- User customization is the key to consumer-facing algorithms, but agents can use them to their advantage, too.
- According to Zillow, 47 percent of sellers reduce their price before selling, and they receive an average of 2 percent less.
- Overpriced property takes 113 days longer to sell on average, but customizable algorithms can help fight the overpricing plague.
We are in the midst of an extraordinary sea change. Big data and the burgeoning field of data analytics are transforming how real estate is done.
Big data assembles millions of data points, analyzes them and uses predictive analytics to translate that data into actionable steps for our clients and our businesses.
The name of the game is user customization.
My article last week, “The death of competitive market analyses as we know them,” fired up the Inman readership. Comments questioned whether I had ever sold real estate (I’ve been a broker since 1982, and I’ve run the training for 4,000 agents), complained about the inaccuracy of algorithms, raised issues about whether appraisers would ever adjust and questioned the definition of “best” agents. Other commenters agreed with my assessment.
Algorithms — friend or foe?
Granted, algorithms have their limitations, largely due to how statistics work; we can use statistics to illustrate what is happening at a macro level, but they might be way off on a micro (individual house) level.
To illustrate this point, Zillow says that its Zestimate values are accurate plus or minus 7 percent, 95 percent of the time. In other words, if your Zestimate says that a property is worth $200,000, there is a 95 percent chance that the subject property will sell between $186,000 and $214,000. There is a 5 percent chance that the home will sell outside those parameters.
Skilled agents can generally exceed this degree of accuracy, but this occurs only when the agent has a deep knowledge of both the inventory and the lifestyle in this area. This group is primarily composed of the 10 percent of agents doing 90 percent of the business.
Zillow makes its Zestimates obsolete
Zestimates are a 10-year-old technology. If you haven’t visited Zillow recently, check out its “price your home” feature.
As I mentioned in last week’s column, the new algorithms allow for user customization. Its “price your home” feature is the slickest pricing tool that I have seen because it lets you adjust the algorithm to make adjustments automatically for square footage, location, condition, bedrooms and baths plus compiling the photos for all comps as well. It’s free, and you don’t have to register.
The question you must answer for your business is: “Can your current CMA tool allow you to make all of those adjustments and arrive at a price in less than two minutes?”
Yes, you still need to check your MLS data for accuracy and comps that Zillow didn’t cover. Nevertheless, when sellers use a Zestimate to object to their listing price, use Zillow’s “price your home” feature to adjust the algorithm for the best comps and other factors. There’s nothing like fighting the Zestimate battle with Zillow’s newest tool.
Use Zillow’s big data to illustrate to sellers of the perils of overpricing
In Zillow Talk, Spencer Rascoff outlined how they used data analytics from over one million homes to reveal the best time of year to list a house, words that can enhance or harm your marketing efforts as well as the perils of overpricing.
According to Zillow’s analysis, 47 percent of all sellers reduce their price before they sell and receive an average 2 percent less.
Overpriced properties take 220 days to sell on average versus 107 days for properties that are priced properly. This data backs up what agents have always known but have been hard-pressed to explain — overpricing results in a lower price and a much longer time on the market.
UtilityScore (MyUtilityScore.com): A case study using big data
I recently sat down with UtilityScore founder, Brian Gitt, who said a number of people had complained that UtilityScore wasn’t accurate for their property. As noted above, the macro data should be pretty accurate though individual house data can vary substantially.
The point is that buyers almost never see utility data until they place a property under contract. Both buyer’s and seller’s agents can now share this data with clients as a preliminary comparison that can be documented accurately with actual data on the subject property at the point of listing or sale.
Again, the customizable nature of the algorithm educates owners on how their expenses change based on where they set their thermostat, the number of people in the household and whether someone is home during the day.
There’s no simple, accurate way to project these changing numbers without using big data tools.
UtilityScore’s big data analytics go beyond just utility data by:
- Making recommendations about qualified professionals and verifying their licensing status
- Screening local reviews and combing social media to locate both negative and positive comments about which contractor to choose
- Identifying which home improvement products have the best reviews, the greatest efficiency and then matching that data with local rebate programs
- Providing green financing options that can assist owners in making repairs or upgrades to their homes, even when they might not have the cash or the credit to do the work
Today’s buyers and sellers want simplicity and customization
The tools we see today are merely the first wave of what’s coming in the future. Consumers want tools that are simple to use and allow them to make the adjustments that fit their unique needs.
Rather than fighting the trend, seize the opportunity as a way to have a competitive advantage over those who are slower to adapt. As my former boss, Jon Douglas used to say, “All we have to do is to be six months ahead of the competition, and we will dominate the market.”
Bernice Ross, CEO of RealEstateCoach.com, is a national speaker, author and trainer with over 1,000 published articles and two best-selling real estate books. Learn about her training programs at www.RealEstateCoach.com/
