The Road to MVP

Project Focus

PropTech + CRM + CMS + B2C

Empowering real estate agents with modern technology

Generative Research

Design Thinking

Lean UX

Facilitation

Challenge

Create a unified platform (Website, CRM, CMS) that empowers the 10,000+ real estate agents, across 43 brokerages, with the technology to better serve their clients.

Outcomes

This is a collection of outcomes that the design team in whole contributed to.

New broker sites

4

User tests conducted

200+

Workshops facilitated

20+

Background

My entire time with Berkshire Hathaway HomeServices was spent building, validating, and preparing this platform for its debut. This case study will focus on the first launched broker site.

PROPTECH

The real estate platform wars

THE PROBLEM
The competition in real estate has gotten fierce over the last few years. Big disruptors, new business models, emerging technologies, and a new kid on the block, millennials. Not to mention the crash of 2008 which reshaped how a lot of people approach homeownership. These variables have all contributed to where we are today, savvier buyers and sellers, iBuyers — companies that use tech to make instant offers on homes, real estate agents fighting for their position in the transaction, and brokerages making bold moves. There's a lot to consider here, and it took me some time to really understand the landscape, trends, and forecasts for this market. But as designers, this fundamental understanding is critical to making meaningful contributions and adding value as quickly as we can.

THE VISION

Create a platform that empowers real estate agents with modern technology that makes buying or selling a home delightful and simple for consumers.

Each brokerage had a specific system in place that they accredited to their success. These systems are complex and composed of third-party vendors, in-house tech, refined processes, marketing teams and much more.

Additionally, the systems varied among real estate agents, real estate teams, and at the cornerstone of all this was the homeowner or seller. The number of inter-dependencies and overarching goals among the personas was incredible, we had a magnitude of work ahead of us.

Strategy

Working against a story map I would own the CMS and Agent Profile experiences. These were net new experience that required generative research methods up front and evaluative methods after.

Design Workshops

Inspired by design sprints and retrofitted to our unique needs

After several months of participating in our established sprint cycle, I identified an opportunity to increase velocity, alignment, and collaboration. While creating this proposal I drew inspiration from Google Design Sprints, LEAN UX design studios, and Invisions design sprints. Something to note is that this workshop was created to fill a gap in our product cycle, not as a replacement of a typical design sprint. The workshop aided in:

  • Contributing to the shared understanding of the design problem and the solution
  • Identifying assumptions, constraints, and unknowns that could impact the success of the sprint
  • Rallying teams around a design direction
  • Creating high-level design artifacts
  • Increasing the feeling of ownership across the team
After my team was successful in creating notable change in the areas above the workshops were adopted across other teams.


USER RESEARCH

Research was at the heart of our company. I learned new methods and grew more confident with analysis and synthesis.

Our entire process was organized around the LEAN methodology and validated learnings. This constant loop of user feedback would shape the product into something that was not only usable but also valuable. I conducted countless sessions that included:

  • Tree Testing to audit the information architecture
  • Card Sorting to to pair with tree testing results
  • Stakeholder Interviews to understand business objectives
  • Subject Matter Expert Interviews to understand real estate agents needs
  • Usability Testing to evaluate my designs with real users
The amount of research being conducted was a forcing function for a few tools and processes I would instrument to help with analysis such as:
  • Google script to create data visualization from usertesting results
  • Custom node scripts to parse user feedback and return a sentiment analysis score




Execution

We would need to support the largest real estate company in the country with 42,000+ real estate professionals, 1200+ offices, and located across 47 states. We would need to have a design language that stayed true to the Berkshire Hathaway brand but was flexible enough for its agents and brokerages to showcase their own unique identities.

Design Language

Creating a themable component library

AUDITING THE CURRENT DESIGN LIBRARY
I had experience architecting a design library from my previous role at Gaugebox and would lead the efforts. This allowed me to quickly segment components and identify patterns in our design files. After a comprehensive audit, I was able to surface what our design language looked like across the site. The usual culprits were found such as color variants, excessive fonts, similar but different components, range CTA's, and others. From this high level view we identified a path forward for the design language.

THE RIGHT TOOL FOR THE JOB
Axure was our tool of choice initially. It served us well for creating highly interactive prototypes which we used for user testing but it came at a high cost. Propagation. Changes made in one file would not propagate to files shared by the team. This immediately introduced fragmentation into our designs. I had worked exclusively with sketch up till point in my career so when I first noticed the lack of symbols, text styles, and palettes I was concerned. After a few presentations to the team we began testing several tools to replace Axure. Ultimately we landed on Figma.




The Agent Profile

Promoting honest and transparent relationships between agents and clients

Initially this was intended to be an agent website. The website would mimic the functionality of the main brokerage site with opportunities for Agents to customize certain content. This got complicated fast from not a tech perspective but from the users experience. This was validated when we looked at the data, which showed that few agents were actually maintaining their websites, and usability tests with low success rates and NPS scores.


After I conducted interviews with a number of Real Estate Agents to understand what the value of a website was. This turned out to be more of a percieved value though. Each brokerage offered it's agents a websites as a value prop to join their team. The agents would then be 100% responsible for maintaing that website with quality content, compelling calls to action, and lead management.
I would then propose a portfolio solution. That would put their online presenece on auto pilot and free them to do more meaningful work. The beauty of the portfolio was that not only was a better solution for the agent but it also lowered tech overhead signficantly and allowed to focus and optimize one portfolio for all agents.

The CMS

Providing Brokerages with a new way to manage teams, agents, and their websites

Their would be wide range of content, roles, and permissions this needed to support. Additionally, we were allowing for multiple ways of editing content via a Live-editor or text editor. These multiple ways of editing was to support the two types of users that would use the CMS.


TYPES OF EDITORS
We had the The "power user" persona, which was a Broker Admin who would be creating and managing all agents. The quick edit form would be ideal for this type of batch work.
Then we had the Agents, who would want to make finer updates and test different types of images and text to optimize their profile. The Live-Editor gave them the opportunity to view these changes in real-time.
INTERFACING WITH THE MLS
Their was one huge tech limitation in place. The fact that all the data we recieved for listings would be read only from each MLS. This meant that we would be able to fetch updates from the MLS but we couldn't write to it. That would get in the way of content syndication and require redundant updates in our CMS to update the new website content. I worked closely with engineering to understand how the MLS sync would work and then design around it.



Final Thoughts

My time at Berkshire Hathaway has allowed me to grow in ways I wouldn't have imagined as a designer. I applied learnings from previous mistakes and made new mistakes. I've conducted more experiments than ever before and I've built some amazing relationships with some very talented people. I hope this has given you a clear look into what it's like to be apart of a growing product team at a large enterprise.