Why Do People Use Social Networks ?

Believe it or not, it was really tough trying to figure out exactly why people use social networks and services - looking at both the industry in general and at networks in particular. The first conclusion we wanted to jump at was to say "People use Social Networks because of what they offer" - and to an extent that is definitely true, particularly among the niche services. But after we profiled the 100th network or so, we began to see that they all pretty much offer the same stuff - and that just complicated our initial conclusion beyond utility.

So we started to look at the social networking interests, and identified three core players in the arena:

• The Userbase
• The Social Networks
• The Content Providers

Then we started looking at their growth over the years, and noticed some important patterns regarding their numbers that correlated to market trends.

The Userbase was growing at a highly exponential rate largely due to:

• increased net access for end-users
• increased SNS choices - networks, applications, store - everything added in id
• niche SNS targeting was bringing more people online.

The combination of more people, more options, and a diversity of options turned the userbase growth rate explosive.

The Social Networks were also growing exponentially (though at a slower rate) due to diminishing barriers to entry. In 2003 starting a Social Network meant hiring a lot of smart coders and building something from scratch. By 2007 the collective knowledge of the industry eliminated most needs for R&D: the bulk of Open Source languages had grown sophisticated templating languages and Module repositories that competed with Perl's accomplishments, and you began to see Open Source frameworks and commercial or free whitelabel solutions. What once took know-how, talent, and millions of dollars five years ago can be accomplished with a modicum of financing and knowledge.

The Publishers were growing as well - but differently, at a linear rate, which is where things got interesting. With millions of people accessing content on thousands of sites, there is really only a few hundred content originators that truly capture attention. While the online arenas and consumers exploded, there simply wasn't an explosion in the offline growth of artists, authors, musicians, actors, scientists publishing papers, etc. Technology made it easier to communicate online - but it couldn't provide the fan-based dialogue or the cultural or academic discourse that our societies have.

The ultimate conclusion we came to was that to be a successful network startup you would need a certain amount of people to actually use your network - the critical mass. You also need some amount of exclusive/premier content - as most is syndicated and can be found elsewhere. Above and beyond anything else though, the primary attraction of people to Social Networking Systems are the basic utilities and features used to access the information. This experience quality is the sole thing that differentiates one community from another - and is not necessarily an advantage in technology or usability , it often is the culture of the community or something else ephemeral and intangible that affects the user experience.

What Is A "Social Network" ?

Once you start looking at online networks critically you ask "Why are people using these things?" Eventually get to the point where you start to ask "Well exactly what are these 'Social Networks' anyways?" - and that is a question that is often answered incorrectly.

A lot of people like to think of Social Networks as publishers, which they're not. While SNSs often will create some amount of original network-sponsored or endorsed content, it is miniscule when compared to the syndicated content or user-generated content that they mediate. Because Social Networks don't filter or edit the user-generated content , and users will often post that same content in multiple places , the term 'publisher' becomes misleading.

People often think of Social Networks as social networks too (notice the difference in capitalization here). Social Networks are computer programs, applications, digitized relations spanning a domain . They're not a network of people - just a representation of some relations with activity that is mediated by a third party (the network). This sounds like I'm splitting hairs semantically - and to an extent I am , so i can better illustrate that a Social Network isn't the group of people who use it -- it is the application and what it offers.

My definition of a Social Network often sounds incendiary to those in the industry - but it is the essence of what they are... Social Networks are interfaces. They're the utility that bridges the interaction of users with one another, and with content / content originators.

If the industry can get past its hubris and try to except that definition, they'll see that there's nothing wrong with it. Actually, its really good.

If you are operating a Social Network, aside from income, you have two concerns:

• How do I source content ?
• How do I acquire and retain users

Talk to anyone who works at a network and ask them about their core business concerns. If they said anything else, I'd be amazed.

Social Networks are simple interfaces. If we begin to approach them in those terms, it becomes fairly simple to identify the tools needed to handle acquisition and retention concerns - which can be outsourced them through Social Network Portability and Content Syndication. We can let the networks focus on their own features -- the way users interact on their site, and the non-generic content and activities. If networks were to "Open Up" to portability concerns, and stop thinking like Cellphone Carriers, they could refocus their efforts on what makes them special and unique.

The Interface Layer

There is a short, yet richly detailed history of Social Network migrations - how groups of users move from one network to another - which shows some very interesting trends. While there is an obvious element of Critical Mass involved in migrations ( wherein people simply migrate because everyone else is ; the "orange is the new red!" school of thought ) , if we peel away the layers of history to see why the critical mass actually moved... it is almost universally due to an interface layer change.

When I talk about "The Interface Layer", I don't mean a pretty User Interface (though that can be a part of it), I mean to talk about the way in which people can interact in general. People migrate to the easiest, most efficient and entertaining/rewarding way of communicating with one another. MySpace pulled away hordes of Friendster users because you could befriend bands, and quickly message anyone in your network of two-hundred-billion-friends. Facebook offered a cleaner, more intuitive alternative to MySpace. People still use Evite and new calendaring services - despite having all these functions offered by the premier Social Networks - because they're so easy and open to use. People aren't using these networks and applications just because they're friends do -- just look at your friends list and your email addressbook: you can contact just about everyone you want to, directly, through email. People use networks because of the advantages these interfaces offer.

What Is A "social network" ?

Once we've defined "Social Network" as an application, we're left with pondering the term "social network" in general.

Whenever I ask this question, I'm almost always answered with the phrase "The real social network is the internet!"

While that answer is correct, its a bit limiting - the concept of a social network is much bigger than that. The social network is the real world: not just online , but off. The social network is our connections - our real human connections. It connects people to people, and people to entities.

Social Networks are closed and proprietary. They track relations within a system -- within applications and corporate partnerships and domain names. They operate within a box.

The social network is not proprietary. We're more than just an online ID or url.

"I am not a number, I am a free man."

You can't realistically quantify an actual human relation under an online network - and you can't expect or require everyone to join the same applications.

You need to think outside the box - literally: The social network spans multiple Social Networks.

Initial Market Resistance

FindMeOn was up and running by August '06; September through November of that year was our first round of investor pitching and alliance building - and it was a nightmare.

Our stripped down pitch was very simple: the average internet user has multiple online identities/profiles that create or display information; our Identity Management and Open Networking services gives the control to link all these accounts together and push content notifications between systems - with privacy they control. We gave networks a chance to keep their userbase profiles up-to-date through user acquisition, retention and syndication systems; and offered advanced userbase profiling tools to help with advertising endeavors.

Just about everyone we spoke to didn't get that concept at all - most saw no need ever for a mechanism that bridges identities across networks; they were very much sharing a belief that dominant market forces were set in stone and people would never switch networks. They dismissed our arguments of MySpace converting Friendster users, and Facebook converting MySpace. This wasn't a matter of them not liking our idea - they just didn't fundamentally understand it.

A handful of groups did get our concept - and thought we were absolutely crazy. They thought our business would fail if MySpace blocked our widgets - and perceived a need for backing by 800lb Gorillas. They didn't care that a mere handful of small-to-medium sized networks encompass the same user-base, or that these small and fledgling networks were struggling for userbase acquisition and open to our ideas. They didn't care about increased fragmentation due to niche networks, and the explosion in smaller sites.

"These gardens are walled for a reason..." was a phrase we often heard. "No one outside of the top 10 networks cares about that" was our standard reply, "and where do you think we'll be 3 years from now? The small-to-medium networks we're talking to are your next big networks. Think towards the future".

They didn't care about the future. They wanted the present.

A few VC firms wanted us to build interfaces that would automate and sync all of the networks together - like a desktop app or meta-console. When I'd ask "Um, wouldn't the networks block us and sue us ASAP for stealing their users and blocking their revenue stream?" I'd get a bunch of "Oh yeah.. just throwing an improvement idea out there..."

Trying to work with open standards groups was even worse of a nightmare. The first place we tried was OpenID, sending requests to chat with the lead developers over the concerns of identity conflation -- to see if there were any ways to address that within the system architecture itself. Out of a half-dozen intros, I got a single response which essentially said "Yeah, we're really focused on this as a protocol, so thats not a concern." Months later it would be marketed by everyone as a complete identity solution.

I approached the MicroFormats group to try and get the open standards released for findmeon and OpenSN compatible with their group; they not only wanted nothing to do with the concept at the time, but I was barraged with what was essentially hate-mail for weeks by list members. Low-level contributers didn't like our licensing terms (which were actually more open and flexible than most Open Standards, including their own), disagreed with the approach of making something that just worked for the existing problem of networks (instead of making an academically ideal solution), and many simply hated the notion that someone would come up with a format elsewhere and approach them for integration - they wanted to originate. hCard and FOAF were dead-in-the-water though -- there were so many different online profile elements that kept expanding beyond their capabilities - an intermediary format like our OpenSN was the only solution.

At that point we just said "Screw it" and started talking with corporate groups.

The First Wave - Timid Market Acceptance

Things started to shift in the Winter of '06/'07. The market started to see an explosion in new online identity aggregators. Several dozen companies positioning themselves as either "Identity Aggregators" or "People Search" formed in October through December, launching products through March. There was a free-for-all: everyone wanted a piece of the action, there were no rules, overcrowding was rampant; sites started launching that were nearly direct clones of other sites with form and function. We even had a certain company clone our widget and API systems, then dump a few hundred grand into a PR campaign - we laughed it off and said "Can't wait until our patents clear!".

As eager as entrepreneurs and developers were though, they failed to convince network operators to buy-in. MySpace and the like were continuing to lockdown their profile pages, turning off widgets, disabling flash functionality, and doing almost anything they could imagine to retain users. As 'aggregators' became more commonplace, large sites started adding anti-hotlinking features to further lock the userbase down. The more people wanted to open networks, the more the networks locked themselves down.

This might sound user-antagonistic, but it kind of wasn't. For the most part, users didn't care. None of the Identity Aggregators or People Search systems ever really caught on. Pundits blogged enthusiastically about them -- but none of the players acquired a significant userbase. Casual users liked the idea, but the products weren't tailored for them so they were never motivated/excited. Continual network lockdowns didn't help the situation - but the product marketing was never on-key: Aggregators were designing themselves for the techno-nerds, people search groups were in their own little world, and groups like us were focused on R&D for Patents.

Around March '07 the market was completely flooded, and it looked like people might finally bite, but the blogosphere caught onto a new meme - one that we had been advocating for a year : the danger of conflating online IDs. Some large blogging figured started to question the ramifications of work and personal lives being connected online -- FindMeOn was (and still is ) the only system designed to address those concerns.

The Second Wave - Market Acceptance

Towards the summer of '07 things started to shift again. In order to stay competitive, Facebook launched their API/Platform - which would allow third party developers to create applications for their userbase. LinkedIn and MySpace quickly announced plans to follow suit. Plaxo began work on their Pulse network, and Viacom teamed up with Tagworld to create SocialProject. Just recently, Google announced their attempt with OpenSocial. In just a few months the larger networks and solutions providers went from reinforcing their walls to tearing them down. The floodgates were opening -- why?

Intro

I listed three core social networking interests earlier:

• The Userbase
• The Social Networks
• The Content Providers

There is actually a fourth:

• Advertising Concerns

I limited the scope of the earlier discussion as advertising concerns are a special group- they often overlap the other groups.

• Advertising Concerns are often found in the form of Advertising Networks
• Many Social Networks do internal ad sales / operate their own Advertising Network
• Many Content Providers are advertising oriented - celebrity , music , politicians , films , etc - they often have a marketing concern
• Advertising Concerns can also have nothing to do with the internet, looking at social networking data as a source of demographic information for lifestyle marketing and brand/product management

With that in mind, lets take a deeper look at these groups and analyze at the reasons why the market has shifted to a model of acceptance.

Why do these people want to open the walled gardens?

Understanding: The Userbase

Users are the easiest group to understand - they want/need portability of their accounts and information. On a daily basis users sign into multiple websites, looking at different sets of friends. Their 'event notification' streams are generally system-specific, their profiles are mismatched/out-of-date, every day they're pointed to a new online service and need to signup again. Despite how much users will claim allegiance to one-and-only-one network, they almost always have multiple accounts and are increasingly drawn to niche services. Users are irritated when forced to migrate to new networks because their friends have moved on, and they hate keeping active profiles on old networks because of stragglers who have yet to move on.

Understanding: The Content Providers

Content Providers (or Publishers) are interesting because they're almost a crossing of Social Networks and Advertising Interests. If you look at some of the more popular content providers out there - musicians, celebrities, politicians, non-profit groups - you'll note that they're all brand-centric identities with loyal constituents scattered across networks.

The only thing that motivates publishers is keeping in constant contact with the userbase. You'll almost never see a publisher doing an "exclusive community" on one network over another - they will homestead wherever possible. Smart publishers will offer "exclusive content" or contests on a network tailored to better engage those users on an intimate level, but they will never lock down to a single network.

Publisher use social networks to increase their brand recognition and reenforce brand loyalty. They communicate with fans not just to maintain interest, but to get a better understanding of them - so they can tailor their marketing campaigns, and sometimes even their products and services.

Understanding: The Social Networks

Social Networks are actually the most complicated interests to understand: as the consummate interface, networks are tasked with having to service both Users and Publishers ; as a commercial interest, networks must monetize their products and services.

The primary concern for Social Networks is that of Userbase Acquisition and Retention. Every network wants to acquire a userbase and keep it active: a growing active userbase is a sign of vitality. More importantly, a userbase is a network's primary income base.

Keeping the userbase active and happy is a chore. Networks can only use a set amount of syndicated publisher content- they must offer some sort of exclusivity to their users, which is why we see the various contests, targeted messaging, and network-specific content for the same publishers across networks. During a recent consulting meeting with a popular major label band , the members got phonecalls from three separate Social and Music Networks who were pitching new ideas for exclusive content.

The userbase is unfortunately fickle, with people migrating from network to network almost randomly. Once a critical mass has left a network, its extremely hard to regain traction with the same demographic. Facebook has capitalized on almost the entirety of the old Friendster core demographic ; Friendster is still growing but in different demographics, such as Southern Asia.

The secondary concern for networks is how they can monetize their userbase - and while some networks do this based on subscription services, the bulk of online networks are fueled by advertising. In fact, the only revenue stream for most networks is online advertising - repeatedly showing CPM or PPC ads to their userbase , and earning $2 or so for every one thousand page views.

In order to maximize revenue generation, networks typically do the following:

• Employ some sort of ad-targeting optimizer - displaying more relevant ads means more clicks
• Show as many ads as possible on a page
• Inflate the number of pages viewed to perform a task , so more ads are encountered ( i.e. the 3-step tasks on MySpace )

With this in mind, networks are in a very awkward position regarding Open Networking. While Open Networks can offer quick solutions to Acquisition, they ease the ability for users to leave.

Understanding: The Advertising Interests

Advertising Interests are easy- they want to make more money by increasing their ROI on marketing.

So Why The Change ?

The reason for change is simple: Increased revenue through advertising.

Until recently few people saw the value in Social Networks for effective advertising. Performance was so poor , that it couldn't be taken seriously.

Online Advertising is a multi-billion dollar industry -- but the social network segment alone is projected to only reach $865MM this year.

Why? The average CPM for a social network is reported to be between .10¢ and $4; premium networks with niche demographics can charge $2-8; highly targeted blogs earn upwards of $20.

Clickrates are theoretically improved by better targeting (that, and tricking a user).

• Un Targeted Ads - generic screenspace buy
• Semi Targeted Ads - categorized screenspace buy
• Semi Targeted Ads - contextual ad placement
• Targeted Ad - within a social network
• Targeted Ad - behavioral / clickstream

The larger behavioral tracking firms ( Blue Lithium, Tacoda, Revenue Science, DoubleClick, Coremetrics, etc ) are all pulling together disparate portions of your click history , and sometimes pulling in your Social Profile Data through a partnership the network. They're all trying to find ways to analyze your activity across the internet. These firms have been working with publishers , retailers, and networks to affect advertising based on usage patterns ( which is simply brilliant ), but few people have looked at social network profiles as an opt-in marketing system.

Not until the Summer of '07 does MySpace leak that they began testing a Social Advertising platform (now called HyperTargeting)- which targets ads based on profile information... and their effectiveness nearly doubled. Facebook followed suit with their SocialAds platform - which uses Facebook data for offsite ad displays. Google announce OpenSocial... and well, Google puts ads on everything.

Everyone made a sudden push to better monetize social networks when they realized how bad the performance has been , and trying to correct themselves.

When I first architected FindMeOn, I noted five main vectors for the monetization of online interaction that networks just consistently missed out on - and sought ways to capitalize on them.

• Targeting a user based on their profile. Believe it or not - as obvious as this is, networks have not been doing this since day one. ( Show an ad on MySpace based on MySpace profile )
• Bringing in external profile information to better target users. ( Show an ad on MySpace based on Last.FM profile )
• Exporting profile information to better target users. ( Show an ad on Last.FM based on MySpace profile )
• Turning anonymized site profile information into productized demographics for consumer market research and media planning. ( Create a demographic report on MySpace users )
• Turning anonymized multi-site profile information into productized demographics for consumer market research and media planning. ( Create a demographic report on MySpace + Facebook user overlap )

When I present the consumer side of FindMeOn, I talk about the Meta-Identity being the aggregate of someone's online identities.

When I talk about the corporate side of FindMeOn, I use another name which correlates to the world of consumer marketing: The Meta-Profile.

The Meta-Profile is a collection of a person's interests across the internet - as we look at different websites as parts of a consumer, we can fill-in-the-blanks and achieve more data with overlap. The more websites information we pull in, the more complete your profile is.

The kind of information that we're looking at gets to be quite amazing too - the General Purpose Networks provide a very good base of information with your favorite things nicely listed. But are your favorite bands your MySpace indexed ones? Or should we amplify your profile with the last 100 tracks you listed to on iLike and your top-10 bands on Last.fm ? Your favorite books and movies can quickly be correlated from the information you're sharing and commenting on in Amazon and Netflix. Even simple associations provide valuable clues to your Meta-Profile: joining sports sites shows which sports and teams you like, joining a website for moms suggests motherhood, joining a first-time-homteown forum adds even more context.

The benefits of Social Mapping Technologies aren't limited to individual targeting, and are better demonstrated with group applications.

By properly indexing and bucketing users across social networks, we can create actionable intelligence on group trends. Group applications are also much easier, as fewer privacy issues are encountered : personally identifiable information can only leave FindMeOn's servers on an explicit opt-in basis, demographic reporting uses only anonymized and aggregate data.

• Hillary Clinton has 150k friends on MySpace and 150k friends on Facebook. Does she have 300k friends in real life ? Or does she have come number between 150k and 300k due to overlap?
• Should a media planner spend $10k on $2/CPMs on MySpace, or $5k on $8/CPMs across 3 niche networks that have the same overlap with users.
• CocaCola wants to sponsor a Summer Concert Series. What are the most popular bands for 13-24yr olds in the markets they target? What other brands do they like for a corporate tie-in with other sponsors?

This might sound big brother-ish, but remember that we're talking about indexing public facing data, and maintaining privacy views depending on the source and consumer. Existing companies like Experian, and even the major Political parties do all of this with census data, public records, and credit reports. The industry term is "Data Overlays" - an euphemism for cross-referencing online and offline data from multiple sources to complete consumer profiles. When we designed FindMeOn, I mandated the system could monetize this information responsibly- ensuring user privacy and only dealing with aggregate data. We've been deathly scared of the ramifications that backtracking can have when global-ids are involved.

Five years ago, advertisers had to mine this data -- now its out and public, it just needs to be correlated to the right people. Products like Google's OpenSocial aren't designed for opening the internet -- they're just thinly veiled attempts at making it easier to Google to access and standardize profile material. We spent months getting OpenSN to support our target networks and allow users to import/export profiles -- and simple CSS changes can sadly break our regular-expression mappings. Widely adopted systems like OpenSocial become quick function calls that push the needed integration work onto the networks, not the standardizer.

There's been a recent explosion in the People Search market - Why? People search is utterly worthless on its own. If you're looking for someone, you can find them with Google or your addressbook -- almost always with better results. So why are people building and funding these companies for user interaction? They're not -- the goal of People Search technology is demographics and targeting. I'm not stating this as a skeptic, I'm saying this as someone who has been developing and patenting People Search technology for years: the only real monetization-point in people search is cross-site advertising.

In the Spring of 2007 Social Networks were scared of Open Networks because they heralded the democratization of online advertising. Networks didn't want to improve their advertising if it meant improving someone else's as well. Developments by Facebook and Google have given the networks the impetus to join up - or lose out. The democratization of advertising is not a bad thing - the networks themselves don't fully understand their own users - by integrating cross-site data, they can have a better idea of what their users are interested in.

Designing Our Solution

Its easy for me to talk about the Facebook , Google and MySpace initiatives, because we were working on the same stuff several years ago. When we architected FindMeOn, we set very specific goals that would balance the interests of all the players.

Users

• Centrally Manage Online IDs
• Provide linkage between identities, and track how people visit them
• Perform Cross-Site Searching and Contextual Friends - bring 6°-Of-Friendship across networks
• Open the networks to allow for quick integration across their personas - content update notifications and portable relations
• User Privacy - Protect people from People: Stalkers, ID Thieves, Compromised Workplace
• User Privacy - Protect people from Data Overlay exploitation by publicly linked ids
• User Privacy - Protect personal information from being discovered through advertising systems

Publishers

• Map users across networks to simplify the ways in which you continually engage with them on this networks
• Better understand your userbase through aggregate reporting on cross-network social trends
• • Use this information to better engage your userbase based on their interests
• • Use this information to better plan where you engage your userbase

Networks

• Always be beneficial and not antagonistic
• • Don't steal users or content , that is their revenue stream
• • Don't offer competing services
• • Make it easier for publishers to communicate on networks , don't bypass the networks
• Address Acquisition of users through Rapid Registration
• Address Retention of users through Content Stream Notifications and the ability to internally focus efforts on better interface & unique promotions
• Should a user leave or be inactive for months, refresh their profile off the network specific information they grant through FindMeOn. Even if they no longer use your service, other people will stay - to them the information is up to date.
• Better understand your userbase through aggregate reporting on cross-network social trends , translate that into higher rates for certain ad demographics

Advertising Interests - Brands

• Better understand consumer base through cross-network-profiling for aggregate trends analysis
• Better plan media buying by finding the networks that your target demographic interacts with better

Advertising Interests - Advertising Systems

• Better target consumers based on cross-network meta-profile

When we first released our platform in August '06, the market wasn't ready. Now- that's a different story. Over the past four months we've seen every single one of our predictions hold true and initial tests proven true:

• Users began to worry about social identity conflation
• Advertising based on Social Profiles proved itself to be incredibly efficient
• Networks accepted the opening of their doors on the premisses that
• • For every person that leaves, you're making it easier for 10 more people to join.
• • They do not want to end up like Social Network ____ with 20MM user profiles, 80% of which haven't logged in since 2004?
• • They see the utility in driving up their own advertising rates by aligning themselves with other networks to complete consumer profile information

Closed Networks

Closed networks are what we're all used to: a captive userbase that can only interact with applications offered by the network. The classic example is MySpace or LinkedIn ; these are networks that are "user-antagonistic" in that the network owns the data and the relations.

Opened Networks (Leased Networks)

Opened networks rent/lease their userbase to third party developers. They're not truly open because everything is mediated through the network, and an external account isn't created. The Facebook API is the best example of a closed network ; leased networks are "developer-antagonistic" in that they own the userbase and can terminate their involvement at any moment. Leased networks have also been known to use 3rd party applications as a model for their own official versions, eventually competing against (and possibly blocking) those they lease users to. There is no conversion of users from one network to another, the platforms only have certain things open, and its very clear that the operators want to convert and retain every person in the world to their network.

Open Platforms is really a euphemism. They're a step in the right direction, but are actually "Limited APIs". In terms of marketing buzz, 'Open' is a lot better than 'Limited'. I think an appropriate summation would be: You can call it Firefox if you want, but it looks a whole lot like internet explorer to me.

Open Networks

Open Networks allow users to wholly own their data, and migrate their data between networks of their choosing. They often use CreativeCommons licensing on OpenStandards and APIs and offer OpenSource solutions. Open networks can either do centralized data migration (FindMeOn) or decentralized migration (Google). The only con of Open Networks is that they can conflate user identities and relations unless they are designed with user privacy in mind.

Option 1 - Collapse the "Social Graph"

The first school of thought is to collapse the Social Graph - at its essence, this means to relate everything to a unified person or identity.

Collapsing the graph is really popular with the OpenID crowd - but its a concept that ( just like OpenID ) is suited to a certain purpose - and not advisable for others.

OpenID started as a decentralized ID system used for blog comments... and it quickly spread to some blog oriented social networks and platforms. Under that usage pattern , it makes a lot of sense - its a single facet of someone's identity and they want all that to be connected. The OpenID team came up with a great framework for their target audience - bloggers and livejournal-ists who lived outward lives and wanted everyone to know everything about them.

But there's a large difference between that and people who - as they get older - start to worry about the different facets of their lives coming in contact with one another. When you collapse the social graph, all of these elements point to a universal identifier.

Once items are all linked to a universal identifier, a little bit of creative googling means that each facet and blog post can be attributed to the rest -- you have a full backtrace possible: your information is consolidated neatly for the exploiters, and your work and casual lives can influence one another. Not good.

I find it hard to discuss this approaching without ranting. Calling this "Not Good" is a gross understatement -- this approach is utterly irresponsible. Promoting OpenID and social graph unification like this is tantamount to tossing a bunch of preschoolers box-cutters and saying "have fun and draw on each other!". It is such a bad idea , and I'm completely amazed that this has received any bit of traction -- much less the widespread appeal its had. This 'solution' does nothing but conflate personas and lose the factor of 'network significance' that is so important.

Option 2 - Map the "Social Graph"

The other school of thought is to intelligently Map the Social landscape - this way we retain the network significance and are able to keep our identity facets separate.

As my business parter like to say- this lets us "Get Found or Stay Lost".

The idea behind this is really simple - if we know where all the points on the global "social graph" are (i.e. who is who on which network) , then we can treat them as simple nodes on a global network.

It's also neat, because we get to 'plot' the graph/map -- which is one of the few times these allegories make sense.

With this approach we can retain the contextual information of the network - which is extremely important to the end user: the network a friendship exists on is an important contextual descriptor about a relationship - much like an XFN attribute. The relationship you have with your average LinkedIn connection is much different than your Facebook or MySpace; Do you want your next employer to be reading your Facebook feed ?

To the networks themselves, the contextual clue of the network is like having a brand recognition built into the relationship - which is an important concern when debating if they should 'open up'.

The stereotypical image that people have been passing around the internet to explain relationship porting is the classic description of 'Network Ownership' of friends -- where your friends are clustered under networks.

To counter that, people have been advocating an identity-based mapping of online relations - where the networks are clustered under the connections.

The problem with this solution, is that you then expose your identity and relations on networks that you may have wanted hidden. In this example we weren't friends with Adam on LinkedIn or David on Myspace - by unifying the graph and going for a solitary ID , we can expose ourselves to them on the wrong networks. If that doesn't sound like an issue to you, how about this: David is the Managing Partner of a VC firm that we're negotiating with , Adam is an old friend who is in a punk-rock band and just entered rehab on court-order.

We can avoid all of these problems if we map (or index) the graph instead of collapsing it.

By preserving the networks in Social Mapping - not as the owners of the relations, but as contextual clues to the relationships - we're able to connect-the-dots on our own terms.

• We can expose only certain-networks/elements to our contacts and vice versa.
• We retain the ability to have network-only privacy
• We gain the ability to browse through the lens of friends

Indexing the graph means constantly interpreting and aliasing the data based on your current positions -- and having the linkage as possible waypoints when traversing from one network to another. I'd love to think of my own relations with people as not-network specific -- but they are, and I enthusiastically don't want a collapsed / unified graph.

With a properly indexed and mapped graph, we can create 'views' of the information - I can group my relations to Carl based on who he is , but only be privy to the parts of Carl he wants me to see. Proper indexing gives me the flexibility to view relations by person, but the makes me respect permissions rules that affect visibility based on network and overrides ( i.e.: friend, family, etc )