SAN FRANCISCO — Have you ever felt your eyes begin to cross after hours of sifting through listings on the MLS?
Trust me — your clients have, too. In a time where both consumers and real estate agents have so much information at their fingertips, it can be overwhelming. We begin to have trouble recalling special property features, what listing has what, or even how many listings we looked at.
RealScout co-founder and CEO Andrew Flachner likened it to sifting through a jar of jelly beans at Inman Connect San Francisco last week, noting that the difficulty of recalling property features is not unlike the difficulty of recalling one among hundreds of specific textures, colors or flavors.
“You wouldn’t want to mix up a buttered popcorn with a piña colada,” he joked as he drew the correlation of mixing up an open layout with a galley kitchen.
Flachner makes the point that while jelly beans may make us laugh, the differences in property features that are arguably just as difficult to identify can be “potential dealbreakers for clients.”
And just as listeners were feeling discouraged that they’d ever be able to recall many of the slight differences that are so important to their clients, Flachner brought in the role that computer vision plays in the process.
Don’t be so quick to think that the bot has finally beaten the human. Computer vision — much like the technology used in the photo tagging feature on Facebook — is, ironically, something that must be created by humans.
Flachner said the “ultimate goal of computer visioning is to model, or replicate, human vision using software and hardware at different levels.”
However, humans still have an extremely important role in determining the accuracy of computer vision. Or, in English, “hundreds of thousands — sometimes even millions — of photos must be processed by humans in order to train the model to identify people, faces and room types like kitchens, dining rooms, living rooms, bedrooms, bathrooms, etc.”
So, what will this future of computer vision begin to look like in real estate? Well, given that consumers crave a highly photo-centric homebuying search, computer vision really may be the future of what search looks like in our industry.
Computer vision can effectively shift the homebuying experience from sifting through thousands of photos to an experience where consumers can simply search by feature.
For example, Flachner’s company reinvented the photo carousel on property pages to allow consumers to simply click on the room type they’re interested in searching.
Why does this matter? Well, without computer vision, search had traditionally been reliant on agent inputs into the MLS.
When a computer (with human assistance) can identify room types and features of high-quality photos, the experience is enhanced. Consumers are able to more easily identify properties that fit their needs and interests, and agents are able to provide better quality information.
But Flachner understands that home search is more than just searching, and more about comparing. RealScout released a “compare” feature as well, so now consumers can not only search their desired features but compare properties side by side, ultimately, searching and identifying properties more effectively.
What does the future look like? Flachner argues that we are in a world where the “home search experience will once again change.”
The question simply is, “who will usher in this new era of home search?”
I’ll put my money on Flachner being the one. He understands that this technology, while still in its infancy, presents a huge opportunity for our industry, yet he values the human connections that he creates.
I’ll be keeping my eye on him and his partnerships, and I think you should, too.
Alyssa Hellman is the director of the Go School of Real Estate.