Artificial intelligence - although adopted by some of the biggest tech companies in existence today is, to some, still a futuristic concept that hasn’t yet arrived, says Dominik Pogorzelski, the head of product management at leading Spanish real estate tech company Restb.ai
The VC-funded machine learning company specializes in visual recognition for real estate and automatically analyzes, tags and classifies property images with real estate-specific information. Recently the group sponsored the Property Portal Watch Conference in Lisbon.
"I think one of the things that shock us is that people still think that AI is this thing that is coming – it’s a futuristic thing. But the reality is that artificial intelligence, computer vision/image recognition - it’s available today and it’s quick and affordable.”
Restb.ai offers a solution to real estate portals who have databases containing millions of property images. Porgorzelski says often portals don’t know about the specifics of their photos and this is where image recognition can help with a website’s SEO.
“We basically scan through your images and tell you: ‘this image is a living room and in this living room, you have a fireplace and hard wood floors’” he says.
“It’s these images that are often one of the biggest digital assets.”
There are three ways that portals can benefit from using image recognition technology to drive their site’s performance, he says.
Generate SEO traffic
“By tagging your images you’re basically able to optimize the SEO so that Google sees the relevance of your images and it ranks your listings higher in order to generate more organic traffic,” Pogorzelski says.
Increase user engagement
“By knowing what’s in an image, you’re able to make a lot of interesting user experience improvements in order to capture the interest of people,” he notes. “So you show the image that the viewer is most interested in, instead of some default random thumbnail.”
Create new revenue streams
The more control a portal has over their imagery, the less likely it is that these images can be stolen by a third party, Pogorzelski says. It also enables a portal to create new revenue streams.
“A lot of portals are being hijacked by third party sellers without knowing about it because an image has a watermark on it and when you have millions of images, you can’t cycle through all of them,” he says.
“We’re able to detect every single image that has a watermark and tell people who is using their real estate, technically speaking.”
And the best part about this is that it can lead to monetization.
“Once a portal has understanding of who is using their images, it paves the way for them to start monetizing by charging for usage.”