Pandora's box and dangerous thoughts.

Pandora's box and dangerous thoughts.

Honest, officer. I didn't mean to pry open this box I just happen to be standing next to with this eight-foot crowbar I just happen to have in my hand.

It's just that I was delirious with the knowledge that real estate consumers are more and more empowered these days. I keep hearing this from people who would know: academics, regulators, consumer groups, Internet entrepreneurs. The usual 21st century authority figures and objects of idolatry.

Just think! Consumers can now go to a Web site called Homethinking that "helps you find a real estate agent by showing you sales histories and reviews from clients".

Just think! Internet-empowered buyers and sellers can now go to Homethinking and discover that:

1. I've had only one transaction.

2. It was originally priced at $629k.

3. It sold for $470k.

Then, blown away by this gaudy track record, Internet-empowered real estate consumers can contact me by phone or email at their convenience.

I won't hold my breath.

And while I'm not, I'd like to make a few corrections to the rap sheet Homethinking built on me:

1. I don't know how many transactions I've had over the past nine years, but it's enough to put me in the top 8 percent of all Coldwell Banker agents internationally four times and in the top 18 percent once. In addition, I was one of the top 1000 NRT sales associates in Q4 2008, putting me in the top 2 percent nationwide. I think everyone will agree that that's a lot of mileage from one sale.

2. The one transaction Homethinking credits me for was a listing that came on the market in May 2006 at $649k, not $629k, because the listing agent (me) was fooled by a brief uptick in neighborhood sales activity that spring. That brief uptick was, in fact, the last sign of a real estate pulse that neighborhood has had for almost two years. After a few months, my client reduced the list price to $629k. So the original list price, which is the far more important number, was $649k, which I feel sure you'll agree is not $629k.

3. The home eventually sold, not at $470k, but at $629k. That's according to the sales contract, to the Multiple Listing Service, and to a data aggregator called RealQuest that gets its sales price information from county records. To repeat, that's $629k, which I should point out is not $470k. Actually, that's $629k less a $20k credit to the buyer, which is a net sales price of $609k (and lucky to get it; the home is now worth half that) which is the far more important number, which is not $629k but also not $470k.

Aside from that, Homethinking's got me nailed dead to rights. Cool! This changes everything, or whatever it is they say at the product rollouts these days.

I will admit that Homethinking might have a few drawbacks as currently configured. Things it might want to work on or, failing that, disclose.

First, Homethinking's investigation of my track record apparently goes back only to 2006. It doesn't say this, but it also doesn't mention any of my transactions from 1998 to 2005.

Second, Homethinking doesn't show any of the transactions I've had representing buyers since I sold that $629k listing. Apparently when Homethinking says "John Fyten sold", it really means "John Fyten listed", which is a definition of "sold" that's incomplete and needs disclosure. It's also a definition that puts agents who work only with buyers at a real disadvantage. It also doesn't do any favors for newer agents, who typically start by working mostly with buyers.

Third, Homethinking doesn't mention my $1.45M listing that sold in 2007 the day before it hit the MLS. Apparently when Homethinking says "John Fyten sold", it really means "listings John Fyten had on the MLS". I haven't sold many homes off the MLS, but in some markets, like Silicon Valley's ultra-top end, millions of dollars in sales are private.

Homethinking says its records come from "monitoring of the public web". That's a little vague, at least to a non-geek like me, but it may mean that Homethinking gets its data from Web sites anyone can access, such as Realtor.com. Not from the local MLSs, which charge membership fees. Not from the aggregators of public records such as RealQuest which charge subscription fees. Public listing Web sites aren't comprehensive—for example, they don't have the names of the agents who represented buyers of those listings—but they're free. And if Homethinking pays for its sales price information, well, in my case it deserves a full refund.

I should also mention that Homethinking calculates an average transaction price for each agent so you can know his or her price range. My price range is $470k, of course, since that's the only transaction I've ever had. But the funny thing is that an agent who got into the business even just a few years before prices skyrocketed might well have an average this low. Today's million-dollar house was yesterday's $750k house and, not long before that, a $500k house. Condos? Back in the day you could pick them up for a few hundred thousand or less, even here in Silicon Valley. So an agent who's been in real estate just since 2006 will have a much higher average transaction price than a veteran agent, even if the newer agent works a lower price range as newer agents often do.

Homethinking might respond (although it doesn't) that it levels the average-transaction-price playing field by going back just to 2006 (although it doesn't say it does), but if that's what it's doing (and it doesn't say it is), it's still distorting agent transaction histories by ignoring the number of sales that veteran agents had prior to 2006.

But it's all good: Homethinking says that all I as an agent have to do to update (or in my case, correct) my transaction record is to register. Fair enough, except that by letting agents update their records, Homethinking seems to be opening the door to inadvertently allowing agents to compromise its data.

Here's a real breakthrough: Homethinking lets former clients post comments on their agents' performance. Fine, except that how do Homethinking's users know that these comments come from former clients and not from the wives, mothers and close friends of agents—or from the agents themselves—or from cranks? How are online reviews any better than reading an agent's testimonials and wondering if the agent wrote them himself? Might they be even worse?

Which shows how hard it is to harness the indiscriminate data-grabbing ability of the Internet to something as nuanced as real estate agent performance. It's a great idea, or it can look that way to unsophisticated investors reading a business plan or to unsophisticated consumers reading a Web site's home page. After all, data don't lie...unless they're incorrect, or incomplete, or they've been bent to fit an agenda, and we know none of that ever happens. But nothing in the real estate transaction—not the property, not the agent, not the buyer nor the seller—is a cookie-cutter commodity that lends itself to robotic data collection. The actors in real estate transactions aren't Sport Utility Vehicles or digital cameras or laser printers or any of the other identical mass-produced units that can be turned into mass-produced units of online consumer information. It's real estate's nuances that have tripped up so many Next Big Things launched by so many bright entrepreneurs with so much technical knowledge and Web marketing experience and zero experience in fusty old real estate.

You can almost hear these savvy entrepreneurs asking, "How can real estate be nuanced? It's just a bunch of knuckle-dragging used car salesmen". Answer: how would you know?

So I wasn't surprised that Homethinking's two co-founders identify themselves as a former online advertising and media analyst and as a specialist in "crawling architecture and information retrieval". If they've ever spent a minute in real estate, they're too modest to admit it.

Yes, friends, it's another "triumph" of technology over experience, of data (correct or not, interpreted well or not) over knowledge. Just like virtually every Internet venture into real estate, from the flossiest Web site to the funkiest bubble blog.

Which might cast doubt on just how empowering the Internet is for real estate consumers.

Which might put a great big ding in the idea that the Internet has reduced or eliminated the need for real estate agents.

Which might blast a hole the size of Rhode Island in the theories of academics, regulators and other conspiracy buffs.

Which might explain why the full-service full-commission brokerage model doesn't wither away.

Which might expose the whole—

"Put that lid back on!!! put your hands up and BACK away slowly from that box!!!"

What's the charge, officer?

"Laughing out loud in the temple of technology worship and refusing to take the Internet as seriously as it takes itself. We got laws in this town."