AgentMatch:'s Misguided Use Of Statistics For Rating Real Estate Agents

Any long time reader of this blog will know that I absolutely love statistics. However, as I’ve often pointed out in the past, statistics can be seriously misused and the notion of producing real estate agent ratings from meaningless statistics would be a fairly misguided effort. But that is exactly what the National Association of Realtors is in the process of doing with their development of the AgentMatch Web site, now in beta. Here is the concept in a nutshell, and notice that  the focus here is on selecting listing agents, not buyers’ agents:® realized there was something missing in the real-estate world; an easy way for sellers to find experienced and compatible listing agents. AgentMatch uses data, not word-of-mouth. When looking to sell your home, we understand what information matters. Rather than selecting agents by advertising or promotional incentives, we leverage statistics and data from mls’s to display agents that are active and recommended. We then give you lots of rich profile content about these agents to explore if you think they might be a fit…and no gimmicks either, just the phone number and email address if you want to contact them!

Actually, I applaud the concept. It would be great if there were objective data that could be compared between real estate agents to help you separate the good ones from the bad ones. But, alas, in reality all they/ we have to work with is the following:

  • Agents’ list/ sell ratio
  • Number of homes sold by the agent in the “area”
  • Average days on the market for the agent

In the past I’ve elaborated ad nauseum about the shortcomings in using these metrics to select a realtor but for those of you not familiar with the issues I’ll give you the short explanation:

  • Metrics like list/ sell ratio and average days on the market don’t just reflect the agent’s abilities. In fact, more than anything they reflect the type of business that that agent takes on. So if they take on a lot of clients with high expectations then their statistics are going to look “worse” than their competitors. It’s like hospital mortality statistics. The better hospitals could actually have higher mortality simply because they take on the more challenging cases.
  • These statistics could actually reflect the opposite of what you think. An agent that underprices his listings is going to have shorter market times and higher sale/ list ratios. Is that really what you want?
  • The volume of homes sold has more to do with how well established the agent is and how well they market themselves than with how well they sell homes. Real estate is a self amplifying feedback loop. The more business an agent closes the more business they get and the more business they get the more business they close. And that’s what keeps commissions high.
  • As with the other statistics production volume could actually be a negative indicator. Unless an agent employs a team high production volume pretty much ensures that an agent is stretched pretty thin and will not be able to give you much attention.
  • These statistics don’t actually reflect what is most important – what is the agent going to DO for you?

These statistics are so perverse that if they were widely adopted then a handful of agents would dominate all the business and commissions would stay exorbitantly high. In fact, if these statistics were widely adopted they might very well result in some agents cherry picking those listings that will only help their statistics. Call them up with a challenging transaction and you might find yourself begging a realtor to take you on.
I’m not saying that the realtor data should be buried. I have no problem with showing the statistics for individual listings simply because the consumer can view that data in the context of the specific attributes of that listing that might have contributed to those results. My problem is when they start aggregating the data and the consumer loses access to the context that allows proper interpretation of the results.
It turns out that this might be one of the few times that I’m in agreement with lots of other real estate agents. Apparently there has been considerable backlash from real estate agents opposed to AgentMatch. And Keller Williams issued a statement advising it’s agents to petition their local MLS to not participate in AgentMatch. In their statement they said:

While we agree that providing online access to feedback on agent performance and service level are necessary due to consumer demand, we believe that it should be in the form of reviews that are submitted by your customers.
We feel that using the data passed through from the MLS will result in rankings that are completely arbitrary and create lopsided opportunities for some.

However, on a more positive note, AgentMatch is including client reviews as part of their agent database and, as I’ve previously posted, I think that using agent ratings to select a realtor makes perfect sense. The only problem is that this part of the Agent Match Web site might get overlooked in an attempt to boil down the selection process to a handful of simple statistics.
If you want to keep up to date on the Chicago real estate market, get an insider’s view of the seamy underbelly of the real estate industry, or you just think I’m the next Kurt Vonnegut you can Subscribe to Getting Real by Email. Please be sure to verify your email address when you receive the verification notice.

Leave a Reply