Entries in social network (2)

Monday
Jan112010

Using the social graph to facilitate higher trust interactions

Stik.com begins with the premise that the current commissioned sales model is broken and that your professional interactions  could be vastly improved if you stuck to people already trusted and vetted by your network. In fee-based service industries, the incentive for many salespeople is to extract excessively high margins on transactions and buyers too often do not have the savvy to know the fair price. In a market where people are often simply web searching to find a mortgage broker, financial planner, or accountant, "lemon" salespeople may continue to thrive unchecked by any review process that alerts new prospective clients to their poor behavior.

To confront this challenge, Stik.com utilizes Facebook Connect to allow you to search for the service provider you need within the networks of your friends. Co-founder Nathan Labenz explained to me recently via e-mail: "By situating the sales processes in a social context and making reputation durable, we hope to create interactions that begin with trust and proceed quickly´╗┐."

With Facebook as the platform, my main question was whether the main filtering value that the site offers is to signal trustworthiness of the seller rather than competence or if it could do both. After all, I like my friends on Facebook, but I usually wouldn't be able to speak much to their professional abilities, and I'm not "friends" with my insurance agent who I might want to recommend. Labenz had two answers for this: first, that they will be introducing a review and recommendation function to highlight competence; and second, and more interestingly, that they are deliberately targeting specific use cases where the competence is really secondary to providing a good deal. By this, I understand that he is arguing that in industries such as accounting and insurance, you can assume that all prospective providers in your network are baseline competent to help you with the paperwork, so the real question is how good of a deal you can get rather than can you get the *best* accountant out there. In that case, the signal may be the most important.

With Facebook Connect as the platform, the power of the site does seem to be a bit limited (I personally would be better able to recommend professionally via my LinkedIn or Twitter contacts), but Stik.com does see the potential in mashing different networks or maybe Facebook will evolve to be less about friends and more about contacts. In many of these industries, however, trust is at a premium and I'm not aware of any mechanisms to find really already vetted providers besides, of course, asking your friends and family. Stik.com takes that idea and makes it scale.

I'm intrigued by the use of social networks to bring better transparency to the sales market, potentially moving it to a new equilibrium where buyers get better deals, scrupulous salespeople get rewarded with more business, and lemons are shut out. Stik.com was accepted into the Founder Institute incubator program last year, launched in October, and now has a serious funding commitment from a Silicon Valley investment firm.

Saturday
Feb072009

Daylight Network launches and tries its hand at futarchy

A new social network launched this week aiming to provide a platform “for America to once again become a nation ‘of the people and by the people’.” To facilitate this goal, the Daylight Network offers three features:

  1. Tracking government spending
  2. Participating in prediction markets on government decisions and policies (a service powered by Inkling Markets) with the end goal of influencing policymakers
  3. Connecting with like-minded individuals to discuss government activities and propose alternatives

Features 1 and 3 are traditional government watchdog/social networking offerings, but let’s look at Feature 2. The Daylight Foundation incentivizing thoughtful participation in their prediction markets by providing (fake) trading currency to each trader, and paying out the winning traders with (real) cash prizes.

Many of the trades offered are similar to those found on sites like Hubdub and Intrade, such as estimating the final size of the stimulus package and predicting whether a given cabinet nominee will be confirmed. These prediction results are interesting, but are of limited value to the site’s mission as they have no influence on the actual events themselves.

One market within the site aims higher. Clearly drawing upon the futarchy theory of Robin Hanson (which receives no mention), this market breaks down the Senate stimulus package line-by-line and asks users to rate the provisions from 1 to 5 stars. As we’ve discussed before, in futarchy voters don’t vote on which policies they want (as in a referendum), but rather on which policies they believe are most likely to increase national welfare.

This is where the Senate stimulus package prediction market falls short. The crude formula identified by the Daylight Network is: “create jobs, minimize the impact on the federal debt, and promote fairness to all taxpayers.” Of these three variables, the first is clear and measurable, the second is measurable but less valuable (after all, this is a spending plan and a hit to the federal debt is a given, so the results should really be the focus), while the third is totally incomprehensible (how does one spend to promote fairness TO everyone?) Together, they become an impossible muddle of a standard to evaluate proposed policies against.

With enough users, the site hopes to gain enough momentum to influence the Senate’s deliberations. Unfortunately, by failing to clearly identify the future state that bettors should evaluate each line item against, the market is just collecting a list of voter favorites and least favorites rather than a collective impartial judgment of which items will most likely contribute towards reaching that future desired state. Still, the site is an effective forum to engage voters in talking about these issues and systematically capture their opinions.