Anywhere Real Estate - Referral Platform

A platform accelerating lead delivery, management, and conversion that connects all involved parties to help simplify real estate transactions.

Background

Built primarily on the Salesforce (SFDC) Sales Cloud and Experience ecosystem, my main problem to solve was closing the gaps to unify the agent search and lead placement feature areas. With a hefty investment in a business rules engine to match consumers with the right agents, ensuring high utilization was a key north star metric.

Role

  • Product Manager

  • UX Designer

Product Tools

  • MIRO

  • SFDC Design System

Timeline

  • Q4 2022

Problem Definition

How might we normalize the input patterns of the referral's destination data so that the business rules engine for agent assignment is utilized more and yields the right results?

There are two paths to assign a real estate agent to a referral:

  • Manual assignment

  • Semi-automated assignment based on agent performance and eligibility data

Unstructured areas of interest data received from over 45+ lead web forms managed by external sources

User is prompted with little to no guidance on how to correct this error, which leads to manual assignments

Research

With the program goal to assign 75% of referrals in a semi-automated fashion, data showed the complete opposite.

  • 59% of referrals in the Platform were manually assigned.

Additionally, SFDC's out-of-the-box UI and error handling were very limiting during the initial rollout of the Referral Platform (i.e., April 2021). I led the discovery of how we could amplify the user experience by expanding the usage of SFDC's design system and staying within the declarative programming best practices set by our solution architects.

To ensure we designed a user interface that was simple and reached the program's desired outcome for the agent assignment team, I facilitated usability testing with low and high-fidelity prototypes. This, in turn, protected our engineer's scope of work and limited significant changes throughout the delivery lifecycle.

Results

In the first two weeks since the rollout, we reduced the manual agent assignment rate from 59% to 27% with an "in-the-moment" data correction utility.