Online rating systems are ubiquitous and powerful. But not everything is easy to rate. Restaurants and hotels are pretty easy, in part because the people rating them are consumers who are simply supplying their impressions and opinions. In these scenarios, everyone who rates and comments is equally (un)informed. People know what they know and think what they think and rate accordingly. The value tends not to be in the individual comments (trust me, I’ve read more than my fair share!) but rather in the aggregate view. If a majority of people hate something, you stand very little chance of figuring out exactly what is causing the hate, but at least you know something is wrong. There is still good utility in that.

This could explain why it’s been so much harder for ratings site for professionals to take hold. There urgent need for consumers to get better insight into lawyers, doctors and other professionals is huge. But there are two major complicating factors it: knowledge asymmetry and personal relationships.

First, how does a consumer rate the knowledge and quality of a lawyer? They really can’t, which is why a lot of legal ratings devolve to “great lawyer” (read: she won my case) or “lousy lawyer (read: she lost my case). You also never see nuance such as “she did a great job despite my lousy case.” Services like Avvo have built a name rating lawyers based on public domain data. It’s a start , but it doesn’t provide the color and insight of crowdsourced consumer reviews.

Doctors are a special case. Beyond the knowledge asymmetry, many people feel a personal connection to their physician. That creates a “my doctor can do no wrong” mentality that doesn’t leave much room for ratings services. Moreover, physicians have proven to have very thin skins when it comes to being rated: doctors have been much more active in trying to get patients to sign legal agreements prohibiting their patients from rating them, much more so than lawyers.

One interesting workaround is to ask doctors to rate other doctors and lawyers to rate other lawyers. It’s not a new concept: in fact, Martindale-Hubbell has been rating lawyers this way for 100 years. More recently, Castle-Connelly has built a name with a similar service for doctors. What’s fascinating, though, is the Martindale system, originally built for lawyers to use in picking other lawyers, was something of a dud when offered to consumers. It seems that because consumers didn’t trust lawyers generally, they certainly weren’t going to trust a system where lawyers rated other lawyers. The unexpected outcome of the Castle-Connelly rating system is that doctors tended to give the highest ratings to academic and research physicians, top doctors to be sure, but ones that typically didn’t see patients.

We’re making strides, but there’s still no clear solution to this fascinating problem and massive business opportunity, one that’s been resistant to a technological solution precisely because of the human element.