Eddie Krebs was serving as CIO of an Atlanta real estate broker agency when he noticed how difficult it was for agents to track any of their online leads. He saw potential in some of the agency’s in-house technology that studied the purchasing patterns of potential clients and bought the intellectual property.
Now with his behavior analytics startup Brytecore, brokers can follow digital leads from tepid to fire hot with an easy-to-use platform.
“We use machine learning to predict when home buyers will make their purchase decision,” says Krebs. “The data that fuels our machine learning is online consumer behavior. As a consumer is looking at a real estate property, as they are saving and sharing, all of that information gets pushed into our machine learning platform to look for patterns of behavior.”
“We match the new consumer’s behavior against these collections of patterns that we have from previous customers.”
The machine learning platform invites their primary target audience, real estate brokers, to add Brytecore’s IDX (a Javascript widget) to their websites and mobile apps to track interested home buyers as they navigate their site.
“When the prospective buyer requests an agent or asks for information on a property, we pass that down to an agent and we give the agent the background and history on that person to help them understand,” says Krebs.
That background may include what the potential customer’s price range is, neighborhoods they are interested in, property types, and other data points. The goal is for the agent to have essential information as well as how the lead has progressed over time to make a faster, more effective sale. The technology also classifies each lead into specific profiles like first-time home buyer to investor — making it easier to automatically assign the buyer to the best agent.
According to Krebs, they are analyzing about six million data points a day, live.
“One of the major problems that we have in our industry is with internet leaks — people who are online and end up requesting help from an agent,” says Krebs. “People spend a lot of time online before they actually make a purchase decision. At some point while online, they go and they register for an account so that they can save that property.”
“We continue to follow that consumer and when their behavior online says that they have really made that purchasing decision — when their search history and activities signal that — we will notify the agent. That’s the key to what we do because that really helps everybody in the process. We’re currently tracking about two million people a month.”
For brokers to get access to these analytics and lead boosters, Brytecore follows a very simple subscription revenue model of $500/month plus a fee for each hot lead. The platform handles all leads and automatically finds, assigns, and texts the right agent for the prospective buyer.
While Krebs has seen the rise of machine learning in the real estate industry, he says that it has been mostly in the listing side. He hopes to dive deeper into this technology to also add it to the customer side, such as an AI virtual assistant-like product that allows you to search properties by voice.
“Based on who you are, where you live, what’s important to you, instead of you searching online for a property, we’ll just let you know when one pops up that’s gonna fit your needs, that you’re gonna like,” says Krebs. “That will get smarter and smarter.”
After raising a seed round last spring and presenting at Venture Atlanta in October, Brytecore is concentrating on product development and scaling their reach. Krebs believes they will seek out their next round in Q4 2019 to boost growth.