SPLITTING THE ATTOM
pre-mover leads— modeled public record data and listing data — generate leads that are too hot in the sense that a high proportion are all sizzle and no steak: people who may never actually move or whose move is at some indeterminate point in the future. The modeled leads specifically represent a list of properties that are likely to sell at some point in the near future, thus representing a potential mover. The likelihood to sell is based on an algorithm that takes into account a variety of factors that might include how long the property has been owned, how much equity the homeowner has and personal characteristics of the homeowner such as age, marital status, number and age of kids. In a typical modeled lead structure, all properties in a given market are assigned a “likelihood-to-sell” score and the top 10 percent of scores are designated as the leads.
This is all very sexy stuff, and many smart data scientists are getting much better at creating highly sophisticated predictive modeling to develop these leads. But at the end of the day, modeled leads fall short — at least as pre-mover leads; they work much better as listing leads or leads for real estate investors — because they ultimately don’t represent a person with an imminent intent to move, let alone a concrete moving date. These are folks who haven’t even made the decision to sell yet, so when they receive solicitation for moving services, it doesn’t fit their current situation or mindset. Faulty Assumption Foundation The second of the “too hot” leads are generated from homes listed for sale on the local multiple listing service (MLS). Not a lot of sexy data science or predictive modeling here, just a basic assumption that if someone lists a home for sale the occupants will soon
be moving out of that home. The only problem is that’s not a safe assumption. Only about 55 percent of all homes listed on the MLS end up selling, according to an analysis of MLS data by Clear Capital. Even for the 55 percent that do end up selling, the listing provides no information about when that will occur. That means companies marketing to the occupants leaving those properties are shooting at a target while blindfolded. They know the target is there; they just can’t see to nail the bullseye or anywhere close to the bullseye. Not to mention that the target isn’t even there almost half the time.
Time Travel Not Included The third traditional method for
generating pre-mover leads— sales deed data also obtained from public records data — is too cold in the sense that the lead by definition lags the actual move.
But at the end of the day, modeled leads fall short ... because they ultimately don’t represent a person with an imminent intent to move, let alone a concrete moving date.”
ATTOM Data Solutions • P16
Made with FlippingBook Online newsletter