Best of US Homes Listing Predictive Analytics Program
We’re experiencing a market inefficiency in the luxury home market here in Birmingham, AL. We at Best of US Homes believe it’s driven by the Baby Boomer Sell-Off. Over the past six years we’ve experienced a 50% sell through on homes listed for $750,000 and up. Yet there are some homes that sell in less than 30 days, some in less than a week, but half linger on the market for over a year, some for four to five years.
We’ve been studying the use of Predictive Analytics in the retail, insurance, the used car market, the stock markets and telecommunication industries and believe that this same technology is adaptable to the real estate industry. In the past, we’ve relied on Comparable Listings to assist our listing clients in pricing their homes. Using Comparables we found homes that were of comparable size, in the same neighborhood and had similar amenities and concluded that our new listing should carry a similar price. But if 50% of the homes are not selling, what is the value of a Comp?
It’s evident over history that people buy homes based on trends, the same way they buy automobiles. Being raised in the 50’s I saw the flight to the suburbs and the switch from city two and three-story homes to the ranch home on large lots. Then as we moved into the 70’s those ranch style homes were replaced with bigger two-story homes on smaller lots and then in the 90’s the Baby Boomers started building their McMansions behind gates adjacent to a golf course. We’re now experiencing a migration back into the city, but the well-healed Generation-X’ers continue to buy the McMansions, there just aren’t enough of them to consume the high inventory being sold by the Baby Boomers. So, it’s important that sellers know exactly what the few buyers that do exist are looking for and what sells and what doesn’t sell.
This can be predicted using Predictive Analytics; there are six tool options that can be used to produce predictions of future behavior, logistic regression, neural nets, decision trees, Bayesian classifiers, support vector machines, or genetic programs. These tools ate referred to as algorithms. Not all algorithms create equally accurate predictions when applied to the same data sets, so the modeling must be suited for the data that the modeler has available and the type of prediction that he’s attempting to produce.
The easiest parallel to PA in real estate is what Kelly Blue Book has done for the used car industry. Kelly Blue Book gathers resale data on millions of used cars then sorts them by make, model, condition, mileage, and area of the country. This information makes it possible for you to take your used car into CarMax and within 30 minutes get an offer to purchase your car. This same technology can be adapted to the used housing industry.
Here’s how Best of US Homes Predictive Analytics works in the luxury home real estate market.
Predict how much your home will sell for and how long it will take it to sell.
We download specific data on all the homes that have sold in a neighborhood over the past six years from our MLS sales data. This gives us approximately 150 data points (size of home, cost per square foot, number of bedrooms and bathrooms, number of fireplaces, garage spaces and location of same, size of lot, pool, lake front, golf course location, etc.). We then scan the pictures of the sold homes and add an additional 150 data points. These would include such things as color of kitchen cabinets, surface of kitchen and bath countertops, type of lighting used in the kitchen, use of foe painting, color and coverings of floors, ceiling molding, media room, finished basement, wine cellar, type of roof, type of construction we also pull keyword features from the description of the home.
We can then measure this data against the number of days the house spent on the market, and how close to list price the house sold. We find that specific data points cluster among the homes that sell quickly and closest to asking price. Doing this by neighborhood allows us to then enter our listing clients home data into the algorithm and give them an estimate of how much their home will sell for and how many days they can expect it to take to sell their home. We can then modify their data input; we might paint the kitchen cabinets and install granite counter tops for example, and tell them how their sales projection will change with any improvements or upgrades made to their home. This also gives a ROI for any of the changes.
We can also refine our projections by entering the data on all the other homes in their neighborhood that are currently on the market. This direct comparison will be measured against the number of expected buyers, again derived from the past years sales records, and tell our listing clients which houses will sell and when their home will sell relative to their competition.
Our next step will be to move ahead of the market by monitoring social media using text analytics to determine any change in trends in the real estate market. Zillow is already doing text analytics by mining the descriptions of listed homes and finding what keywords appear in the descriptions of those homes that sell fastest and closest to the asking price.
Analytics powers almost every facet of our business, including home valuation processes, market research, customer analytics, financial forecasting and demand planning.
What’s in the future?
Best of US Homes predicts that the CarMax business model will make its way into the real estate industry some day in the future.
It’s being done by Opendoor.com in Phoenix and Dallas in the $200,000 to $500,000 housing market. In the past two years, they’ve helped over 4,000 homeowners buy and sell homes and they’re currently transacting over $60,000,000 in home volume each month. Opendoor’s services are notable because they focus on two problem spots for home-buying: Deferred maintenance and inspection. Fixing up wear and tear is often challenging for home sellers because they don’t have the cash up front. Buyers aren’t usually open to taking it on either. To help with that, Opendoor takes care of inspecting, designing, and fully fixing up the house before listing it.
Kerry Grinkmeyer is a real estate agent in Birmingham, AL specializing in the marketing and sales of luxury homes. Kerry's a retired financial advisor, he sold his firm, one of the largest in the Ameriprise Financial Advisor system in 2005 to his son, daughter and nephew. Now he's building one of the largest boutique real estate agencies that he'll eventually sell to his grandchildren. He the author of the children’s book The Christmas Web- A Family Christmas Tradition. Kerry competes in the Senior Olympics in to 50M, 100M and 200M dash as well as the 5k and 10k time trial cycling events.
Kerry offers you his financial background, knowledge of the community, love of business and family and energy to assist you in one of the most important financial decisions you'll make in your lifetime.
205 919 6006