Why did Zillow Offers fail?
It didn’t have to.
I’ve written previously in the abstract about why Zillow Offers struggled: pricing homes is hard.
But now, three years after its failure, I want to write more specifically about the structural corporate pathology that I believe brought down Zillow’s nascent iBuying business:
Lack of ownership.
To be clear, I loved working for Zillow and would recommend them to anyone as an employer. I would eagerly work there again. All companies struggle and fail sometimes. And this is not about finger pointing. As you will see, there is no specific person at whom to point the finger. Instead, Zillow Offers had lots of highly talented people and leaders in a bad structure that made the sum less than its parts.
THE BACKGROUND
Zillow Offers began buying homes in 2018 as a copy of Opendoor. Zillow made offers to homeowners via their industry-leading real estate website and then sold them. They were basically a flipper but with very small margins.
The timing was unfortunate, as only two years later the real estate market would be convulsed by the COVID pandemic. Expecting a real estate plunge, Zillow paused all home buying in early 2020.
To the surprise of almost everyone, the opposite happened. The real estate market got white hot and took off. Zillow’s offers struggled to keep up with the market as homeowners could get better deals in multiple offer situations. Homeowners saw little value in a “guaranteed sale” program that lowballed them. Zillow was not buying enough homes in order to justify its cost.
Leadership decided they had to do something to keep up with the white hot market. The solution to low conversion was higher offers (a program Zillow code named Project Ketchup).
Project Ketchup is the consensus shorthand reason for Zillow Offers’ failure.
But this is not the point at which Zillow committed its fatal error.
Zillow had already committed its fatal error many years before.
THE FAILURE
Zillow 2.0 failed because of ownership.
The problem that predated even COVID was both structural and cultural.
In this case, no organization in Zillow “owned” the pricing result.
As Zillow set about to correct the conservative bias, four different departments were engaged:
· Pricing came up with a nowcasted CMA estimate of the home and now sought to include pending comp sale estimates and other tactics for mitigating conservative bias.
· Data science adjusted that price based on past performance. Because Zillow had been underpaying for homes, the offer was adjusted higher. They also appended new home price appreciation (HPA) onto resale forecasts.
· Asset management added their own gross price overlay to all offers on top of the other adjustments to improve offers.
· Resale was selling everything over list price, so the offer was further juiced in the underwriting to account for higher Sale Price/List Price Ratios.
None of these departments owned the result, however. There was not a single employee at Zillow who looked at a price sent to homeowners and said to themselves “I’ll sign my name and stake my reputation on the home selling for this number”.
One had to trace the organization’s chain of command all the way to CEO Rich Barton before finding a common hierarchical leader shared by all four organizations.
The result was that all four departments succeeded in solving the underpricing problem.
Separately.
Suddenly, Zillow was overpricing homes by the summer of 2021.
THE COLLAPSE
And then the market cooled.
In the first week of July, 2021, analysts began to notice the already aggressive pricing was egregiously inconsistent with comparable sales. The signs began to show that the market had peaked. Conversion on offers skyrocketed as homeowners began accepting too-good-to-be-true offers, an outcome that was initially greeted with applause from leadership as it had been the goal of Project Ketchup. But analysts took to Slack to raise the warning, but the response was that everything was “working as intended”.
Because no single leader owned the pricing result, Zillow was slow to react.
Exacerbating the problem, Zillow’s leadership culture emphasized data-driven consensus. While superficially an attractive method of decision-making, data-driven consensus is not highly responsive to rapidly changing conditions. Real estate has a very long sales cycle. It takes a month for a home to go from contract to closing and the data to be available. Leaders waiting for transaction data to confirm their decisions may be a month or more late to the decision, and that is assuming they can quickly get their colleagues’ consensus on the appropriate pivot.
Sure enough, in mid-August, six weeks after the signs of overpricing had been obvious to many, Zillow finally reacted and introduced a pricing pivot.
It was too late.
Zillow had, by its own estimation, overspent by nearly half a billion dollars in those six weeks.
Zillow stopped buying altogether and would shut down Zillow Offers and 25% of its headcount (including yours truly) before the end of the year.
CONCLUSION
Rich Barton, Zillow’s CEO, attributed the failure to volatility, saying “the unpredictability in forecasting home prices far exceeds what we anticipated and continuing to scale Zillow Offers would result in too much earnings and balance-sheet volatility”.
And…maybe that’s true.
But that is a strange reason to shut down what Zillow had previously described as a necessary response to an existential threat.
Undoubtedly, the existing, profitable business of Zillow.com made it easy for leaders to flinch when Zillow Offers experienced its first major challenge. Zillow, who had refused to burn its boats in pursuit of its new iBuying theory, got spooked and sailed back for safer harbors. Competitors Offerpad and Opendoor, meanwhile, had no boats to burn. They will live or die on the iBuying model and therefore had no choice but to see it through.
It’s also possible my analysis is tainted by domain dependence. I am in pricing. The problems with pricing are more obvious to me versus other parts of the business. I’m inclined to diagnose pricing problems as the primary business pathology.
But in the case of iBuying, I think that is correct. Pricing is the main effort of iBuying. By Zillow’s own admission, it was pricing volatility that was the critical point of failure.
If you have a centralized, data-driven, consensus-led decision making culture, consider Zillow’s result. That culture can work 95% of the time, until a Black Swan-like event like the COVID housing volatility demands rapid adjustments that you are culturally unprepared to make. It doesn’t matter if your business thrives 95% of the time only to be wiped out the other 5% of the time.
Instead, had ownership been structurally devolved to lower leaders, Zillow might have pivoted several weeks faster than what actually happened. Zillow might have mitigated the damage done by overpricing. And perhaps Zillow Offers would still be in business today.
I often wonder if Zillow has seller’s remorse regarding their iBuying experiment. Neither Offerpad nor Opendoor’s prospects are certain as of this writing. Both appear to have survived yet more volatility in 2022 and 2023. But are both trading near all time lows at 90% off their peak price, and have disappointed on revenue in 2024.
So it is still possible Zillow’s decision is vindicated.
But the reality is that Zillow never gave iBuying the shot that it deserved.
