Requiem for a Dream

Redfin and Our CEO

Requiem for a Dream

Redfin just took down Scouting Report 1.0 for good. Our latest problem was that the data we used for Scouting Report had problems at the source that weren’t easy for us to fix, mostly because agents work informally in teams, or sometimes don’t formally record when an out-of-town agent represented a buyer in a deal.

So we took Scouting Report down. The post I was going to write about this decision apologized a lot. I’m sorry we didn’t get the data right; it was arrogance to think that we had. I’m sorry we showed up in the Google index, too, and wish we hadn’t run any Redfin ads on another agent’s page, though these problems would have been easy to fix.

But I still think the folks most violently opposed to Scouting Report didn’t hate it because it was wrong but because it was right. I know that consumers loved it and now they can’t get it anywhere else. And I still believe that brokers should be the ones to tell regular folks how agents have performed in different neighborhoods, because we’re the only ones with reliable data.

As for the rest of the story, you can read all about it in the note I sent the company. Thanks to everyone who supported us or reasoned with us for your time & kindness.

Hi Redfinnians,
 
As you may have heard, Redfin just pulled the Scouting Report we released last Thursday to allow consumers to see the deal history and performance metrics of any real estate agent. The basic problem is that the data wasn’t accurate, first because of bugs then ultimately because of problems with the source data in MLSs.

Only a tiny fraction of transactions weren’t accounted for but it would have been enough to destroy our credibility if we had insisted on keeping Scouting Report live.  Had the data been accurate we’d have defended ourselves from MLS and agent complaints ’til hell froze over, but in the absence of accuracy we have no defense.
 
I don’t know when Scouting Report will see the light of day again, though I am hopeful that if we allow any agent to curate her own profile, we will resolve many of the inaccuracies. In the meantime, we’ll use some of the data to tell consumers when a Redfin agent has worked with another agent before.

Some of our competitors will dance on Scouting Report’s grave and say I told you so. I certainly feel frustrated that I wasn’t more thoughtful in launching the service; a beta program involving our partners would have avoided all sorts of problems.
 
We have to account for what happened over the past week as a failure, and try to understand how we can do better. Obviously, it’s my fault. I’ve thought a lot about how hard folks worked to pull this off. I wish you’d had a better leader in me. But the lesson we should draw is to be more thoughtful about making a risk pay, not more cautious in avoiding the risk altogether.
 
There are all sorts of projects that fail at any corporation for a different and less conspicuous reason, because the risk was measured out in teaspoons and the idea was compromised beyond recognition and nobody made decisions and the thing had absolutely no personality and nobody really cared in the end whether it was good or bad — or even knew that it existed. The reason most people give up on a great company like Redfin is because it stops making decisions and stops taking risks. They give up because the company loses its gonads and its heart and then its soul.
 
To which I say nuts. That wasn’t the problem here, and The Great Pumpkin willing it never will be. One of our values is to be fearless: bet big, tinker constantly, fail fast, measure results. If you see people who worked on Scouting Report don’t BS ’em and say the first release was a triumph — man oh man it wasn’t — but maybe thank them for their fearlessness all the same. In the end it’s that spirit that makes me sure we’ll win.
 
Love, Glenn

Show Comments

Real estate redefined