We recently sat down with Tom Gamble, the co-founder and CEO of TKI, a software development company that is using artificial intelligence and predictive analytics to power sales enablement tools for real estate professionals.
I wanted to start with a simple question, what is predictive analytics?
Several years ago, many people were using the term “big data.” While it became a buzzword, this was actually when the use of data went mainstream as industries like retail, financial services and automotive studied everything from sales to inventory statistics. They overlaid it with demographic information and customer habits and started to predict buying patterns. They eventually utilized artificial intelligence to parse data to make highly sophisticated assumptions allowing for better target marketing, promotions, in-store and out-of-store experiences. Back in 2012, Forbes ran an article with this headline, “How Target Figured Out a Teen Girl Was Pregnant Before Her Father Did” That was when predictive analytics went mainstream.
How does this play out in real estate?
We have been using data for years in real estate. About 10 years ago, we were one of the first to harness available information when we built CBx for Coldwell Banker. If you remember, CBx shared where prospective buyers were coming from which was especially useful to determine and target feeder markets, and, from a demographic perspective, know who they are. For example, 40ish young couples with two kids with a household income of $100,000. This was a game-changer for so many agents who used this information not only to get a listing, but then use social and digital advertising to target that potential prospect pool.
We then worked to close a huge intelligence gap in the highly desirable “seller” prospecting processes as more and more data became available. With three years of incredible data gathering, mining, crunching, learning and adjusting nSkope, we now have more than 300 data points and a proprietary algorithm that is able to predict which homes will come on the market within 6-12 months.
Here is an example. West Deptford, N.J, is a Philadelphia suburb with about 20,000 residents. Like everywhere else they have a listing shortage. Teams and brokerages are flooding the entire community with “now is a good time to sell” messaging. They are wasting a lot of money because at least 80% of the households have no intent to sell. But nSkope recently ran all of the town’s households through an algorithm and predicted that 1,432 homes might come on the market over a 6–12-month period. In this example, 11% of these predictions came true. Compare this to traditional lead conversions of 1-2%.
One of our clients, the Nancy Kowalik Group, serves West Deptford and previously spent a small fortune on farming materials, online and traditional advertising and email campaigns with the hopes of getting the phone to ring or capture an email address of a potential seller. Using nSkope, Nancy and her team now focus on the homes that are most likely to list. Not only has she reduced her marketing spend by targeting ONLY these potentially sellers, she is also turning leads into prospects using her own lead cultivation system and tools.
How does it work?
It’s essentially the same as what retail has been doing for years. We understand the triggers that historically drive people to sell their homes – relationships, marriage, kids and more kids, job changes, empty-nesting, illness and others. But it is almost impossible to track what is occurring in the 5,000 households in a town like West Deptford without the help of technology.
We started by analyzing a multitude of available data points to determine which are right for real estate. We then built appropriate algorithms, tested and re-tested them a million times to determine the right mix and weight of each data point. The artificial intelligence elements learn as we go forward, making predictions stronger and stronger over time. And we continuously look for new household data that we can throw in the mix to enhance not only the predictive models but also to give sales people more contextual prospective leads. We set a high bar for ourselves. Obviously we want to be above standard market conditions of the number of homes that typically go on the market. For example, there are about 80 million owner-occupied homes in the U.S. and last year there were about 6 million sales. That give us a 7.5% annual turnover.
Let’s round that up to 8% and carry it to a local community. Eight out of 100 homes will sell. But which ones? And from a time and marketing expense, how much is needed to reach all 100?
Predictive analytics tools like nSkope reduce the pool. We might predict that only 20 of the 100 homes are actually in play. If eight of them actually come on the market, the original 8% potential success rate now jumps to 40%!
Now that you have lowered the pool of potential clients, what should brokerages and agents do with it?
Predictive analytics in real estate is relatively new, but previously the industry has looked at the spreadsheet of potential listings as leads. They are not leads.
I’ve been in real estate for a long time and remember the early days of the Zillow and Realtor.com leads. We eventually got to the point where it seemed like there were a billion leads for only 6 million sales. After the initial excitement of the lead gen era, agents saw most of these leads were useless and stopped answering. Brokerages picked up the slack and started to invest in “scrubbing” leads for the agents. This is obviously important, but really costly, too. CRMs evolved and we continue marketing to these email addresses.
At the same time, few were focused on listings – the breadwinner in real estate.
Not only does nSkope help you know who might sell, we go a major step beyond. We share that potential seller’s profile so agents can understand why we believe they are potential prospects. This allows agents and brokerages to use the right imagery, content and relationship building tactics. But it still requires that the work be put in. While nSkope reduces the playing field and has proven to be effective, the data by itself does nothing. It’s what you do with the data that matters.
We don’t usually like to talk about others because real estate is a huge industry with room for different ways of adding value, but one thing that differentiates us – and we are very proud of is that, from the beginning, we wanted nSkope to go beyond the data and provide context as a storyteller. We have stayed laser focused of giving brokerages and teams great information so they can use their own marketing tools, processes, materials and software for the outreach. As we know agents have their own ways of farming, we just give them targets and suggest where they should focus
Do have tips for those who want to use predictive analytics?
There are a few really important ones. The first is to make sure that you are going to use the data and are willing to create a process to work with these prospective prospects. Otherwise, don’t waste your time and money.
After that, you really have to go through the addresses and potential prospects carefully. Someone on your team, or in the brokerage, likely already knows them. This allows you to renew relationships. But remember, because we are predicting 6-12 months out, you may be in touch with them before they even know they are going to move. Therefore, don’t use hard selling techniques. But certainly engage in social, email, mailings, etc. to be in front of them.
Also, predictive analytics allows for true 1:1 marketing and engagement using intelligent profiling/segmenting. For example, if nSkope identifies a potential prospect as an empty-nester, don’t use imagery of young families. Instead use appropriate images and share content about where other empty-nesters have moved, successes with other empty-nester clients and have past clients serve as referrals. You also want to act quickly. Our research shows that on average most of those we predict to go on the market do so within 6-8 months. Get in front of them ASAP.
And, most importantly, as I shared before, predictive analytics is a guide to help you succeed. Without the hard work and marketing efforts, the data by itself does not win you business.
A reverse mortgage is a type of consumer home loan that allows homeowners ages 62…
JPMorganChase has said it’s cutting jobs in its mortgage lending business amid tightening monetary policy…
Pending home sales increased in May, breaking a streak of six successive monthly declines, but…
Industry Coach Tom Ferry Headlines Event JPAR® – Real Estate, America’s #1 fastest-growing, 100% commission…
Reputation in real estate is everything. The team at Custom Fit Real Estate will tell…
Workforce planning can help you meet organizational goals in any economy. It’s been a challenging…