A new real estate valuation Web site developed by a former commodities trader will rely on the collective wisdom of users to estimate the value of individual homes and generate price indexes by ZIP code and neighborhood.
If enough people participate, My-Currency.com would be better at predicting the true market value of properties than sites that rely on public records, comparable sales and other factors to generate estimates using mathematical formulas, said Karim Tahawi, founder and chief executive officer.
Tahawi thinks real estate professionals will participate in My-Currency.com — not only providing valuations, but answering questions, blogging and contributing articles to a Wiki — because the site will help them land clients and generate business through the site.
Registered users will be scored on the input they provide to the site, earning them “reputation currency” that can put their profiles at the top of lists generated by visitors seeking experts in a particular ZIP code or neighborhood.
My-Currency.com will harness the wisdom of the crowd not only to achieve valuations, but to determine “who is the expert — the wisest of the crowd,” Tahawi said.
“The value to real estate professionals is in the age of the Internet, 80 percent of people are starting their search online,” Tahawi said. “At the end of the day, when they need a human being, we’re providing depth and meaning to the names and faces.”
My-Currency.com is expected to provide agents and brokers better visibility on the Internet because the site, once it has a larger user base, will show up high in search-engine results, according to Tahawi.
Visitors will be able to ask for valuations of a property by entering an address and basic information, or submitting a link to a listing. Other users will submit their valuation estimates (which the site treats as trades). The closer those estimates are to the market consensus that emerges over time — and ultimately, to the price of the house if it sells — the more reputation currency is earned.
Real estate professionals seeking to boost their status as experts will also be able to enter their estimates of the median value per square foot of homes in a ZIP code or neighborhood. Those estimates will be used to generate price indexes that consumers can use as an aid in valuing properties.
For now, My-Currency.com is limited to the San Francisco Bay Area, and the site is still in an “Alpha” demonstration mode that showcases some of its capabilities, but doesn’t include full functionality of all planned features. Tahawi said a more advanced Beta test version should debut by the end of February.
One obvious potential pitfall for the site is that users may seek ways to game the system to boost their reputations, so they will show up on top of the lists of experts the site will deliver to consumers.
What, for example, would stop users from following the pack — entering valuations that simply mirror the established collective wisdom in order to earn “reputation currency”? Or what if a property owner wanted to inflate the value of a home by having friends submit inflated estimations of its value to the site?
Tahawi, a veteran commodities trader and former vice-chairman of the San Francisco-based Pacific Exchange, says the site will address such problems in several ways.
The algorithm used to rate users will analyze their behavior, and look for suspicious actions.
“If you put in five trades in a row, one right after the other, we’d turn your market impact to zero,” Tahawi said.
Attempts by a group of users to inflate a property’s value would be counteracted by others who, seeing a property that’s overvalued, would have an incentive to place “trades” at the market value to earn reputation currency.
“If it’s a house for sale, ultimately we’ll have the answer” as to its true market value, Tahawi said. “If it was valued at $1.5 million, and it sold for $1.5 million, maybe that was the market value,” assuming, that is, that the actual sales price is not affected by the site’s estimate.
Those who placed artificially inflated estimates would lose reputation currency if their predictions turned out to be off the mark. To further minimize the potential for users to game the system, the site’s valuations would give greater weight to the opinion of longtime users with proven track records, Tahawi said.
Users will also be able to earn reputation currency by answering questions posed by other users, and writing blog posts and Wiki articles. Those contributions will also be evaluated by an algorithm, rather than human judges. That could also lead some users to try to bolster their reputations by generating a flood of blog posts of little or no value to the site’s users.
Tahawi said a user’s contributions will be scored not only by quantity but by their popularity or effectiveness as determined by the number of page views or comments.
All in all, there will be three “dimensions” to a user’s reputation, he said: performance in providing valuations, the quantity of their contributions to the site, and the popularity or effectiveness of contributions.