Case Study

Tracking WeWork's Real Estate Deals

On Oct. 10 2019, WeWork had 622 locations open in 123 cities across the globe. There was a huge opportunity for technology to support WeWork’s real estate team during this expansion. This project was a zero to one effort to first model the different roles on the real estate team then design software to support the workflows and collaboration across the team.

By understanding the team’s needs and behaviors, and not designing around fleeting roles and responsibilities, we were able to create a proprietary piece of software called “Dealtracker” that was grounded in human behavior and successfully scaled with the company as roles changed. Dealtracker increased efficiency and data integrity, supporting WeWork’s rapid growth.

Senior UX Researcher
Key Metric
Decrease production hours, Increase collaboration
10 weeks
The cross-functional delivery team sketching together.


For some time, it didn’t make sense to put technology in place to facilitate workflows of internal WeWork teams. The teams were growing and the work was evolving, so whatever was working last week would not work today. It was not a priority for the team and mostly third-party tools filled the need.

When I joined WeWork, the company was moving into a new stage. Patterns in ways of working across teams and continents emerged. This was a strong enough signal for leadership to prioritize building proprietary applications to support the internal teams. 

However, titles, day-to-day responsibilities based on role and other nuances were still very different depending on team, region and timeline or scope of the project. 
Our team needed to model out activities, mindsets, goals and workflows agnostic of corporate titles and team-based responsibilities in order that we designed a system that supported the global workforce.

This effort focused on mapping the workflows and modeling its users. 

The second stage was designing the software based on what was modeled. 


Our team delivered personas and two experience maps in a friendly “Meet Your Users” guide. Video below and images further down.

The User Guide gave the team: 

  • A way to communicate with each other and stakeholders about design decisions. 
  • The tools to stay connected to the user throughout the product development process. 
  • Signal for impact measurement and instrumentation by knowing what should be measured as successful behavior
  • An easier way for people to relate to the work being done and empathize with the people the team was helping.

Ultimately, this work propelled the team forward to deliver a successful product with confidence and clarity. 


Dealtrack, WeWork's proprietary real estate software.


My role was to design and execute end-to-end UX research across the generative and evaluative product development cycles. I collaborated alongside a team of product design, engineering, data science, product management and SME’s from the real estate team. 


Observing the team in their environment was particularly important for this project because of the environmental considerations. For example, deal managers were often walking sites under construction with tenants. This required gloves, hard hats and paying attention to avoid hazards. 

Multitasking for a deal manager involved walking a site with a potential client and the construction team on an open roof on the 10th floor while simultaneously participating in a legal call about another deal. 

Team members not out in the field were often doing dictation of information from people walking sites, jumping into planes, trains and automobiles. The same people doing the dictation were also doing a lot of the manual deal tracking and dictating back to those in the field when a change happened throughout the day. We spoke to both during in depth interviews and observations.


To model personas, we first created behavioral continuums and plotted our interviews along the continuums. Then, we captured clusters of participants, looking for patterns in those clusters. Clusters of the same people were the beginning of our models. 

People’s titles did not necessarily translate to universal truths about the activities they did. 

Mindsets and activities were variable depending on the team’s size and maturity. 


We wanted to communicate what we learned in various ways because we knew based on the audience and situation, different cuts of the data would be more useful than others. 

We went through several iterations of the personas and designed them to be easily added to product or engineering presentations in addition to being included in the user guide.

Once product managers started their demos and kickoff meeting talking about “John” and “Leah,” including them in presentation decks, we knew we’d made progress. 

Persona for presentations.

Persona for presentations.

Persona for the user guide.

Persona for the user guide.

Experimenting with Experience Mapping 

The UX lead and I worked very closely on the experience maps. It was an opportunity for each of us to lean into our strengths but also push each other through constructive, actionable feedback. 

We started out sharpie to paper, iterating quickly on the design with feedback from the team. Both maps were original, innovative and communicated different cuts of what we learned in a compelling way. 

I focused on mapping the "day-to-day," highlighting information needs.

The UX Lead, Sam Carmichael focused on tensions along the deal process as well as what channels people were using to complete a task.

Putting the User Guide into Action

Just like the storytelling artifacts, we needed to approach communicating what we learned to different audiences in different ways. 

For our designers and engineers, I facilitated sketching sessions based on the persona opportunity statements. 

Product managers and our senior leadership used the data in prioritization meetings I facilitated.

Full Timeline

Week One: Design the generative research approach. 

Week Two and Three: Begin all logistics related to conducting the research including travel plans, recruiting participants and resourcing with cross-functional teams of design, product and SME’s.  

Week Four: Conduct the research. 

Week Five: Synthesis including persona modeling and user journey mapping. 

Week Six: Feedback on personas and user journeys. 

Week Seven: Collaborative workshops to create concepts with rapid prototyping & user feedback sessions. 

Week Eight: Define the MLP roadmap. Create user stories from our new understanding and begin feature prioritization. 

The Team

Carson Andrews, Product Designer

Laura Cochran, UX Researcher

Sam Carmichael, UX Lead

Amal Muzaffar, Product Lead

View More Case Studies

Building a Leading Consumer Digital Bank
With a robust asset base of approximately $42 billion, Valley Bank, a regional entity, operates over 230 branches catering to both personal and commercial customers. As the bank experienced continuous growth, attracting an increasing number of customers daily, the leadership team recognized the necessity to create a seamless digital onboarding experience.
Saving Editors 100's of Production Hours
This project focused on a feature enhancement to Copilot, a proprietary content management system used by Condé Nast editors. Copilot powers the production and distribution of content for ~70 Condé Nast brand teams across 10 countries and in 8 languages. It has 4600+ users. Some editors were manually cropping up to 300 images a day to meet responsive design requirements as the number of distribution channels continued to expand.