Laserlike

People respond to incentives.  

These four words are at the heart of the social sciences and the key to hacking human nature.  Mankind has developed positive incentives and negative incentives.  Positive incentives include fame, praise, respect, degrees, badges, medals, [good] grades, points, and various monetary rewards (which can be traded for most other goods).  Negative incentives include prison, the dunce hat, [bad] grades, gossip, ridicule, physical intimidation / harm, and monetary loss (lawsuits, depreciation of assets in disrepair, etc.).  It’s important to remember that “incentive” does not necessarily mean “money.”

In the early days of the internet a small community of academics relied on community norms to govern the system.  The systems governing the exchange of email were limited.  What incentive would one academic have to spam another?  The same goes for usenet, which was a very open message board with limited governance.  The first generation of web search engines made assumptions that publishers of content on the web would accurately tag their own sites so that search crawlers could help index the web.  Who knows better than the content creator?  In small communities like the early internet, the cost of being a bad actor were high — transparency and the fact that it was a small community were powerful incentives.  

But then email became pervasive outside of academia and SMTP was (and still is) ill equipped to deal with a world with bad actors — even a small percentage of spammers can ruin the experience for everyone else.  Usenet was quickly overrun by spam as the internet went commercial.  And the idea of having publishers tag their own content for better search and indexing failed once incentives to cheat grew and the cost of doing so decreased.

While most people think about incentives on the web being monetary things like About.com paying their guides or virtual currency in name-your-favorite Facebook game, understanding incentives has been at the core of many of the web’s greatest inventions.  And I suspect that will continue to be the case.  

1.  Google’s PageRank.

Google Search “won” by hacking human nature to fight the spam problem.  The insight behind PageRank is that you can figure out what a site is about and its relative value by what other people say about the page rather than the publisher herself.  Publishers speak with their links (and the content around those links).  Not only do you need high quality sites linking to you, but you also get punished for linking to bad content.  Brilliant.

Crappy content fades away and bad actors live under the threat of being removed from Google’s index.  A potential benefit of having a monopolist in the search market is that the threat of being removed from Google’s index could literally lead to bankruptcy for many potential spammers (not the case in a fragmented world of search providers).  That’s a painful pill to swallow and one that arguably leads to a better search experience for all consumers.  

2.  Facebook’s domain-based authentication.

Early social networks had many of the same issues [and still do] as usenet.  Consequently, users [particularly women] were often afraid to share authentic details about themselves.  

Until 2006, the only way to get into Facebook was with a university domain (e.g., name@harvard.edu).  Domain-based user authentication not only limited fraud, but allowed Facebook to automate the creation of groups by University (and later corporation, etc.).  Furthermore, the cost of being a bad actor surely carried greater risk than having some lightly authenticated name@hotmail.com account.  

What other mechanisms of authentication can improve authenticity?  How about sending physical letters with a unique code to every home on the planet to create a location-authenticated social network?

3.  Less is more with IM buddy lists.

Of all of my online collections of friends, the smallest by far is my IM buddy list.  There is a huge cost of interruption.  With a small list of people, I can leave Adium running in the background whenever my computer is on.  While I’m fairly open to receiving and making invites on other social networks, I cultivate my IM buddy list like no other.  LinkedIn smartly suggests that you only invite and receive invites from people you know.  But instant messenger applications needn’t make such suggestions, as users quickly learn the costs of too many or unknown friends.  I get more email spam in a day than IM spam in a year.