Will AI ease resi agent pain points?

Sunnyvale-based 3D mapping firm Matterport has released its latest features, several of which employ AI to ease pain points for residential agents who use its system to showcase homes online.

“The level of data capture coupled with the generative AI means that Matterport can make the job of real estate agents easier and provide better outcomes for clients,” said Preethy Vaidyanathan, the company’s vice president of product.

The newly unveiled “auto-defurnish” feature, which removes unsightly furnishings and other decor from listing pictures and 3-D models, and “auto-description,” which uses AI to write property descriptions, are the two highest-profile additions now available to Matterport’s subscribers, who number more than 1 million. Those subscribers are often the listing photographers, not the agents themselves.

While brokerages already use third-party large-language-model AI like ChatGPT to create property descriptions, Vaidyanathan said Matterport’s new auto-description feature creates “the most powerful, holistic and automated description of every property” by understanding the property layout and dimensions of every room, including ceiling height, and by factoring in key features like fireplaces from images. It also has filters such as “aspirational” or “informative” that agents can use to set the tone for the listing.

The defurnish feature differs from existing virtual staging products because it also works in 3D, Vaidyanathan said, “allowing the users, whether they are a buyer, renter or agent, to walk around and experience a fully unfurnished space.” It is also the first step to eventually add 3D virtual furniture and decor, which is not ready for this release but is in development.

The defurnish and description features took over a year of research and several months of development to prepare for release and came as a direct result of “customer use case and need,” Vaidyanathan said. Other new Matterport AI features responding to those needs include an automated detailed report with room measurements that can be sent to the agent’s local MLS and business card capabilities within the virtual tour so agents can capture leads.

Compass’ Vice President of Northern California Kevin Patsel said the brokerage’s Compass AI tool, launched last year, can’t declutter virtual spaces but it does offer support for drafting automatic listing descriptions, as well as “designing effective marketing strategies, creating engaging social media posts, sending personalized emails and more.”

The firm’s internal AI “enables Compass agents to focus on human connections that drive their business while tapping technology to streamline marketing and communication,” he said.

The brokerage has spent more than $1.5 billion on tech in the last 10 years, he said, and continues to spend $100 million annually on research and development to create offerings that “involve solutions for managing agent workflow and administrative tasks.”

That ongoing investment combined with public and proprietary data will eventually lead to an internal AI experience that will be hard for outside companies to top, according to Compass’ Head of Engineering Shay Artzi.

“Over time, we expect our AI to be more personalized and to offer more unique insights than the insights that could be arrived at by point solutions like Matterport, which do not have a holistic view across the entire real estate transaction in the same way that is possible via the Compass platform,” he said.