Zillow Lost $1.4 Billion Flipping Houses — Then Quit

In November 2021, Zillow announced it was shutting down Zillow Offers — its home-buying business — after losing $1.4 billion flipping houses.

This wasn't a small side project. Zillow Offers represented 56% of Zillow's total revenue — $2.6 billion out of $4.2 billion in 2021. The company bet big on the idea that algorithms could predict home prices better than humans.

The algorithm failed.

What Happened

Translation: We bought too many houses for too much money, and now we can't sell them for what we paid.

In Q3 2021 alone, Zillow bought 9,680 homes — more than in the previous 18 months combined. But they only sold 3,032. The inventory of unsold houses ballooned to $3.8 billion in costs, up from $491 million in December 2020.

The average gross profit per home sold? A loss of $80,771.

The Losses

The Timeline

Why It Failed

The Fundamental Problem

Here's what Zillow (and Opendoor, and every iBuyer) faces:

To make money flipping houses, you have to buy low.

But if you're buying thousands of houses a month with an algorithm, you can't buy low. You're the whale in the market — your own buying pressure inflates prices. You're bidding against yourself.

And when the market turns? You're holding billions in depreciating inventory that you overpaid for.

Zillow's CEO admitted it:

If the company that invented the Zestimate can't predict home prices accurately enough to flip houses profitably... maybe algorithms can't solve real estate.

What About Opendoor?

Opendoor survived where Zillow failed — but they're not profitable either. They've lost $3.7 billion since inception. The difference is they saw the market cooling and adjusted faster.

The iBuyer model itself is the problem. It requires:

1. Perfect market timing — know when prices will rise or fall
2. Razor-thin margins — compete on convenience, not price
3. Massive scale — make it up in volume
4. Cheap capital — borrow billions at low rates

When any of those break down — which they always do eventually — the whole model collapses.

The Lesson

Zillow thought technology could remove the risk from real estate. It can't.

Real estate is local, emotional, and unpredictable. The house next door sold for $50K over asking because two families fell in love with it. The one across the street sat for 90 days because of a weird floor plan. No algorithm captures that.

If you're considering selling to any "instant offer" company — Opendoor, Offerpad, or whoever comes next — remember:

Their business model requires buying your house for less than it's worth. Otherwise the math doesn't work. They're not doing you a favor. They're betting they can flip your house for a profit — and taking a cut either way.

Zillow's $1.4 billion lesson: that bet doesn't always pay off.