What is Predictive Analytics?
Predictive Analytics is the creation of data points that are indicative of a likely future event. A single predictive data point can create an immediate actionable moment, while multiple points of data, i.e. a robust descriptive profile of a consumer, is the framework for a more accurate prediction of a specific moment, as well as, other multiple other actionable moments.
When we talk about big data, we’re often looking at descriptive data. Understanding a consumer’s age, location, job, and background tells us much about who that consumer is today. This information is used to create targeted customer demographic profiles for marketing purposes.
There are a variety of predictive data types. The simplest is event-driven marketing, which seeks to create sales prospects based on public life events. Divorces, probate proceedings, FSBO listings, and mortgage defaults can all identify a more opportune target market than simple geographic farming.
As straightforward as this tactic is, it’s still not yet a facet of many real estate companies’ marketing strategies. Blanket-farming a neighborhood is still a highly popular marketing campaign, even as it focuses on a consumer pool that may be 10 times less likely to deliver a customer in the short term than the targets that have been identified by a predictive event.
An early application of predictive analytics would be the process analyzing a user’s credit patterns and payment history in order to determine an individuals likelihood of making future credit payments on time. That is however simply the tip of an iceberg of possible applications.
The days of marketing to every homeowner within a city’s borders are long past. Selective targeting today allows a marketer to select prospects via an unnerving plethora of personal data points:
- Relationship Status
- Career, Industry, Specific Workplace
- Type of Home and Ownership Status
- Political Affiliation
- Household Wealth
- Past Travel Destinations
These kinds of descriptive data points serve as a framework for general demographics. At the confluence of a select demographic set and a career, relationship, or financial event, they quickly become sharp predictors of immediate sales potential–the holy grail of target marketing.
Predictive Analytics in Real Estate
Real estate brokers and agents don’t have to use complex data tools to leverage predictive analytics. There are a number of easily accessible tools that can be directly employed to draw out the events and interactions that portend a moment of sales potential.
And some companies are off to a great start. For example, in 2014 Redfin launched a consumer-facing predictive sales timing model dubbed “Hot Homes”. By using an algorithm that weighs over 500 local variables and recent sales statistics, it predicts and labels homes that will sell quickly. Properties with a 70 percent or higher likelihood of selling within the first two weeks on the market are labeled “Hot Homes” when consumers view them on the website.
Redfin’s tool uses big data to deliver preliminary insights to an individual consumer, much like Zillow’s Zestimate. They crunch more data than consumers ever could on their own, and deliver simple answers. The margin of error in these tools should generate a healthy level of skepticism from a consumer researching a single home. But used as research points for broader market results, these tools are informative. More than that, they’re marketing genius from an advertiser’s perspective, because consumers don’t see them as advertising. They view them as tools that make them more informed. Zestimates have been viewed over a billion times, and Hot Homes will likely garner a significant number of consumer eyeballs themselves over time.
What are you doing in the world of predictive analytics?
If you don’t know, or are not sure, grab a copy of the 2016 Swanepoel Trends Report (164 pages) – due for publication end of January 2016. The Report details a large number of these tools such as: ReboGateway, Brytescore, SmartZip, Zurple and others.