleshed out all the main flows for the dashboard, I used Dropbox Paper to keep an inventory of all the components needed and keep track of notes. I used Overflow to create the flows with arrows so people could at a glance understand.

We used Figma comments to get alignment and feedback on the copy between PM, CX and Content Strategy.

Extra details & launching

I wanted to explore custom loaders since there could be long moments of waiting while the web app loaded and also for things like contracts that could take up to 10 secs. The team was really excited about them and felt like it added another level of polish to the experience.

We launched in San Antonio first and we immediately saw a decrease in time with customers in completing tasks. For example, before it would take 2-3 days for customers to schedule their assessment and we saw customers completing it within ~12 mins after signing their contract.

Opendoor Closing Dashboard