Reputation Graph- 3 mins
Over the last week I’ve come across two distinct suggestions that one of the next disruptive and innovative markets web start-ups will target is online reputation. There are no major service providers of reputation online, only those who generate limited reputation information in pursuit of some other goal. I recently commented on a blog post by Jon Bischke about the Reputation Graph by stating that reputation is most prominently relegated to Seller / Buyer feedback on eBay and Amazon, and credit score information, among similar other contexts. There’s tremendous opportunity both online and offline for good reputation information to improve business transactions and interpersonal interactions.
I started thinking critically about reputation service providers (RSP?) during a meeting of Babson’s Idea Live! entrepreneurship group. My co-founder (currently a Babson MBA student) and I participate in the program because we’re really interested in start-up culture, to get critical feedback on our plans for fetchmob, and help other start-up founders by providing honest feedback as they finesse their business plan. One of the group’s members was considering offering a service that helps roommates coordinate and pay bills. I thought the idea was interesting and had potential if done right. The group, after discussing iterations on the idea, eventually arrived at permitting renters to offer information on their good standing to property owners. The benefit to property owners would be better information about probable tenant responsibility, while tenants with a relatively good record will have an easier time renting their next apartment. It dawned on me then that there are numerous business and inter-personal transactions that would benefit from better information about the person on the other side of the table.
We already use informal networks, service providers, and searching the web to gather information on others. However, the qualify of information retrieved may be poor for certain contexts. For example, consider the renting situation I mentioned in the prior paragraph. Another similar example would be AirBnB, where the hotelier is not interested in offering their apartment to bad guests. Now, AirBnB provides some mechanism particular to their site to tease out this information, but the pool of data it uses to characterize potential guests is probably limited*. In general, I expect P2P services like AirBnB to blossom in the future and consumers / personal service providers will demand some way to mitigate risk.
As commentators in the attached blog post mentioned, building a system that provides accurate reputation system is not a trivial endeavor. I imagine the solution will take the form of a company that allows users to connect their facebook, google, twitter, etc. graph or data, completes various analysis, then offers categorized measures of the user’s reputation. Relying on users to willingly submit specific information is problematic for two reasons: some will be dishonest, others will not participate. There is strong incentive to be dishonest and portray yourself in the best light if you an get away with it, and the thought of providing various types of disparate information to the reputation service provider doesn’t sound very fun or interesting (from a user’s perspective). Therefore making the aggregation passive, and relying on analysis of their activities and third-party sentiment from their graph to arrive at a user’s score(s) makes sense.
It may be the case that the web isn’t quite ready for this type of service. The web is still being filled out with services that capture different types of information about a user (Facebook, Twitter: social graph, Google: personal interests, shopping interest, Mint: financial history and information, Fetchmob: etc.). Secondly, single sign-on services are not pervasive across all data sources, and there isn’t a standard scheme for representing this type of data that’s exposed to the world via APIs. Further, users in general (including me) may not be ready for what privacy means (or doesn’t mean) in such a world.
That said, If any start-up is working on this and is interested in discussing integration into fetchmob, or just wants to talk about the idea, feel free to reach out to me @CrissonJCharles or via email at crisson at fetchmob dot com.
* I don’t have intimate information on AirBnB’s reputation system