my-currency Crowdsources Home Valuations

By Joel Burslem on Tuesday, January 30, 2007 Filed Under: Internet

my-currency is a site that launches later this week at the DEMO conference in Palm Desert, California. It’s hoping to out-Zillow a Zestimate by tapping into social media; blogs, wikis and user generated content to help finesse a home’s value.

The idea is that the ‘wisdom of the crowds’ will be far more accurate than any machine generated valuation. Like a fantasy stock market game, users are encourage to estimate the value of a home, to place a “trade” and vote whether they think the price is too high or too low. They then get points on how accurate their prediction is based on the final selling price of the home.

The idea is the more people involved is predicting the price means the accuracy of the CrowdValue^TM of the home will get closer and closer to the “real” value of the home. Real estate professionals are also encouraged to get involved in the process so they can demonstrate their local market expertise.

Over time, so the theory goes, the true experts will begin to appear.

I like the concept, but I see a few problems with the model right off the bat.

An accurate assessment of a home’s value really requires first hand observation of the property. Unlike stocks, whose value can be predicted on financial figures, past performance, etc. - a home’s value is much harder to determine over the Internet and really requires (to borrow a metaphor) “boots on the ground”. This is one of the big criticisms of Zillow’s AVM model - the so-called “Unzillowable” factor.

my-currency’s model requires crowd participation to estimate a home’s value, but I’m not sure how qualified the average user is to make an assessment of a home’s value. Even for the most web savvy among them, just looking at comparable sales and past performance of the neighborhood can only be one part of the equation and I think it’d be tough to find much more information on a home by simply surfing the Net and not seeing it in person.

That said, if the real estate industry really takes to this idea - then I suppose it could be a great way to vet a proposed listing price to community of interested parties.

In any case, the site, launched by Karim Tahawi, is in early Alpha stage right now and is in fairly rudimentary form right now. Data is not present in all markets but keep your eyes on it as it evolves.

Share this post:

Sphere: Related Content

If you enjoyed this post, make sure you subscribe to my RSS feed!

Conversation - Add Your Thoughts Here

8 Comment(s)

mark | Jan 30, 2007 | Reply

Rumor has it that Trulia tried to buy these guys late last year to pre-empt the DEMO hype and buzz. It was of course too early for MyCurrency to take the money and run. Plus, MyC thought that real estate was a beach head for more applications and verticals to come.

I guess Trulia will have to build their own AVM now. Not that it shouldn’t be hard to do. Clearly, anyone who can do a little math thinks they can be in the AVM game.
jf.sellsius | Jan 30, 2007 | Reply

Nice find Joel. Maybe Jim Suroweicki will stop by and let us know what he thinks about the wisdom of crowd valuation.

Either way, CV should draw a crowd of the curious. If they gave the experts dollars instead of points, that would be interesting

As you point out, nothing beats first hand inspection of a home, as well knowledge of the current market conditions and pool of competing homes (and all the other unzillowables), when determining what a home will sell for, which is really it’s value for practical purposes.
Incredible Agent | Jan 30, 2007 | Reply

We all know a home’s real value is only as much as someone is willing to pay for it…not what some calculator says it’s worth or even what some agent says it’s worth. (Supply vs Demand=Price not Sales History+Agent Knowledge=Price) There are some enourmous economic flaws with the entire AVM concept, not just the “unzillowables”.

I wonder what would happen if we tracked the actual price of what a home sold for vs what all the AVM’s said it was worth at the time the home was listed? Then we’ll be able to see who’s numbers are more accurate.

Will the crowds win over the calculator?
John | Jan 30, 2007 | Reply

“I wonder what would happen if we tracked the actual price of what a home sold for vs what all the AVM’s said it was worth at the time the home was listed?”

Incredible Agent, that is what Zillow is doing (only for their own algorithm would be my guess). They’re plugging into the local MLS’ to get fresh sold data to compare their zestimates to recently sold homes in order to guage their margin of error and improve their algorithm.
Joanne | Jan 30, 2007 | Reply

That’s all we need every buyer who reads in the paper about the bubble and listens to what the TV says will make an opinion on what a house is worth. It is bad enough that the buyers make offers $50 to $100,000 below asking!
Jake | Jan 30, 2007 | Reply

All of these new AVN’s count on user input to generate accurate data, but none of them I see have any kind of plan to generate enough traffic to their websites to get any kind of accurate data.

Jake
Kevin | Feb 1, 2007 | Reply

Brilliant idea. Perfect venue for real estate professionals to showcase their expertise.
Andy Denton | Feb 10, 2007 | Reply

Joel,

I agree with you that the home valuations must be completed by “boots on the ground” activity. I’m not so sure opeing this application to the masses is the sweet-spot. Modularizing it, making it an agent tool which can be used at open-houses seems more applicable.

But, as we all know, the crowd suggested home valuations are provided when a prospective buyer presents an offer. Allowing non-participatory critics to value homes will yield unrealistic modeling.

http://www.andydenton.com/2007/02/10/my-currency-puts-home-valuations-up-for-debate/

Best,
Andy