Learning from Zillow

Lack of Focus and Long-Term Thinking

Nov 09, 2021

If you’ve been following the news recently, you probably noticed that Zillow recently shut down Zillow Offers, its algorithm-powered home-flipping division.

First time flipping houses meme - Product Thinking by Kyle Evans

If you’re unfamiliar with this type of technology, let’s do a quick overview.

(Want to listen? Check out the podcast version below:)

There are many companies in the iBuying space—these are startups or large companies that buy and sell homes “instantly”. An iBuyer will make an offer for your home based on its assessment of market value, cutting down the long process of selling. Other companies, like Opendoor, have been doing this since ~2014. Zillow entered the fray in 2018 with Zillow Offers.

These companies then do whatever work they deem necessary to get the home ready to sell, and list it. Large-scale flipping operations.

Zillow announced at the beginning of November that it would pause home purchases through Zillow Offers, laying of 25% of its workforce, and taking around a $500 million writedown on the business as it works to offload thousands of homes for which it overpaid.

How Did Zillow Get Here

Much of the home buying business for iBuyers is based on algorithms. At the beginning of 2021, Zillow Offers wasn’t aggressive enough, especially in a hot housing market, so they adjusted their algorithm to be more aggressive. As the market cooled, Zillow found themselves overpaying for houses.

According to Bloomberg:

It’s become clear that Zillow misjudged the housing market, making more aggressive offers just as competitors Opendoor Technologies Inc. and Offerpad Solutions Inc. were growing more cautious.

The volatility proved too much, and rather than continue to adjust its process, Zillow opted to exit the iBuying market.

So, what can we learn?

Lessons From Zillow Offers

Need for Focus

For almost all of its history, Zillow has been a real estate marketplace. It has had broad appeal because you can get information on home prices and values (something many of us do, even just to see for fun), create and browse listings, and get connected with real estate services. This has been the bread-and-butter.

With Zillow Offers, Zillow moved to become a market maker (and also became a risk taker, though less by choice). They wanted to facilitate transactions and take a cut of the profit. Which, frankly, is needed in real estate. It is extremely illiquid and takes a long time to buy and sell. Though, whether iBuyers will end up solving this or exacerbating it is still an open question.

But Zillow was caught in two worlds, and never seemed willing to focus sufficiently. It is great to expand into new businesses, especially when you have competitive advantages. But ensuring you do it right is critical.

With the fallback of its core marketplace business, its possible (and even probable) that Zillow didn’t feel the pressure to focus intensely on its new business unit.

As Ben Thompson discusses:

What is also worth noting is that Opendoor has two more advantages that come from being a startup laser-focused on home-buying: first, while Zillow started with a Zestimate tool that was about attracting customers to the top of the real estate funnel, and thus only ever had to be directionally correct, Opendoor knew from day one its entire fate as a business rested on its model’s accuracy; it’s definitely plausible to imagine its accuracy being much better as a result.

Many of us have been involved in situations like this, though probably not as large or extreme. Our company wants to invest in a new feature or product or opportunity, but isn’t willing to fully commit, either through investing in enough people to create the product or enough time to fully see it through.

Which leads us to:

Long-Term Perspective

I’ve written about Taking the Long-Term View frequently. Whether it is building a meaningful business or exploring the unknowns of space, we need to have a long-term perspective:

Space exploration has generally been about long-term thinking. It takes a long time (speaking in terms of our software-driven world) to plan and execute missions that send robots or rockets or probes to other worlds. And the payoff is extremely uncertain. You fail a lot.

And again in Planting Trees:

The tree can’t win in that type of environment. A tree is a 10 year investment, and so few companies have the vision or willpower to see that far ahead. To understand that success isn’t just about immediately increasing sales or increasing stock prices, but about creating something meaningful that will last. It’s about planting a tree that will turn into a mighty oak.

The long-term view is something I’ve admired about Amazon, despite its flaws. It will make big bets that won’t pay off immediately. But it has the fortitude to see them through.

Zillow, for better or worse, had a few rough quarters and decided to shut it down. From the most recent earnings call:

We have been unable to accurately forecast future home prices at different times in both directions by much more than we modeled as possible, with Zillow Offers unit economics on a quarterly basis swinging from plus 576 basis points in Q2 to an expected minus 500 to minus 700 basis points in Q4. Put simply, our observed error rate has been far more volatile than we ever expected possible and makes us look far more like a leveraged housing trader than the market maker we set out to be.

There is something to be said about the complexity of housing, and for the willingness to shut down things that aren’t working. But can you tell if it isn’t working after a few months? Especially at this scale?

What is the ultimate vision driving you or your business? I suspect for Zillow the answer is “to post better quarterly earnings and increase the stock price”. And some will argue that is what it should be. But if that is all you’re doing, will you ever be able to do anything meaningful?

In product, having a clear, long-term vision is critical for success. We may have to make adjustments along the way for what is working and what isn’t, but a poor quarter or some bad metrics won’t thrash us around. If our goal, for example, was to create a frictionless buying and selling process for our users, we may have some lessons from the volatility in the housing market, but certainly wouldn’t give up because of a poor quarter.

Product thinking is about a holistic, long-term view for our features, products, and companies.

Art and Science

Finally, we have to acknowledge that much of what we do is as much art as it is science.

Even with all the data, Zillow was unable to reasonably forecast home prices or even put together profitable offers for flipping.

As Nima Shahbazi states:

It's one thing to build a model on a website that's often reasonably accurate. It's another to then try to use that model in the real world to make very costly bets — and do so at scale

Additionally, there are many factors that go into home pricing, and many intangibles. Certain buyers may value hard-to-quantify things, and many sellers may understand that. It’s difficult to create variables for those in an algorithm.

As NPR’s Planet Money discussed recently, there is often a problem with lemons and lemonade. Bad houses (lemons) may make good sense to sell to an iBuyer, while the good ones (lemonade) make more sense to sell directly to another potential homeowner who will pay top dollar and cut out the middle man.

All this is to say that data can’t make all our decisions for us. It’s true in real estate, and it’s true in product development. Data can point us in the right direction and inform our decisions, but we need a human touch still.

So…

I personally won’t be losing any sleep over Zillow failing at house flipping. But there are several lessons we can learn from their failure. To create great businesses and great products, we have to focus, think long-term, and understand it’s as much art as science.