Yes, big data usually infers ‘quantitative’ analysis, as in “relying on numbers.” The Zestimate legacy of profound inaccuracy finally reached a devastating conclusion with the collapse of Zillow Offers this week and the loss of hundreds of millions in shareholder equity. Zillow never figured out the qualitative part that enables the actual precision in the pricing of a home sale.

There is a lot of talk right now about how other iBuyers are continuing to buy and sell properties so the space is still viable – business as usual. But step back for a moment and think about this:

  • The iBuyer market is currently overcrowded, even with the loss of Zillow Offers.
  • The iBuyer bold-faced name is OpenDoor who was the unicorn of Softbank who famously backed WeWork without any apparent due diligence.
  • The iBuyer segment is characterized by its razor-thin margins and billions of investment required.
  • It was created and run in a rising market, most of it a boom, and was recently turned off during the recent downturn.
  • It is wholly dependent on housing markets with homogenous housing stock and will always need high volume just to survive.

I feel pretty confident there will be further fallout over time, but the iBuyer space will settle into a small segment of the overall transaction universe. It has been wildly overhyped (at real estate brokers and real estate appraiser’s expense) as investors, burdened with high volumes of capital, desperate for upside in housing in this fintech boom.

Tags: , , ,


3 Responses to “Zillow Offers As A Proxy For ‘Big Data’ Shows The Lack Of Qualitative Analysis”

  1. Anna Coatsworth says:

    As always, Jonathan, I agree with you 100%! I knew it was just a matter of time before the iBuyer segment imploded, with Zillow having absolutely no clue what they were doing and no appreciation for the fact that it is an art as much as a science to determine a property’s correct pricing. This is a win for those of us who take our profession and craft seriously, and who actually know what we’re doing.

  2. Vince Slupski, MAI says:

    It seems like the appraisal industry will be quick to dunk and taunt Zillow for the perceived failure of its Zestimates. As Mr. Miller pointed out in another blog post, Zillow boasts median accuracy of 7% if the property is not on the market, 2% if on the market (and presumably the hard work of pricing has been performed by an agent with a CMA). Are appraisers really any better? Would the median appraisal accuracy be 7% prior to property being placed on the market? And how much accuracy is really necessary for an 80% loan, or a 50% loan? How much is the lender willing to pay for a few percentage points of accuracy on a 50% loan that’s in a large portfolio of loans?

    • Jonathan Miller says:

      Thanks Vince. I appreciate the insights but I think you are missing the broader point. The trend towards automation and the big data narrative has now been clouded by this failure. Expectations of financial institutions have been dominated by the idea of quantitative solutions when in reality, the qualitative is missing. For a home equity line, automation already dominates for reasons you describe. But your logic is only valid if low level of precision is implied.