Yesterday, Derek Stacey, a colleague of mine at Razorfish sent me this article:  http://gigaom.com/2008/12/08/yahoo-plans-to-launch-a-mail-app-platform/. We worked together on a nifty project aiming to aggregate and blend both social and email experiences from numerous sources into a seamless and light web interface.  The idea did not come as a surprise, but rather as a confirmation of a fact we figured would become mainstream soon: social graphs are everywhere.

Social graphs are born from repeated conversationAnywhere.

Concepts around what a social graph really is or could be started becoming clear to us as we tried to conceptualize an aggregated address book, with sources such as various social networks and email address books. We ran into a number of challenges, such as the inability to equate profiles across networks with varying information restrictions. This lead to many very interesting spins around how to use social network graphs and aggregate them into a form of  “super graph”, which lead us again to believe the following: social graphs are a by-product of sustained conversation using a communications medium.

The idea is pretty well illustrated when we think of our Facebook friends list, our Twitter follows and followers, our email address book, our IM friends list and more. Each of these belongs to a specific communication service (FB, Twitter, Yahoo!, Gmail, GTalk, AOL, etc.) and we build those graphs from sustained and repeated activity.

Email: One heck of a social graph…

Kudos to Yahoo! for a move to leverage the oldest social graph in digital history. Who can think of a communication vehicle stronger and more ubiquitous than email. We can get a bunch of us digitophiles in a room and list out who doesn’t have an Orkut, Hi5 or MySpace profile but I can bet you any amount of money the list of those who don’t have an email address is going to be empty.  Everybody has one if not multiple email accounts. We all have an address book, full of contacts, IMs, phone numbers, distribution lists, groups, even calendar entries. The email medium is in fact where you’ll find the richest and most stable social graph.

Beyond FB Connect. More freedom to choose.

While FB Connect and all the other APIs out there, old and new, are a great start for portable social graphs, I see the realization of email as a social graph and vehicle very disruptive. Here is why:

  • Email as a communication technology  is open and supports old, proven and stable APIs
  • Email has unparalleled support for a variety of content types and delivery options
  • Email is probably the most widely adopted and used technology in the world
  • The barrier of entry for integration is very low: think Yahoo! Social APIs, Google Data APIs (mail, calendar, etc), MSN Live APIs and more.
  • Email as a social graphs gives you identifiable leads, not just virtual clients protected by API restrictions

So come to think about it, referring to the great article Shiv, Jessy and others have written about Amazon + iTunes implementing FBConnect (http://www.goingsocialnow.com/2008/12/imagine-if-amazon-integrated-f.html), here’s how one can think of an Amazon decision strategy:

  • Analyze their user profiles by email address. Find out what portion of them belong to an “accessible” provider (such as Yahoo!, Gmail, MSN and many others)
  • Run some analytics magic to figure out what proportion of Amazon unique users have also been on Facebook. This can apply to other networks as well.
  • Make an educated decision… I’m willing to bet going the email route will prove more powerful.

Possibilities are endless…

Now the exciting part. Yahoo! and many others have and will realize that tapping into “human area networks” is easier than ever. Facebook and others have forged the way to raise awareness to the benefits and wonders of graph based communication while creating circles of trust between users, their friends and third party applications.

What’s exciting about using email is the ability to openly integrate messaging and content delivery in a centralized or de-centralized fashion, the sky’s the limit in terms of ways to use email to re-create any kind of social experience.  A tiny but perfect example is how you can send a picture via email directly into a Flickr album corresponding to that email address.  Does this get you thinking of the limitless capabilities of the integration medium?

What’s even more exciting about using email is the idea of re-inventing permission based email marketing in a much more targeted, social, trusted and efficient way.


We’ve been noticing a lot of press, internal and external buzz around FacebookConnect. Surely because of the latest marketing push Facebook has been attempting.  I will cite one ofthe many articles recently published, painting the picture of Facebook’s attempts to find “distributed” ways to monetize. It is interesting that Facebook has waited so long to publicly engage third parties in leveraging the Facebook social graphs in order to increase site stickyness. FacebookConnect incorrectly seems like the groundbreaking new and unique opportunity everybody’s been waiting for. I’d like to say that we, our clients and Facebook, justifiably have just been passive about weaving Facebook (or other social networks) into websites. While it’s great that such announcements have captured the interest of the masses, it’s important to know all the options that are and have been on the table. 

Facebook (and other) third party integration has been around for long…

I can find technical posts that date back as much as May 2006 that answer questions around the integration of the FB REST  API. This API, at the time, already supported authentication, friends lists, profile information, photos and more.

Beyond Facebook, let’s not forget the fact that sites like LinkedIn and many others offered the ability to find friends using your address book. While some services resorted to scraping, address book APIs have been around for a long time from providers such as MSN, Yahoo! and Google. What people ignored is that these same services also had distributed authentication schemes such as OAuth, OpenID, Live (Passport). The ability to make sites more social and to avoid having users register 1000 times (haven’t we all) on 1000 sites has been around for a long long time. 

As for MySpace, a similar picture can be painted. While Facebook was in fact the first to release this type of integration capabiltiies, MySpace closely followed. MySpace has, for over a year now, implemented the OpenSocial container and supports REST services around much of it’s social functionality and data, with permissioning, authentication, etc., etc.

The real question: Why should a site give up on their profiles?

The real question to be asked an answered in order to shed some light on the shift that seems to be occuring is the following. “Why should marketers and service providers give up on their user base and switch to one owned and operated by Facebook?”

This is in fact a tough question to answer. And marketers, with digital programs that heavily focused on lead generation, email, user profile,  behavior tracking and personalization, had very little time to entertain the option of letting go of their data in exchange for a very controlled third party set of profiles.

But the game is changing. Marketers are starting to believe in the accuracy, relevancy and scale of interactions within social graphs. Suddenly, every marketing touchpoint can become viral. The marketer’s voice can be carried on by users within their own realms. After all, who knows people best than the acquaintances they’ve “allowed” to interact with them. The concept is almost unbelievable. The reach is brutally powerful. The opportunities abound. This is just the beginning…

Needless to say, the question is no longer whether marketers “should” replace their user base, but “how” they should replace their user base.  And in that field, some have been agressive (CBS) and others remain cautious, rightfully so.

Why Facebook should pay close attention: there are other options out there.

While Facebook has always been the first to make the right moves (much respect!), this is a critical time for them to understand the value of openness. Facebook figured out they couldn’t monetize solely by driving traffic and advertising on the facebook.com property.  A distributed interactions platform is now the way they want to monetize. And they are right on to think so. But what will make or break them is the perception integrators and marketers have of the platform’s openness. Here’s why…

Anyone who’s played around with FB Connect or API based integration can testify to the fact that these platforms were carefully crafted with control in mind. Fear of losing their audience, of providing the tools for people to build alternative, even better experiences outside of Facebook. For example, authentication is a semi-complex process which takes a user to Facebook, to return to the site they were on. A site implementing FB Connect cannot obtain an email address from a profile and is limited in the numbers of notifications they can send users. This is the whole Beacon story, which has really never been appropriately solved.

While marketers are excited to embrace FB Connect, the next “reality checkpoint” will be about the plethora of options available out there as alternative social graph platforms. Google, MSN and Yahoo! all have very rich social APIs.  I am sure that represents nearly 95% of the entire target population. As a marketer, I would not only rely on FB to connect my users, but on the ability to let users connect via the mechanism they feel would be most efficient (i.e. via Google, MSN, Yahoo, etc.). For this very reason, Facebook will have to implement a status of openness comparable to these platforms in order to keep competing.

The money factor: How the heck will Facebook setup revenue share? 

Distributed, interaction based marketing. How does Facebook monetize? Do marketers engage in “affiliate-marketing-like” contracts with Facebook? For example, if Amazon were to implement Facebook connect, to they pay commision for purchases that were measurably influenced by social graph conversations? Is so, how? Now that’s a tough nut to crack… It’s too soon to tell.

 Final point: If users are now marketers, why don’t they get paid?

Leaving you with something to think about… If users are empowered to market on behalf of businesses, why don’t they get paid? What do I mean? Simple… Let’s create a digital marketing platform that creates or aggregates social graphs. The platform can enable interactions on any sites, much like FB Connect. However this time, the user’s influence of purchases and conversions is analyzed and remunerated. Wouldn’t that be swell? Finally, I get paid for convincing those 10-15 friends to buy an iPhone.