Feature concept testing

Problem

How might we get market-fit signals from sales meetings?

One of our company goals was to create a demo site for our sales team to use in sales meetings and mortgage industry conferences. One aspect of this demo site was to put feature concepts in front potential clients for feedback.

Visual design for the demo site

A new design system

When I started at Xpanse, the existing design language difficult to work with. The neomorphic design language could be found in applications with limited functionality. Xpanse’s product offerings had screens dense with content and functionality. The layered dropshadows and gradients were becoming difficult to work with. Additionally, I received signals from our leadership team and product team that the design was not appealing.

Nebula design system

I socialized the idea of updating the design system every chance I had. Our team eventually won over leadership and we created OKRs related to the new design system. After a couple of quarters of design iteration and cooperation with the front-end team, we had a new design system called Nebula.

Best foot forward

I was paired a brilliant front-end engineer for the demo site. We agreed to build all of the new functionality in the new design system and to retrofit the existing pages with the new look and feel. Everyone agreed this would make our product more appealing.

A five week deadline

There was an upcoming mortgage conference with plenty of opportunity to create connections and get feedback on our potential future features. We decided to create a standalone site with fake data that the sales team could use. We had 5 weeks to design and build the site!

Positive signals

Our product team and sales team had in-depth knowledge of the mortgage industry. Based on their knowledge and conversations with operations managers, we had positive signals for two experimental features:

1. Provider insights dashboard

2. Encouragement to use our smart provider allocation system

Provider insights dashboard

There is a lot of variability in the performance of appraisal companies (AMCs) for a given region. Reporting on AMC performance within the mortgage industry is reactive to past trends. We aimed to give lenders more information and control so that they could select the AMC with the best price and turnaround time.

We would start with dashboard where operations managers could view AMC performance down to a county level. We had existing research supporting a persona for operations managers. Next, I pulled some existing dashboards into a inspiration board. I began daily syncs with the sales and marketing team members to get immediate feedback and hone the concepts.

Choose a region to filter by

One concept that immediately drew praise was a suggestion by the front-end engineer to use a map to filter by state or county. I was able to quickly make a Figma prototype of how it might behave. The sales and marketing teams were sold and the engineer was able to build it quicker than estimated.

Video of the map-filtering Figma prototype

Smart opportunity

As an addition to the provider dashboard, we also want to show the operations managers that our smart allocation system would benefit them. The smart allocation system used machine learning to make estimates of how an AMC would perform for a given region.

Response and future considerations

The presentations of the developed demo features sparked interest at the mortgage conference. The sales team continued to use the demo site in future sales meetings and elements of these prototyped features were added to the roadmap for future development.

Looking back, I would have created a better definition of success. Everyone was very positive about the responses the site received, but I didn't attend the conference or sales calls. It would have been great to have gather specific, actionable details that we could use to continue to improve.