Facebook Marketplace Real Estate

From 2016-2019, I led design for Facebook Marketplace Real Estate, creating & shipping dozens of features focused on optimizing Facebook to become a trusted destination for connecting shoppers with local homes.

Role

Lead product designer

Duration

2017-2019

Product

Facebook Marketplace

Property feed, home listing page, contact agent form

Context (How did Facebook get into the real estate business?)

At launch, Facebook Marketplace was a minimal product experiencing rapid adoption with limited inventory, shopping features and seller tools. Among the emergent categories, home rentals & sales gained significant traction organically.

Problem

Facebook didn't have specific enough tools for buying or selling homes, so users listed properties as generic items on Marketplace. This made it hard for buyers to find home listings, and sellers could only share basic details in plain text, often missing important info like fees and pet policies. This led to mistrust among early adopters considering Facebook as a home shopping destination amidst alternatives.

Opportunity

We asked ourselves how we might explore and enhance support for listing and searching for homes on Facebook while staying aligned with current app usage patterns. To decide where and how to invest, we needed to prioritize an initial, measurable experience.

Organic real estate behaviors in Facebook Groups

Launch & learn

We designed & shipped a minimally satisfactory Real Estate category experience for Marketplace. Support for core shopper and seller features were built, tested, deployed and evaluated in under 6 months.

Facebook users could browse a feed of home listings or list a property they manage directly. Understanding that users expect a comprehensive view of available inventory, we partnered with home listing aggregators to include their listings alongside direct ones.

Real Estate category experience category feed, home listing composer

Early feedback

User research studies and feedback from inventory partnerships showed clear themes in reaction to our initial launch:

Low quality, limited totally inventory from direct Facebook listings
Lack of property details, neighborhood data & map browsing support
High volumes of low quality leads coming from Facebook Marketplace compared to other home inventory classifieds platforms.

Home listing detail page, property contact form

Visualizing our strategy

I helped the team clarify our focus by creating a visual roadmap during post-launch planning. This allowed us to explore solutions quickly and inspire product direction. I mapped two tracks of feature work:

Build trust: essential user-requested and inventory coverage features to improved perception of Facebook Real Estate in comparison to others.
Grow through differentiation: we couldn't simply match competitors' functionality. Instead, we must focus on leveraging Facebook's dynamic, feed-based, social commerce strengths.

Expected impact

Success was gauged by improvements in listing quality, increased engagement with properties on Facebook, and the quality of leads generated for property managers.

Early concepts to provide clarity in strategy through a visual roadmap

Solutions

We focused on developing promising features, sometimes undergoing multiple iterations to ensure satisfactory results based on customer feedback. The success of Facebook Real Estate didn't hinge on a single flashy industry feature, but rather on a series of targeted concepts refined through experimentation. We adopted a long-term approach, observing users throughout their home search to remove bias from short-term metrics fluctuations. Here are some highlights that contributed to realizing this vision...

Redesign of home listing pages for Facebook Design System updates

Interface refresh

During my tenure, I overhauled the Real Estate category to align with the new Facebook Design System ("Tetra"), initially aimed at addressing accessibility concerns. This initiative turned into a chance for designers company-wide to reimagine existing interfaces. I concentrated on refining typography and spacing to enhance clarity and hierarchy in home listings, which were previously cluttered and difficult to navigate.

Providing contextual neighborhood integrated into home listing detail pages

Neighborhoods

When deciding where to live, people consider their entire life, not just the property details. To address this, we partnered with local data providers to highlight key information on public transit, schools, and walkability. This required creating content patterns for incorporating third-party data in a lightweight manner.

At first, I explored the option of letting users toggle local places on/off in the main map view, but the number of pins was overwhelming alongside home listings. So, I opted for a tertiary map view specific to each listing. This way, users could focus on local data related to the listing they were interested in, rather than the entire market."

Providing contextual neighborhood integrated into home listing detail pages

Defining new patterns for multi-item listings

Originally, Facebook Marketplace had a design for single items with a price, description, photos, and category. When we extended support to apartment buildings, I had to reimagine the baseline design. There was also the question of whether each unit in a building should have its own home listing page. To address the volume and lack of precedent, I established rules to guide the solutions logically."

Listing detail pages are limited to 1 additional level of nested sub-views.
Group listings in the same building into a single home listing page.
Maintain consistent a visual language for building and unit attributes for a uniform appearance across properties, yet diverse in specifics.
Allow users to contact a property at the unit level.

Home listing contact form experience

Helping home shoppers express their interest

Inventory partners reported low-quality leads from Facebook. To improve this, I investigated ways for home shoppers to show nuanced interest by requesting tours and specifying their property preferences. To make this easy, I used quick toggle interaction patterns. Suggesting prompts reduces the need for manual typing, making expressing intent more effortless."

Sort & filter

Shopper-drive search curation

On average, a home search can take anywhere between 30 to 90 days, so it was essential to provide tools for shoppers to narrow their focus to the particulars they care about most in a home. We supported this through both push and pull features.

Pull: Early on, users filter inventory for availability. I added an inventory count of matching homes to build trust. My principle here was to take advantage of every small moment to help users feel informed.
Push: Let users create alerts for future matches due to the perception that homes go quickly. I combined filters and alerts entry points since they are intertwined tools that work better together.

Saved search alerts

Pull

Early on, users filter inventory for availability. I added an inventory count of matching homes to build trust. My principle here was to take advantage of every small moment to help users feel informed.

Push

Let users create alerts for future matches due to the perception that homes go quickly. I combined filters and alerts entry points since they are intertwined tools that work better together.

Rental search activity summary

Helping users pick up their search across app sessions

Early on, a major complaint was the difficulty in finding previously viewed home listings. With Facebook's constantly refreshing feed design, tracking down a post or Marketplace listing you had already seen was nearly impossible.

People solved this by creating spreadsheets with home URLs. To address this, I proposed a new activity summary section as a hub for all homes a user had engaged with. It organized homes into previously viewed, contacted, or shared categories.

I believed that we should focus on making the home search on Facebook effortless, so we should anticipate what users would want to do next as they continued their search.

Home listing composer

Enhancing direct listing quality through new patterns

We noticed a clear difference in the quality of direct home listings (via the Facebook app) versus those from partners, measured by the amount of photos and detailed property information provided. Initially, our approach favored simplicity over depth, using a basic Marketplace listing format. However, we recognized that real estate transactions require richer content. Research showed users desired structured details for better matching.

Consequently, I revamped the home listing composer to allow for comprehensive property information input. I streamlined data entry with quick toggles, and added advanced search for more detailed cases.

Stats

Impact

200K+ increase in active home listing inventory from post MVP features.
10% increase in home shopper leads for inventory partners from Facebook Rentals.
50%+ increase in quality of direct listings in comparison to partner listings, defined by # of structured fields.

Skill contributions

Visual design
Interaction design
Prototyping
User research
Experiment design
Strategy & planning
Design systems

Acknowledgements

Thanks to Nathan Garvie, Aaron James, Billy Sweeney, Caylee Betts, Mayank Yadav, Cristin Carey, Eric Shell, Adria Saracino, Farhan Saleem & Juliette Gilligan for their contributions and feedback that informed the many iterations of the design.