The Likelihood That an Agent Will Sell a Listing? Less Than 50%

A couple of weeks ago, Redfin engineers got together for a hackathon to prototype features we’d like to see on the site. One team, featuring Jane Nemenman, Jamie DeMichele, Dane Brandon and Llewellyn Botelho, built a Redfin.com widget for each listing that showed the listing agent’s track record: how many listings he had on the market, what his average discount to list price was, how long it had been since he closed a deal.

It was a great idea. But it didn’t all come from the engineers. The original insight started with our San Francisco agents, who like to size up a seller’s agent before deciding how to represent a buyer in a negotiation, on the theory that negotiating strategy is often influenced as much by the listing agent’s state of mind as by her client’s. Some agents are chronic over-pricers, expecting to give part of that away at the negotiation table. Others stand firm. And still others just need to get a deal done.

Jane, Jamie, Dane and Llewellyn wanted to give everyone this information, so that anyone using Redfin’s site could know what she was up against going into a negotiation. Then we dug into the rules that govern how we use listing data, and decided that using the broker’s database of listings to embarrass brokers publicly wasn’t a fair use of the data.

We’ll still build this into the tools our agents use, so we can help all of our customers know when to hold ‘em and know when to fold ‘em. We’ll also share with everyone the listing stats for our own agents. In the meantime, what I haven’t been able to stop thinking about was how the engineering team reacted as Jane demonstrated the widget, showing the dismal stats for one seller’s agent after another.

Folks were flabbergasted. At first, people thought it was just one agent having a tough year. But after a few minutes of  clicking from one listing agent to the next, everyone began to recognize the truth: that in 2009 it was very hard for any agent to sell a home.

So when we got back to our day jobs, Jamie DeMichele — the man who also created bracket-tracking software for March Madness — looked up the numbers for all the listings put on the market in 2009, to see how many had sold by August 11, 2010. The answer? About half. He emailed me the table below, which summarizes the success rate for broker-listed properties for sale in seven major markets:

We shared the data over the weekend with the Wall Street Journal, which just published its own analysis. As we’ve argued in the past, the basic problem is a stand-off between buyers who expect the world, and sellers who have already taken more losses than they can bear. When no one will compromise, and the banks have been slow to foreclose on overdue mortgages, listings don’t sell.

What does this mean for you if you’re trying to sell a house? Primarily: don’t hire the agent promising the highest price, no matter how flattering that may sound. Hire the agent with the best track record. If 2010 is anything like 2009, odds are that the property won’t sell at all, or at least  not  at the originally promised price.

(Picture of Jamie used with his permission, at his insistence that I use one where he’s wearing cowboy boots)