Imagine a world where agents never have to enter listing data into the Multiple Listing Service again. Instead, either the listing agent or their designated professional real estate photographer uploads the listing photos directly into the MLS.
Agents then review and approve – or make minor corrections – to the listing data and the photo descriptions that it automatically creates once the images are inside the MLS. Productivity immediately increases as it frees agents from the mundane tasks of writing short captions for 30 or more property photos for every listing.
Collectively, this saves hundreds of thousands of hours. As a result, agents can spend more time doing what they do best: working directly to assist their clients and prospects. Does this sound like science fiction? Hardly.
Several new technology firms are bringing the power of computer vision and AI to real estate today. One of the most significant immediate benefits of this cutting-edge technology is the automation of listing data.
Photos contain more than an image
Digital photos shot with almost all digital cameras and even smartphones contain more than just the image. Photos feature metadata that are attached individually to each image. This data is a collection of information, including the photo’s size, dimensions, and, most importantly, location. The date and time are also typically included.
Digital cameras, such as DSLRs, record the shutter speed, aperture, ISO speed, white balance, focal length and camera model. When a photographer captures a 360-degree tour with a floor plan, even more highly valuable real estate data is obtained, as every point in the house is mapped. Add in drone footage, and the power of data potentially collected through photography – and its potential use – is immense.
Most importantly, digital photos contain the precise location for real estate use, as it provides GPS coordinates. This is how a property address can automatically be identified and captured. If you take photos with your smartphone, you can see this information for yourself. Each image contains EXIF data.
If you have a Windows computer, right-click on the photo in question to view the EXIF data and select “Properties.” For a Mac, open the image in Preview, click on “Tools” in the top menu bar, and select “Show Inspection.” Both methods will display the data captured with the photo.
The next step is the MLS
MLSs are already testing out this technology. One of the early adopters in this space is Greg Moore, Chief Technology Officer at Regional Multiple Listing Service, Inc. (RMLS), one of the largest MLSs in the Pacific Northwest.
Moore piloted a test program with one of the leading tech firms in this space, startup Restb.ai. Restb.ai is already processing more than 10 million photos a day. By using computer vision and AI, Restb.ai can automatically “see” the image, allowing it to automate many tasks.
Restb.ai can automatically create the appropriate photo identification tag for each image, an ADA requirement for every website. Its advanced computer vision technology also can determine property conditions, find visual similarities (allowing someone to search for only listings with “black and white kitchens”), detect duplicate photos, and, most importantly, create image captions and descriptions.
Moore’s pilot at RMLS leveraged the power of Restb.ai to demonstrate its ability to describe photos accurately. Moore also is convinced, after the pilot, that photos are the future for capturing listing data.
“It really turned a lightbulb on for us,” Moore said when describing the Restb.ai project. Moore explained that once you have the property address, that’s just a starting point. Public records — tax data, past sales data, property improvements, school district data, and more — can be instantly connected to the listing information. It all starts with the upload of a property photo.
“As the MLS, we want to be the first point of entry,” Moore explained, noting that agents can still control the order of the listing photos and make any necessary tweaks, but computer vision will do the lion’s share of the work.
During the RMLS pilot, Moore noted that they didn’t even send the actual photos to Restb.ai but URLs to the images. “I mean, seconds later, we got the content,” he said.
Power of photos
Snapchat users are creating more than 527,000 photos every minute. Images are driving social media and internet use and engagement. Google Photo users are uploading 1.2 billion images every day.
In real estate, agents understand the power of photography. Research has shown that homes with high-quality photography sell 32 percent faster. Homes with more photos sell faster too, and for more money. In February, the photography website PhotoUp reported that 72.2 percent of agents said that high-quality photography helps them win more listings.
PhotoUp also cited research that 83 percent of buyers said pictures are very important in assisting them in deciding which homes they will visit. In addition, agents who contract professional photographers earn twice as much commission as others. The use of computer vision and AI are significantly increasing the value of real estate photos to agents. And as Moore notes, this is only the beginning.
Joe Jesuele is the founder and CEO of HomeJab, a leading on-demand professional real estate photography and video marketplace for real estate pros, and architect of the real NFT Marketplace. Follow him on LinkedIn and Twitter.