Blog Post

The Race to Create a Web of Reputation

Stay on Top of Emerging Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

One of the big issues with the ongoing explosion of social media, whether it’s blogging or Twitter or Facebook, is a lack of effective ways to filter the signal from the noise — in other words, to figure out who we should pay attention to. Facebook relies on your existing social graph, while Twitter uses its own internal algorithms to suggest people you should follow, and LinkedIn uses your professional status and co-workers or contacts as the benchmark. But the race continues to try and measure online reputation in an effective way. Should it be based on activity? Number of followers? A ranking system in which people can vote on you? All of the above?

One of the latest to jump into this race is Mixtent, which launched today with a voting-based system that uses data from your LinkedIn profile once you log in with your credentials (and will also pull in your Facebook info if you connect that as well). The company says it is “building a professional reputation graph on top of the main social and professional networks” in order to help people hire others and get hired themselves. If Mixtent looks a little familiar, that’s because it appears to be almost identical to a LinkedIn-based game known as Cube Duel that got some attention a couple of weeks ago, in which users vote for co-workers and can “unlock” various badges, and so on.

In trying to measure who has the highest reputation among your co-workers, and therefore who is best qualified to either recommend you or be recommended themselves, Mixtent is going after the same kind of market that other startups such as Honestly (formerly known as Unvarnished), Namesake and BranchOut are aiming at — namely, the professional end of the social web, in which people are looking to network for jobs. In the same way that Mixtent is based on the LinkedIn network, BranchOut uses Facebook as a platform, and leverages all of the people you are connected to via your social graph who might work (or used to work) at other companies.

One issue for BranchOut that I wrote about when the service first launched is that Facebook is primarily personal, and so the overlap between that part of your life and the professional side is haphazard at best, and useless at worst. In a similar way, the game-like aspect of Mixtent might not jibe well with the more professional aspects of LinkedIn for some users. Honestly, meanwhile, is trying to create a reputation-based network that achieves the same thing as LinkedIn or BranchOut — a way of measuring a person’s skills within a certain professional context — but allows for anonymous (and therefore theoretically more honest) input about the people who are being ranked.

Namesake wants to create a personalized network for professional recommendations that is like a more personal or social version of LinkedIn. You can follow people, and recommend them based on what you see as their areas of expertise, and then you can forward or “route” opportunities to them that come from your contacts. The kind of crowdsourced reputation that Namesake is built on also emerges from social networks like Quora and StackExchange, where people answering questions in their area of expertise builds up their reputation (something VC Fred Wilson discussed with me last year). And a company like Klout comes at it from the algorithmic end, by looking at your activity on Twitter and Facebook to try and give you an overall social-media “score.”

One big problem for these services, however, is that each of us has different reputation ranks within our social, and even our professional networks: I might trust my friend Chris when it comes to advice on barbecuing, and I know he is an excellent videographer, but I would never listen to him when he recommends music. And while I know that my friend Rob is a lawyer and understands technology, I have no idea whether to recommend him based on his knowledge of carpentry, or of wills and estates.

This is why Namesake is trying to create a professional network that functions more like a social web — because the ways in which we interact with each other often don’t fall cleanly into one category or another. Will a simple voting system like Mixtent is offering work as a way of measuring reputation? I’m not convinced.

Related GigaOM Pro content (sub req’d):

Post and thumbnail courtesy of Flickr user Danny Cain

12 Responses to “The Race to Create a Web of Reputation”

  1. It seems fairly obvious that reputation is multiple things and cannot be define by one metric.
    From top of my head here’s a few things I would like to know about someone and that can be somewhat measured online :

    * Sociability : how much does he share, comment and interact with other users. Can tell a bit about personality type

    * Expertise(s) : How much does he know about a particular topic or issue ? This is topic specific.

    * Reach/Audience : When he talks, tweets, comments or blogs, how large is his audience ?

    * Likeability : Do people like him ? Is he someone I want to do business/ hire / work / partner with ?

    * Trustability : In general (no related to a topic) is he seen as someone you can trust ? Someone reputable ? Spammer is not.

    There could be many more, but that’s a beginning.

  2. The filter versus noise is a challenge – according to some stats I’ve seen white collar workers spend 25% of their day managing this issue. Throw in useless meetings and the numbers rocket up.

    I don’t see this ever being resolved by figuring out WHO we should pay attention to, that’s a dead-end. The answer lies in WHAT we should pay attention to. Cards on the table, that’s what we’re betting on here at Cohuman. To get nerdy: Synthesizing Salience through Emergent Prioritization. Leverage the actions of the team to determine individual priorities.

  3. Certainly an interesting comment– but will people be more honest here than they are on LinkedIn? I can’t count how many times I read a recommendation and know that its completely fabricated.

  4. At some point, we’ll be spending more than 50% of our time answering questions and participating in forums online in order to ‘prove ourselves’ professionally.

  5. Joseph Thornley

    The one best way to filter noise out of content is to read and make up our own minds. To entrust any algorithm to determine for us what is quality content is the equivalent of leaving it up to Reader’s Digest to determine what we read. And that’s really shortchanging ourselves.


    • Nicole Solis

      It’s true. We were a little too quick to hit the publish button on this one, and so some of our RSS readers got an inadvertent inside peek into how the GigaOM sausage gets made.