One Graph To Rule Them All?

My partner Albert posted this on his blog yesterday:

But I see at least one flaw with this plan for domination. I simply don’t believe that there is a single social graph that makes sense. I may very well follow someone’s booksmarks on that I don’t want to have any other relationship with. Or take the group of people that I feel comfortable sharing my foursquare checkins with — these are all people I trust and would enjoy if they showed up right there and then. That group in turn is different from the people I work with on Google docs for various projects which is why I would be nervous about using the Microsoft docs connected to Facebook. Trying to shoe-horn all of these into a single graph is unlikely to work well.

He's talking about Facebook and the new services they announced this week at f8. Clearly Facebook is executing fast and well at a scale that few Internet companies have ever reached. It is very impressive to watch.

But I am with Albert on the issue of one social graph for the entire web. I don't think it will happen either. And we are putting our money where our mouths are by backing companies like Twitter, Tumblr, Foursquare, Disqus, GetGlue, and others (remember that are building different social graphs.

Today Facebook is the mainstream social network and where most people keep their entire social graph. The other services I mentioned, with the exception of Twitter, have not built large mainstream social graphs. But I think they will for the exact reasons Albert articulated.

I want to share some things with the widest group of people that is possible. Those things end up on this blog and/or Twitter. I want to share some things with the smallest group possible (like checkins on Foursquare and financial transactions on Blippy). That behavior requires a very tight, very private social graph. 

Facebook sits in the middle of all of this and has created the largest social graph out there. These other social graphs can and will grow in the wake of Facebook. I am not sure if Facebook's ambition is to create the one social graph to rule them all but if it is, I don't think they will succeed with that. If it is to empower the creation of many social graphs for various activities and to be in the center of that activity and driving it, I think they are already there and will continue to be there for many years to come.

I want to thank my friend Mo Koyfman and my partners Brad and Albert for helping me crystalize my thinking about this over the past few weeks. It is a very important topic for those of us who invest in the social web.

Reblog this post [with Zemanta]

#VC & Technology#Web/Tech

Comments (Archived):

  1. kidmercury

    not going to get anywhere until we have the ability to leave the cloud — until our participation in the cloud is voluntary. i remain skeptical if any of the big platforms can do this, i don’t think they are structured in a way to make it economically viable for themselves.facebook’s 180 on privacy is worth highlighting here IMHO. when you have 400 million+ users, you can’t make massive policy changes like that whenever you feel like it. that sucks for action-oriented entrepreneurs, and i sympathize with that, but the government needs to be setup before you get to 400 million users. imagine if instead of the US constitution being developed when there were 13 colonies they tried to roll it out now, with an incredibly diverse and distributed population. same for facebook, they needed to roll out their constitution back when they were colony-size, not empire-size.of course facebook, like all these companies, is basically a dictatorship. IMHO data portability and governance go hand in hand — the companies that bring us data portability will have a far more meaningful understanding of their role as government. and such companies will be able to bring us the convenience/integration of one world internet while ensuring we are not trapped into just one system.

    1. CJ

      True but you won’t get data portability until data is not the most valuable piece of social media. If Facebook or Twitter let you take your social sphere with you wherever you go, why would you be incentivized to stay? Safety and comfort are two of the biggest forces that impact human behavior. When you set up shop and kick your feet up and feel safe, you don’t want to leave, even if you see the neighborhood change around you. It takes a series of small changes or a huge drastic change to erode that sense of comfort. It’s a main reason why people stay in dead end jobs much too long. You might not like Jim and Bob from accounting but you KNOW them and that makes you feel comfortable and safe, even if you know that those feelings are largely illusionary.

      1. kidmercury

        Facebook and twitter do not have the structure where by they can make it profitable and let you take your data. That is why I think them and that whole class of companies like them cannot bring us a data portable web. For that, we need open source companies, stuff like onesocialweb, or, acquia, drupal….a different class of companies with an entirely different approach, much like how goog has a massively different approach then msft. That is also why I think all of these companies compete too closely with google.

        1. Mark Essel

          Agree with Malcolms behaviorial rationale, but for social networks the ability to rapidly jump ship makes data lock in a weak motivator. Otherwise Friendster or Myspace would still be king of social platforms.Totally agree with you Kid and commented somewhere that open data will be the norm, and high utility touching/filtering will be the dominant business models of tomorrow

  2. Hank Williams

    Well, I think you are right, but I think/fear that it will not be that hard for facebook to add the layer you are saying is missing. The next step for fb will be classifying types of friends, creating some kind of sub grouping. I think they could make, for example, a LinkedIn inside fb, or a foursquare inside fb. It would work just like friending someone does now, but it would be kind of “special friending” within your current friend list. I don’t think there is any technical or even social barrier to doing this. The only problem is potential political backlash like dave winer is doing. But I suspect people still primarily trust fb and so the “dont let fb gain control over identity wont succeed.On another note, I have been repeating what dave and others have been saying, which is that fb is trying to take over “identity”. Waking up this morning, I dont think thats true. I think they want to own your relationships, not your identity. I think that is even scarier. Its the *one* thing they wont let you export.

    1. fredwilson

      i have never seen that type of classification work well in consumer softwarethey already have it to some degree in their privacy settingsand they are a mess. i can’t figure them out.

      1. Hank Williams

        I agree what they have now is too hard, but the fact that what they have done up until now is hard doesnt mean they cant get it right. I am sure I am in the minority, but i think *lots* of things in FB are unintuitive right now, but cleanliness (from an api perspective) of this latest release suggests to me that they very well could get the UI right. Personally, I think it would have to be something you did on an app by app basis. Then it would mirror signing up for a new service. I just dont think betting that fb wont be able to offer a better interface in the future is a good bet. They have already shown themselves to be very nimble.

        1. Scott Carleton

          I see what you’re saying, and on paper I agree that it looks like facebook could institute some of the changes you’re suggesting. However, my feelings are that FB is much more of a slow behemoth with a conglomerate of features that dampen the user’s experience. I have been on it since my early undergrad years when it first debuted, and although I’ll always maintain my FB account, it is just too arduous to use effectively for good on-the-fly social networking. A lot of they’re features require me to be in front of my computer. Good social networking requires little to zero management, is fully accessible from my mobile device and can be as localized or as broad as I intend it to be. This is why I can’t conceive of one network becoming a singular social graph since it won’t be able to become top dog and maintain the accessibility that I require. FB needs to become leaner and more intuitive to even stand a chance but I believe they’re too focused on features right now.

          1. paramendra

            Stand a chance? Facebook just beat Google in page views.

        2. Mark Essel

          Yeah the FF effect on their API is apparent (from what I hear). Brett Taylor and crew are doing some good work at Facebook.I wonder if Mark views the Friendfeed team in the same light as Ev and Twitter saw Surmize?

          1. paramendra

            Good parallel you just drew.

      2. CJ

        If they ever figure out how to make privacy simple AND effective, they could truly rule social media for years to come. I have a lot of friends who are contemplating deleting their account because they feel Facebook is getting much too greedy wrt their thirst for sharing personal information.

      3. neilredding

        Facebook’s Friend Lists was my first thought as well – my second thought was that I’ve tried using these but found it easier to just think of my Facebook graph as monolithic, and wonder what the stats are for adoption of this feature.If FB could make it usable, it has potential to address the multi-graph problem. I have serious doubts about this possibility though – especially given the simplicity of this week’s new functionality. I don’t want to select which friends should see my Like-ing activity when I Like something – too cumbersome.

      4. David Ashwood

        Privacy on Facebook types of platforms is pretty complex to represent and manage.Even MS have issues with their Active Directory when used for Single Sign on.

      5. parsnip

        Why not just have some default groups? “Friends”, “Work”, “Family”, and so on. People can also add their own custom groups.When I have an Microsoft office doc and hit “share” I then just check off my “Work” group and leave the other groups unchecked.Too cumbersome?

        1. richus

          This is a simple approach, and I’d use it for sure. Right now I’ve got sort of FB paralysis…..for instance, I sure wouldn’t post how well a job interview went, because I have a couple of FB friends from work (which I think was a mistake, but I didn’t want to turn down their friend request). More paralysis: I don’t really want to share that I “like” a really funny, but off-color, article on The Onion, because some friends would be really offended.It’s getting to the point that I don’t post much of Facebook because of the wide swath of “friends”.Speaking of job interviews – the first thing I do before a job candidate shows up in my office is check out their web presence. I’m sure when I interview, someone is doing the same. Do I want them knowing my Pandora list, political group, or see pictures from my brother’s bachelor party? Nope. It should be fairly simple to click on a friend’s name, and have their “friend category” list available to check.

      6. muratcannoyan

        To a large degree ease of use is more important than even privacy. Most people have a decent sense of what they are getting into when they connect to a social network so they don’t deal with the complex parts like privacy settings. I feel like that’s why multiple networks will work best for most people.

    2. dipen chaudhary

      I think it will be very hard if not impossible to pull that off and I am sure the thought might have crossed their mind seeing successes of linkedin, foursquare and other networks. The layer or the classification would end up consuming users, there has been lot of talks of building workplaces as social/fun as facebook and one of the original facebook founders Dustin Moskovitz new startup is about making collaboration fun at work. Facebook is interesting to me from the social point of view and bringing office to home has never worked really well to factor out the problems of professional-personal life balance. I think it would be hard.

    3. Mark Essel

      That’s right. And that lack of portability imakes crystal clear the difference between open networks and supposedly open networks.Agree that FB can easily enable subnets inside of it’s identity and relationship gated community.Simple lists

  3. itamarl

    While I don’t believe Facebook Open graph will become the web’s only social graph, I’d tend to disagree with the Wide vs. Small group argument.A social graph is more complex then just a group of connections. Whether in real life or online, a social graph has different layers, from family to friends or your wine tasting group. One open graph that can handle these different aspects could simplify our online activity. The same way that I use only one address book (Google contacts), with private, family, business categories.Facebook will probably not achieve that for me, as I have historically built my Facebook network as a wide graph, and I would indeed never use it on Foursquare.I still believe we might get something out of initiatives such as http://onesocialweb.orgAn open and decentralized social graph that becomes part of the web infrastructure, as every single service we use becomes social, there will be a need for some standardization there.

    1. kidmercury

      yes, i think onesocialweb has the right idea. i think we’ll see competing versions of that, which i think will truly empower users, rather than the current, unportable SaaS model.

    2. CJ

      I agree with the layering and I have a wide variety of Facebook friends from Friends and Family to business contacts to ‘friends’ I’ve never met but share similar interests. I group them all and configure privacy settings to keep things in order. The best thing that Facebook did for me was to allow me such granular control over the content of my page. I can set status messages that only my family can see, how cool is that? That said, my twitter account mostly follows internet and stock type folks and none of them are Facebook friends and I’d feel awkward if they were. I don’t know, I don’t want Facebook running my social network everywhere but I LOVE the ‘Like the Web’ button. We’ll see I guess…

      1. fredwilson

        that’s how to make facebook work but i don’t see the vast majority ofmainstream users doing thati never have and i am not mainstream

        1. Rocky Agrawal

          the UI for managing lists on facebook is awful. if they improved that it would definitely make it more likely. but still it wouldn’t be a substantial portion of the base.what facebook could do is automatically cluster people by their other relationships. my high school friends, college friends, work friends, etc. cluster together. you could use the graph to programatically suggest lists. then the user only has to deal with a small number of random people and people who overlap lists.

      2. Mark Essel

        Nice tweaking but I agree with Fred that most users won’t create the same subnetworks.I only feed Facebook content I generate outside it’s walls and I’m satisfied with that relation to it’s network.

    3. Mark Essel

      I prefer an open standards web with easy access and interoperability as well as being fully distributed. The problems with standards committees is they iterate too slowly. How can we provide a framework where we get the best of startup explosive speed and behavior change and open webstandards?

      1. kidmercury

        first we need to have small businesses built on open standards identify profitable models. than they need to share development costs of their standards. the standards committee thus works to amortize the development costs of the standards for the businesses built on top of those standards. in this way, the platform is really the cost center, while all the apps are the profit centers.the problem we currently have is that everything is in this world where the platform needs to be the profit center. everything on that trajectory is on the “one world internet” trajectory. when the platform becomes the cost center where niche businesses amortize their shared costs, then we’ll get the right environment IMHO.

      2. Daniel Kligerman

        That’s exactly what I’ve been thinking while reading this post and these comments–an open standard, which creates interoperability between all of the above mentioned sites, seems like the ideal open end-state. I think this is what Dave is getting at as well.Isn’t that exactly what Google’s Opensocial is supposed to be? Perhaps we are just not far enough along to see adoption of this type of open standard. If/when we do, it seems that this would allow for interesting relationships between multiple social graphs; you could still maintain each graph for a specific need if you want, but share information between them if desired.

        1. Mark Essel

          A standard is measured by it’s adoption. I could come up with the greatest open standard in the world, but if no one adopts the pattern it’s not really a standard, it’s an idea with perhaps a single implementation. The biggest limitation to the open social standards that Google and others developed a few years(?) back has been wide spread adoption.An interesting example of emerging standards is the twitter api. When wordpress, tumbler,, and tweetie all adopted the API it had the feel of a community standard. Now there may be legal ramifications to mimicking APIs but there’s also increased accessibility, and the potential for greater value flow and growth throughout the network.

  4. Berislav Lopac

    In my opinion, the only sensible future for social networks is for them to become decentralized. That way it’s you, the user, to determine which graphs to belong to, and how and when, instead of waiting for a service to incorporate that feature.

    1. fredwilson

      i agree with thatbuilding them application by application is fun

    2. falicon

      I think this is exactly what we are seeing with the emergence of things like (yesterday) 🙂

      1. Berislav Lopac

        I guess there is irony in the fact that this page prominently displays the Facebook’s “Like” button… 😉

        1. fredwilson

          are you talking about the “share on twitter and facebook” at the bottom of the disqus comment window?

          1. Berislav Lopac

            No, I meant the “Like” button on the OpenLike page Kevin has posted a link to. Sorry, I guess I should have written “that site” instead of “this site”.EDIT: Scratch that. I just realized that’s kinda the point of the whole exercise… Nothing to see here. Move along, move along!

      2. Mark Essel

        So the future plays out like so:Startups and bigger corps introduce new behavior. People recognize the value of opening that behavior up, then the marketing battle for open source mindshare begins…Groovy trends, it pushes companies to faster behavior introducing loops and if they embrace the open advantage of networks they can win the open source mind share battle before it begins.Don’t like what Facebook, Twitter, or even Disqus are doing? Fork it!Ownership of social data & gated meters isn’t a business model that will last. Touching open data and providing utilities of smooth design and user value is.

    3. robertavila

      Facebook annoys not just the anti-social but also those looking for different forms of “society.” My 23 year old daughter complains that “everything and everyone is so phony on Facebook. No one has 327 friends.” An artist/poet friend looking for a an artist community where works could be shared and reviewed by fellow artists spoke savagely about Facebook when it was suggested as a possible venue. Other friends have described the layering of their social relationships and the efforts they make not to mix family and friends and work and professional relationships and then marveled at people who have it all out there on Facebook. A French mother whom my wife met at a play date years ago said that she would be ashamed if her child were “popular.” I suspect there are markets for less generic and more targeted social networks.

      1. dipen chaudhary

        I too think there is lot of scope for niche social networks to keep the noise away, though its interesting to note that all of these niche social networks wishes to connect to facebook graph one way or the other by implementing f connect etc to leverage their user base which in theory looks futile, we took an attempt to understand niche networks some time ago –… Would appreciate input from people on this.

        1. fredwilson

          they need to draft off of facebook to build scale, but then they shouldmostly detatchfoursquare is a good exampleinitially, if you connected to facebook, they sent every checkin to facebooknow it’s a checkbox that is off by default

          1. Laurent Boncenne

            tho to me it would have much more usefulness to share the places I check into with my so called 250ish friends rather than spamming them with my twitter updates…

          2. Rich Ullman

            Hmmm… that sounds like opt-in — which is, according to most people, a best practice. That’s good. Facebook, however, seems to treat it the other way around every time it comes to privacy and social graph issues.

          3. Michael Jung

            Yep, I think we covered that already (if I am not mistaken).”Defaults are powerful”.

      2. Adam

        Many business i speak to feel this way also, Facebook is not a place their customers are comfortable being so why should they bother investing in being there?

  5. Mike McGrath

    Great companies are built around basic needs. Facebook is the highest expression of our need to be connected to others. Human connection becomes depersonalized with a graph, and our need for privacy is almost as strong as for connection. Google tracks you innocuously; Facebook does overtly – which I suspect will bother Main Street a lot more than it does their in-house groupthink.

  6. Kevin

    Kramer: ‘Jerry, don’t you see? This world here, this is George’s sanctuary. If Susan comes into contact with this world, his world’s collide. You know what happens then?’Kramer: ‘Ka shha shha shha Pkooo!’I agree with you Fred – social interactions are all different and most people like it that way. I have vastly different uses for FB, Twitter, etc., and I don’t want them to be all one stream.

    1. fredwilson

      i remember that episodethat was a good one

    2. awaldstein

      I’m with you on this Kevin…both in Kramer appreciation and bifurcating my messages for different networks.

      1. Kevin

        Different messages for different purposes is a big thing for me — but I also think it is an issue of a variety of networks doing things differently and uniquely and better than one big network doing all of them so-so. If FB wants to become the Wal-Mart of social, so be it. Certainly could be worse business models to follow. But they will turn off a lot of people that prefer more custom and better experiences for various aspects of social. My guess is that no matter how big and successful FB is it’s a net win for many smaller players as they make social more acceptable and normal for the people that don’t currently use Twitter, Disqus, etc., etc. I simply don’t trust FB, and I do trust all the others, and I think I am not alone.

        1. awaldstein

          AgreedI’m a believer… and a participator…in the small and the niche.Disqus acts as the connector for me.

          1. Kevin

            Though it is probably my least used of the social networks I think Disqus is my favorite of all of them.

          2. awaldstein

            Facebook is good for sharing with people you know; bad for conversing and making new friends.Disqus is great for real conversations and broadening your contacts on a deeper level.

          3. ShanaC

            I have friends who would totally disagree with you on that- from practical current experience.

          4. awaldstein

            Hey…could be.I’m a public person and use Facebook, Twitter and blogs in different ways. As I look at the growth and value of my networks, the traffic sources (by quality as well of quantity) to my blog, this is how it is playing out for me so farNo doubt it could be different for some folks.

          5. ShanaC

            Off line.

          6. paramendra

            Offline is a whole different ball game. But offline is not great by definition. Between the NYTM after party – something I value enough to show up every time – and the AVC comments sections, I must say the AVC feast is better. Offline? The best would be a combo. I can’t wait to get the USV job so I can start putting together offline gatherings for the local AVC regulars.

          7. ShanaC

            I just wanted to make a private comment that was slightly more specificabout said friends. Certain things you don’t say in public. That was all.

          8. paramendra

            Agreed. I don’t think privacy is old fashioned.

          9. paramendra

            “Disqus is great for real conversations and broadening your contacts on a deeper level.”I need to do another blog post on Disqus. So true. I did not start coming to AVC posts to meet The Regulars. By now I know who they are. And I have to read many comments by one person here before I will click over to their blog or Twitter page. For the longest time I did not know Kid was Indian like me. I was thinking he was a black dude. Just look at his profile picture.

          10. paramendra

            Disqus might not have the buzz, but to those of us who use it – I use it more than Facebook, Twitter, Tumblr – it sure gives the buzz. But here has to be mentioned. It is important to point out the culprit.

      2. paramendra

        I use Facebook to connect, Twitter to broadcast, Tumblr to listen, Disqus to discuss, etc. Blogger remains my favorite social media platform.

        1. awaldstein

          For me..these are different channels for different communities…and all moving to my URL as the center of the world (hopefully ;))

    3. Scott Carleton

      I also feel that I have vastly different uses for FB, linkedIn, Twitter etc. Now what would be useful is having them all on one website but still siloed from each other. That way I could manage their throughput from one location but still have my worlds separate.

      1. Kevin

        I think that is where Gist is headed. It seems like it could be pretty cool and useful by doing what you just said + email, etc. Currently I find it a bit confusing, but they are in Beta. If they get it right I think they are what you want … and so do I.

        1. Scott Carleton

          Thanks for the heads up on Gist

      2. andreaitis

        a mash of friendfeed and tweetdeck?

      3. ShanaC

        I’m still figuring it all out- I find this all very funny.

  7. falicon

    Your second to last paragraph is spot on…I don’t think Facebook really wants to create one master social graph to rule them all (ala Highlander)…but I do think they want to continue to position themselves to be in the center of as many of the social graphs as possible (ala The Godfather).What’s especially interesting to me is that in many ways it seems like there is a higher chance for making money the more focused the social graph is…and the more ‘general’ the social graph is, the harder it is to make money with (because the general social graph lacks intent).

    1. fredwilson

      i think Facebook is going to make a ton of moneybut there is a lot of value in the niches for sure

      1. falicon

        I agree they will make a ton…but not just for having the largest general social graph…I think they will have to segment that graph down into focused and interesting things with intent (like the facebook credits concept for example).

        1. Rocky Agrawal

          having this much visibility into a user’s activity positions facebook very strongly to become a highly targeted ad network. forget the social graph aspect — just knowing the diversity of sites you go to and being able to combine that info with the demos and personal information you share on the fb site. that’s extremely powerful.that would be equivalent (if not stronger) than google combining doubleclick browsing data with google search history.

      2. paramendra

        Same applies to search. Niche search. One could argue Twitter came out with the ultimate niche search product.

    2. Mark Essel

      Intent is a big driver of monetizable action. Chris Dixon has driven that home in a number of posts and that’s a big driver behind Hunch.

    3. kidmercury

      yes. 100%. google has already won the “general social graph” game. if not google, one of the big players has. email and search are the winners is niche, and will totally be about getting maximum share of wallet rather than maximum number of customers. depth over breadth IMHO.

  8. reece

    I’m with you. My Facebook connections now include people I’ve met all over the world, from hometown schoolmates to backpackers in Cambodia… that doesn’t mean I want their influence all the time though.My social circle on Twitter, however, leans very tech/startup/entrepreneur and that’s the way I like it.Regardless of how individuals tailor their connections, I still think individual services are able to provide a better user experience than giants like Facebook. Facebook is becoming an institution and that’s not necessarily a good thing.

    1. Greg Gentschev

      Interesting how many people now have more tailored social circles on the “open” network, Twitter, than on the “closed” network, Facebook, that became everyone’s default friend mechanism. There’s much less social pressure to follow back on Twitter than there is to accept friend requests on Facebook (see South Park).

      1. Fred McClimans

        Greg – I totally agree. I’ve thought for years that Facebook should have an “acquaintance” feature – not quite a friend, but somebody that you want to keep in touch with, that comes with it’s own set of pre-defined/limited privacy/sharing settings. It takes the pressure off that “friend or not” decision immensely.As for Twitter having much more tailored social circles, I think that is because there is one simple setting for privacy: all or none. You know exactly what you are going to get. You either open up your life to the world (and tweet accordingly) or you lock-down your account and restrict access to your “private” circle exclusively. Simple, and it works.

        1. Fernando Gutierrez

          Also, everyone in on Facebook and not so much with Twitter. And the average twitterer is really different from the average facebooker.

      2. reece

        Great episode.

    2. Mark Essel

      My Twitter net is a startup/tech news mashup. Great example.

    3. ShanaC

      I have way too many people on Twitter, and the ones that post a lot are tech/startup entrepreneur- and oddly, I don’t like it- I want to say I’ve seen something else of the world.

  9. Tereza

    I agree very much with this post.I’ve been thinking a bit about social graphing in the context of my business, which people can use to poll others on their appearance. (And apologies for the shameless plug.) But it’s really relevant to what where we want to go.What we’re doing is like bite-sized market research on yourself…before you go out on and check in on Foursquare. (A sub-thesis is: feel confident that you look fabulous while you’re checking in and people might see you)So…you may not want to ask your Foursquare friends, because they’re the ones you’re trying to impress.I’ve designed, managed and executed a decent amount of professional-grade research in my lifetime. As important as asking the right question, is asking the right audience. You ask the wrong audience, and your result can be vastly misleading.The way I see it, there are some questions you would feel OK about asking your own Facebook network (e.g. should I wear my hair curly, or straight?) or maybe a subset (e.g. just the business professionals — if you’re planning for a job interview).Other questions, you may prefer to ask people who are demographically similar to you but perhaps opposite gender (e.g. what should I wear for this date), or are open-minded to certain decisions (e.g. should I get a boob job?)Still other questions, you want experts (e.g. a bunch of cosmetic dentists about whitening vs. veneers). Sometimes you would want a broad cross-section of your own social graph, people that are like you. And other times, a cross-section or people who are not like you (e.g. what to wear to meet your future in-laws). So it is also possible, based on the question, the outsider opinion may be of value to you.So to me the challenge is twofold:(1) can/should a service guide users toward the “right” group is to ask, and (2) is there a ridiculously simple/elegant way to slice and dice people inside and outside of your social graph, for what you need.I want to start extremely simple. But as we all get more sophisticated about our graphs, I see consumers pushing the requirements on that utility and there are opportunities in this.

    1. robertavila

      As always you are so articulate and so ambitious

      1. Tereza

        awww. warm fuzzies, Robert!

    2. Mark Essel

      As always I do my best to slice through the complexity of changing how I appear to different groups. I prefer genuine consistency, to both business, and private connections.”yeah that’s Mark”What you describe sounds like a perspective engine by empowering super human filters that are both manually and automatically identified.I do agree about getting feedback from the right group, so at least on one level I can comprehend this. Usually that group is the one I trust to be genuine and give me their respecitve take.Cool problem space, I think Michelle might better appreciate this as she dresses and prepares vastly differently depending on who we hang out with. Me, I always wear a cumfy shirt and shorts as long as I don’t freeze. Wearing and feeling clothing has always trumped others perception of my appearance in my book. Maybe I’m satisfied with giving the signal “Mark dresses for optimal comfort and cares little for convention”of course I’m wearing a tuxedo at the wedding :()

    3. RichardF

      so “hot or not” meets “chic or cheap” then 🙂

      1. Tereza

        you betcha!

  10. Hank Williams


    1. fredwilson

      90+% of all software is left on default settings

      1. ash bhoopathy

        is there a statistic around that? i could use that as fodder!

      2. David Semeria

        The same is also true of Terms and Conditions for software/sites, as Nic Brisbourne very amusingly showed.

      3. Fred McClimans

        I’ve seen the term “90%” bantered around a good bit. Do you have a source for that nugget of data?And yes, I agree that there is no single graph to rule them all. People use different social networks for different purposes – a point that even Facebook realizes, but doesn’t seem to “really” realize.During one of our regular chat sessions on Twitter yesterday, the subject of this new feature was hotly debated with regard to the use of Facebook pages for corporate and investor relations activities. With Facebook’s new approach, I expect that many Facebook users that currently blend their personal and professional personas will likely back off such an approach. Just a thought.

      4. Scott Carleton

        Exactly. FB compounds the issue by making it confusing for the average user to even find their privacy settings and to understand what they entail. If FB was a more upfront platform, starting people at the most secure level of privacy and allowing them options gradually to become more social then I think they would improve people’s perception of them.

  11. Bruce Barber

    I think you’ve hit upon an important point: that the social Web has become kind of a digital “Swiss Army Knife”, comprised of different tools for different purposes.I use my Facebook page as a very public space to stay connected with different people from my various lives – the kid from Buffalo, the college student, the radio personality. I use LinkedIn and Twitter for business, Disqus for commenting (on only the best blogs!), and Foursquare for connecting with friends and business acquaintances in my physical communities.Any one of these, taken separately, would provide only a snapshot of my graph. Aggregating the feeds makes sense as a business model (a la FriendFeed), but, as you have noted, one loses the ability to “sandbox” different relationships.That said, I believe the “Like” button has the ability to become Facebook’s killer app. Publishers are given an opportunity for content to spread virally on FB’s vast network, and Facebook gains valuable insights into what that network considers important.

    1. fredwilson

      yes, the like button is powerfulbut as albert said in the post i linked to, “share on delicious” has been onthe web for five years at leastand share on twitter is very popular nowi don’t believe most publishers and e-commerce services will use just oneshare button

      1. Bruce Barber

        True – I shared my comment, above, on Twitter, but not on Facebook…

      2. Rocky Agrawal

        i think the big difference here is that facebook exports the data onto the publisher site. while i’m reading an article, i can see the likes/shares of my friends. it drives recirculation while i’m already on the site in addition to traffic from facebook profile/news feed pages. i wish they would also pull in the comments associated with the likes/shares so i can see what my friends had said about the story.agree that there will be multiple share buttons, for the same reason there are multiple graphs. what i share on fb is very different from what i share on twitter… they are very different audiences.

      3. Mark Essel

        In a purely publisher sense, I would use the share button to have access to the largest network possible. But my friends in Facebook already see my posts via twitterfeed and Buzz and a fan page. I could move the fan page to my blog to enable social interaction on page (something I appreciate about comments & blogs). Distributed comments of high quality are a little frustrating as I miss all the value unless I’m omniscient (thank you Google Alerts).

    2. Mark Essel

      The convergence of Like is a thing of beauty but has different meanings on various networks. HackerNews up votes are more powerful in the first few minutes than Twitter retweets, unless the the retweeter has a million followers 😉

  12. RichardF

    I don’t trust Facebook with anything but the trivial. In my opinion they are not to be trusted with anything of mine that is meaningful. I just don’t like the way they act. I’ll allow what I want the world to see of me both offline and online, it’s not for Zuckerberg to decide that. I think it has the potential to be their downfall in the end.I want to be able to have full control over my online profile and relationships in a way that reflects my offline life (so that they can truly merge) , which means someone needs to give me controls and filters that are easy to use and customise.Facebooks privacy controls are a mess for a good reason, they don’t want you to have control over your privacy, it is incompatible with the way that they seek to maximise profit.

    1. Mark Essel

      Network privacy controls are a tricky thing when everything starts private. Opening up information to one third party means the web has access via that third party.Better to default a network at the get go to full public so there is no confusion. Private and web social networks only work with full anonymity. One leak breaks down privacy, the web was designed for easy document distribution, counter to private networks.

      1. RichardF

        That just doesn’t work for me Mark because I have (and want) different relationships with different people. If I have a non pc joke I want to share with my close friends I’m sure as hell not going to publish it in a fully open environment.

        1. Mark Essel

          I don’t see how you can closed publish that message though Richard. Even with email any one of your friends could reproduce that text and reshare it without thinking it may be embarassing to you.

          1. RichardF

            You are absolutely right Mark logically you cannot.Interestingly two of my wife’s colleagues (she is a teacher) have discovered the full horror that Facebook’s default “privacy” settings can cause. Let’s just say the students were most amused and the principal wasn’t.

  13. David Semeria

    It is possible to use one graph to do all the things Albert was talking about.Think of a file system. Many users, all with different permissions on your data.The only problem is that Facebook wants to be the root user.

    1. fredwilson

      yes, and they are building a permissioning system called “privacy controls”but i don’t think the average mainstream user wants to manage apermissioning system

      1. David Semeria

        That’s a very good point.Perhaps some startups should start working on that very problem – chmod for ‘normals’, as Chris Dixon calls them.It’s the only way to have a truly open and controlled social graph.

        1. David Ashwood

          Some of us are – but it’s only going to fly if:(1) it’s delivered naturally via an actual product – rather than floated as a concept ala 4Squared(2) it’s delivered in a way that both consumes and produces content to the channels – too many approaches are only about consumption (twitter consumed by Google Buzz)

      2. whitneymcn

        Exactly! When I reblogged Albert yesterday I noted that Facebook already has something like 36 different switches in their privacy settings (not counting the application-specific settings that often have sharing and privacy implications), and that system isn’t going to get any less complicated.While it’s possible that Facebook’s users will end up taking the path of least resistance and just start shaping their online behavior to fit the service’s friends/FOF/Networks model, the overhead of trying to figure out who will see what—and how to modify that when necessary—may be demoralizing enough to push a nontrivial amount of user activity elsewhere.

        1. ShanaC

          Already my friends are frustrated. I’m also at that awkward age where friendship is ill-defined and constantly changing- so those privacy issues are a big deal, because it is the difference between hanging out safely, experimenting or the mommy state. And yes, the donut has been created. The system is too complicated- and doesn’t serve specific needs nor does it link to life as it circles around me/us/whatever. VRM is a huge huge issue and an awkward issue. And yes, I have see people quit FB (though one person came back).Anybody’s guess though- it is a feature rich location, and it’s hard to say if use will end up shaping behavior as well

      3. awaldstein

        I’m fairly certain you are correct here Fred…in fact, most don’t even understand that there is one. Few folks even manage groups.You know that you are ‘opted-in’ for public sharing as the default.A Facebook friend call me yesterday telling me he liked my Pandora play list…I was auto opted in to share. Didn’t bother me but annoyed a bit as I really had no awareness of it prior.

        1. fredwilson

          we live in public, at least i dobut i don’t think that’s what people want to do unless they choose itexplicitlythat’s why a checkin is powerful

          1. awaldstein


          2. Mihai Badoiu

            I don’t fully. I have different profiles on different sites. (eg, fredmiranda, My name doesn’t even appear on my photography site.I think most people want some degree of privacy.

          3. andyidsinga

            Check-in is very powerful indeed. Not just in the physical world with location oriented apps but also in apps themselves through “sharing” features that allow users to connect with multiple social networking services. Its seems as through the social networking itself is becoming invisible as app features that are useful take the foreground. Apps that provide valuable features to users can easily and loosely couple with many social networks – think wordpress (and other blogs), ustream, disqus, even google docs.I like apps that have the ability to share individual “things” or events to social networks manually – *not* automatically.

          4. raycote

            Yes a balancing act between loosely coupled flexibility or tightly coupled convenience.Most web users do not really have a visceral sense of the privacy issues and so tightly coupled convenience wins hands down and if you are security aware it is still very enticing to just go with the convenience.

        2. RichardF

          The issue that Facebook need to look out for is that if the perception is that there are no privacy controls then people will change (self-moderate) their behaviour which may include not using it all.There are college kids who are changing their names in large numbers on Facebook because they don’t want potential employers looking them up

          1. awaldstein

            Agreed…somewhat, but I think the Facebook team is at their core, really smart and they understand this.But true, for a social open network, they are really bad communicators and not overtly open about what they do unless they get in reactive mode.I think the powerful data they provide for advertisers, the exporting of this aggregated data as a recommendation engine, the potential around fan pages…all of this can be maintained with some controls and they’ll get this.Net net…I agree, but I’m betting on them being smarter (or getting smarter) or third parties building filters. They are not going away anytime soon.

          2. Michael Jung

            I have such a friend.

        3. Jesse

          When you try and opt out (Account > Privacy Settings > Applications and Websites > Instant Personalization > Edit Settings), the small print reads:”Please keep in mind that if you opt out, your friends may still share public Facebook information about you to personalize their experience on these partner sites unless you block the application.”Who has the time/patience to block every single application your ‘friends’ start sharing your information with?!

          1. awaldstein

            You are correct…This is an issue for those that care and a significant blunder if the masses care about this.For a company so smart…and so right about so many things, this non-social way of rolling out stuff is just wrong.

      4. Elie Seidman

        I’ll caveat my next comment by saying that I don’t believe that this is a winner take all space. I’m long FB – especially for the off FB ad network that is coming very shortly – but believe others will build nice small niche businesses in social as well.That being said, I’m also not convinced that for the average mainstream user, maintaining privacy with multiple networks – each with its own set of friends, many of them repetitive – is better than using privacy controls in FB. Comparing multiple networks to the current state of privacy management in FB strikes me as simplistic. To me it’s not where FB is today, it’s where they will be in 12 months and I think that they will massively simplify their default friend setting such that you will have the ability to incredibly easily tag people as “insiders”, “casual acquaintance” and “outsiders” (or whatever other nomenclature). It’s possible now but its far too hard and it’s not a core part of FB. But I believe that their approach today is legacy – derivative from where FB was a few years ago but not appropriate for where it’s going. I would be very surprised if they don’t recognize that. I expect that within the next year they will push people to categorize their “friends” into very simple buckets with intuitive privacy controls. Hard to say if they can make that UI experience simple enough for mainstream users. But they have the advantage of having the user already; for the mainstream user, the cost of adding and maintaining a separate network is not zero either.

        1. ShanaC

          Other issues- we don’t have the language for this kind of friendship experience.

          1. David Ashwood

            Exactly Shana – it’s easy to forget that it takes time for the concepts to sink into the general understanding before they can be acted upon properly.Understanding evolves as it’s applied and people interact with the various solutions to problems.

          2. ShanaC

            I wish we would work around the idea of normalcy. Working around the ideaof a more general population and where to put yourself on a beach-head mightbe more useful….

        2. LukeG

          Good call on the default friend settings/tags, Elie. Didn’t see this until after I commented, but I’m with you.

        3. raycote

          Lets not level mix our GOAL and METHODS here!GOAL:Creating an effective, visually ergonomic, central dash board that enables average-end-user to reach out and easily orchestrate their DISTRIBUTED landscape of social relationshipsPOSSIBLE METHODS:1- centralize that DISTRIBUTED landscape of social relationships under one dominant corporate uber brand, that being whichever brand first reaches the critical mass required to arbitrage it’s network effect advertising potential to every other player. Then let that brand’s uber API define the semantic-web’s metadata lexicon to be just rich enough to reach down and optimize a commoditized advertising layer.2 – Leave that DISTRIBUTED landscape of social relationships in place and created a federated API for accessing a central dash-board metadata service, allowing for some real competition in the SOS(social operating system) user interface space. Is not Meebo kinda heading in that direction?Sending out an SOS – someone please call the internet police!-De Do Do Do, De Da Da DaAfter all, our relationships in real life are both DISTRIBUTED and REDUNDANT. DISTRIBUTED REDUNDANCY is a mandatory attribute of all natural living systems. It is nature’s statistical method of dealing with the inevitable errors and missteps arising out of complexity, nature’s way of not putting all it’s eggs in one basket.BUT… like a moth to a light bulb, we all know we’re go’in down this strange attractor, cul de sac of shot term, course of least resistance, point of no return commoditization FacePlant with FaceBook!As McLuhan points out, most real learning is done in the rear view mirror. Oh… and what’s his name, that other guy, that paraphrased McLuhan with “those that forget to look in the rear view mirror are doomed to keep run into the same things over and over again”.

          1. Philippe Honigman

            The “other guy” is George Santayana.

          2. raycote

            I was just being a smart ass (dumb ass) but you spun me off to wikipedia to discover another of his aphorisms.”Only the dead have seen the end of war”

          3. Mark Essel

            Wonderful analogy between corporate and biological evolution. It makes me think about the proteins that best capture photon energy which lead to vision. Several species independently found “site”, and that’s part of what we want with social web. A better understanding of what friends, colleagues, and influencers are up to in ways that are relevant to our current interests.

        4. brett1211

          Agree here. The company with the most complete view of the data set has the best chance of recognizing patterns and making sense of the pieces. And right now that’s FB.Also, didn’t the success of FB connect pretty much prove that people love one click check in? The only thing better would be zero-click check-in. Oh wait.

        5. fredwilson

          i think its much easier to join another social net where the rules are clearly understood than to manage your privacy settings and try to use FB for everything

          1. Elie Seidman

            Could be – I think there is social network overload for a lot of people. Once you’ve got LinkedIn for work and FB for social, the argument for a niche social network is, I believe, weaker. The cost of rebuilding a social net – with all the invites and invites – is wearing and annoying. So, so long as FB makes privacy remotely manageable, I’m skeptical that the mainstream will find a strong need for yet another network. This is something you think about far more and far more deeply than I do and like you said, you’ve put your money where your mouth is. Time will tell.

          2. Elie Seidman

            I may be changing my mind on this one. Not because Facebook can’t make privacy easier to control and manage (a five year old could design a simpler and better system than the one they have now). But rather because privacy gets in the way of Zuckerbergs larger plans and personal worldview.

      5. Satish Mummareddy

        What if facebook can build a symbiotic relationship with twitter, foursquare, yelp, where it gives something of value to each of those companies and takes the different permission systems to automate that process?

        1. fredwilson

          i think that’s what will happen

      6. Greg Gentschev

        Not to mention a permissioning system where the settings change of their own accord every so often.

      7. LukeG

        but there are probably shortcuts around this. If Facebook pre-built customizable privacy “themes” – e.g. Friends, Family, Work, Keep in Touch – they could potentially make normals comfortable very quickly. FB certainly has the data to look at how and with who its users share.

        1. raycote

          Yes – the Friends, Family, Work, Keep in Touch containment buckets would be very useful but they could create a faults sense of information control as sensitive tidbits will always leak across these boundaries via the digital equivalent of the good old grapevine.- I heard it through the grapevine not much longer will my secret be so confined.

      8. Aaron Klein

        Not only that, but Facebook rocks at almost everything EXCEPT handling privacy.

      9. Michael Jung

        During the press conf (on Scobles YouTube), they said that in ‘December about 50% of users reviewed their privacy settings. That they are very satisfied with that number as it is unprecedented on such a scale in the industry.’I am still rapping my head around all this, should I or should I not remove the information from my profile?

        1. fredwilson

          i reviewed my privacy settings but had no idea what to do after looking at them

        2. Mark Essel

          We’ve found Facebook’s true monetization system, legal privacy advice & consulting 😉

    2. RichardF

      yep agree completely David

    3. Laurent Boncenne

      so true, if done right. Facebook doesn’t seem to be taking that road tho =/

    4. ShanaC

      Too complicated- Although it is data we’re talking about here, I think we’re also talking about more than data, we’re talking about bits and pieces of ourselves.And that is the inherent difficulty- it’s not the social graph per say- it is that the web is flattening an at least 3d person (more dimensions if you consider time).- how do you let trust develop between people when it is so frustrating to be distilled down to flatness. I want my relationships to be different and to grow and change- if I am flat- how do I do that?

      1. David Semeria

        Good point about the graph changing over time.

        1. ShanaC

          Change is a natural state, I suppose I notice it more, because I’m young-yet it is extremely unnatural to assume that what we assume is a graph is agraph- it’s really a web, and an organic web at that.It makes trust difficult, since the “natural state” is always in flux. whatpermissions do you set when they could change tomorrow?

      2. raycote

        The Semantic Web, Syllogism, and WorldviewClay Shirky…Quote”Furthermore, when we see attempts to enforce semantics on human situations, it ends up debasing the semantics, rather then making the connection more informative. Social networking services like Friendster and LinkedIn assume that people will treat links to one another as external signals of deep association, so that the social mesh as represented by the software will be an accurate model of the real world. In fact, the concept of friend, or even the type and depth of connection required to say you know someone, is quite slippery, and as a result, links between people on Friendster have been drained of much of their intended meaning. Trying to express implicit and fuzzy relationships in ways that are explicit and sharp doesn’t clarify the meaning, it destroys it.”a good read!

        1. David Ashwood

          The problem isn’t so much the expression – it’s more with the representation. If it’s a fuzzy relationship – then represent it as something fuzzy :)Visualising complex data is still in its infancy – but there is some real progress being made in the field.Some good overviews and examples can be found and . Ben Fry has a great book on the topic and a great blog –

          1. raycote

            Great links thanks!Tools for transcoding abstract, organically complex, realities into simpler visual formats for the rest of us will probably be key to saving all our dumb ass from ourselves or am I being to optimistic?your links pushed me over the edgeinto opening a new bookmark folder to better focus my hunting and gathering of everything related toPERCEPTUAL ERGONOMICS

          2. David Ashwood

            Over time it should help – but it’s early days. There’s still a lot to learn about the information itself, what we want from it – and what makes sense to people using them.

          3. Mark Essel

            Appreciate those links David. I have been out of step with visualizing social data

        2. ShanaC

          I <3 Shirky. He’s a very logical guy. And in my reader.

  14. Andy Hunn

    Fred – Totally with you here – on a related note, as connection requests have become all encompassing (high school friends, work people, neighbors, close friends) I’m missing the direct connection being drawn that because i’m linked to someone, we share the same beliefs.We’re in the online ad targeting space and have developed a new way of targeting ads based on individuals’ attitudes, values and beliefs. Have met with lots of venture folks in past two months and have had much discussion about those using ‘social graphs’ to target people. I get the ‘if my friend buys life insurance, i’m more likely to buy life insurance’ logic for direct response, but just because I oppose offshore drilling doesn’t mean my friends do. Seems like a fundamental logic flaw is getting missed in the froth.

  15. dave

    Fred there is only one graph.The problem with what Facebook is proposing is the same problem the world Twitter is trying to build has. The company proposing to run the show likes the proposal but everyone else has a problem with it. Ultimately, imho, “everyone else” wins.

    1. fredwilson

      if you mean the web, then i agree with you davecompanies will come and go but the web lives on

      1. thewalrus

        i hope so, but i don’t take the open web for granted. i hope we continue to see a multipolar world, rather the emergence of a google/facebook cold war with everyone else being forced to join a sphere of influence.i think you nailed it when you said fb will “empower the creation of many social graphs for various activities and to be in the center of that activity”. but i don’t buy the argument around settings being too fiddly to prevent this from happening. the scale benefits of their platform are so great that it will allow them to figure that out over time.i agree with dave winer that the best way to push back is to rally the resources of a coalition of the willing. all the major players in the space are scared and have an incentive to offer an open alternative that would deflate all profits around identity and social graph. the question is; are they able to actually rally together and execute on that, and not trip over themselves in bureaucracy and land-grabbing. are they willing to embrace a truly disruptive open model (“users own their own data”, like for real) and empower others to make that happen. ultimately, it must be some startup with a compelling new user experience that will grow the next hyper-value niche, but the big players are needed to help improve the quality of the soil.i think if someone simply tries to build a better facebook moustrap they will fail (i.e. if goog acquires twitter to go nuclear, they will end up like the soviet union).

        1. kidmercury

          yes i agree. i think it is a union of app developers that is most likely to invest in platforms that create open standards.

        2. fredwilson

          i don’t think coalitions are the answernew disruptors areoften more open or totally open

          1. thewalrus

            agree, coalitions and open standards are not the whole answer, they only provide the soil. things like openID continue to have very limited success because there is no instant gratification for users in the form of a compelling use case (i.e. I spend the energy to signup and what benefit do i get right away?). ‘openness’ is not compelling enough on its disrupters with new use cases and more open models are the ones that will dent facebook. the same way facebook didn’t start out to become the ID platform of the web, they were a way for college students to share photos and keep in touch, and that was so compelling that they put themselves in the position they are comment was circular, but what i am saying is someone like twitter, or 4square or whoever now has an even greater opportunity to rally an ecosystem to help their cause. and that facebook has such a dominant position that i think the big players are both needed and primed for co-operation.

          2. David Ashwood

            Neither Twitter or somebody like foursquared are in the position to rally. They’re not bringing anything significant to the table around a compelling story for users. In their current incarnation – they’re just another dataset.

          3. thewalrus

            i hear your point, but beauty is in the eye of the beholder. can’t argue with user growth.every service starts out limited to begin with…..iterate, iterate, iterate….

      2. Michael Jung

        But there are consequences in the meantime whileB and Twitter turns you into an marketer or ad.RE: companies will come and go but the web lives onAnything missed out in the last weeks? Net Neutrality, Digital Economy Bill (UK)?

    2. Mark Essel

      I agree that’s there is only one graph, but like the human nervous system there are regions of richer connectivity. These clusters take on higher and more specialized functionality. Google has been the “brain” for a few years now, Facebook seeks this coveted position.I can’t argue with companies trying to tap into the wealth of collective human creativity with new levels of organization and filtering. But owning the web is beyond any single entities grasp. It is everyone’s and no ones.

      1. dave

        My gut says the winning strategy will be this — a company gets withinstriking distance of taking it all chooses, instead of betting the companyon the unlikely event that they’ll win, bets on the jungle. Sees itself moreas an investment banker than a hotbed of innovation.HIstory has shown that being a hotbed doesn’t scale. That eventually thesecompanies have to tap into the general talent pool and they end up achievingthe same level of mediocrity as the previous dominant one. I’ve seen ithappen to IBM, the minicomputer companies, IBM again, Microsoft, now it’sGoogle’s turn.Zuckerberg is riding a rocket, and no doubt he’s brilliant, but he won’tmake it. What he’ll do is wake up his competitors, and one or more of themwill get a little bit of their shit together, and will break out ofcomplacency. Who knows it might be Twitter. If you see Google give up theirreligion and start to say no to their programming gods, and start doingthings to win markets instead of glorify their genius, that’ll be all itwill take to put a big brick wall that Zuck won’t be able to get around.Microsoft and Oracle are still out there and either of them could create apool of hundreds of millions of dollars to finance an open alternative toTwitter or Facebook.At least Twitter has some ears out there. Fred and Bijan go to their boardmeetings. Hopefully at the next one they’ll consider hitting reset on thego-it-alone strategy they outlined at their conference. A lot of the adviceI gave Microsoft 15 years ago would play well for Twitter today.

        1. Mark Essel

          I haven’t seen the right combination yet. A founder or leadership that shares your perspective on the value of choice, and has the resources or marketing power to sell it to enough of us to push them over the top.Maybe that combination can’t happen and my beliefs are just silly dreams floating along an unrealized potential future. We don’t need to change to adopt Facebook’s social pill, we need to convince Mark that in the long term Facebook can do a lot more than make money. It can rapidly evolve the web by monetizing data touch, versus data containment.*update*Maybe Brett Taylor get’s it. The open graph concept has a very web-like feel to it, maybe I’m drinking the kool aid after watching the f8 keynote (finally found some time).

        2. RichardF

          awesome comment Dave

        3. kidmercury

          if there was a comment of the day badge in fredland, and if fredland citizens were allowed to vote on its distribution, i would vote for this comment.

          1. Matt A. Myers

            I’m going to make time today to read your expanded comment.I’ve been humbled by the knowledge and understanding many people who have commented to Fred’s post.Thank you every for your insights.

          2. Donna Brewington White

            Agree with Kid. Going to the blog now. Thanks!

    3. wd40

      Spot on – reminds me of an old Clay Shirky post on the semantic web……<quote>Many networked projects, including things like business-to-business markets and Web Services, have started with the unobjectionable hypothesis that communication would be easier if everyone described things the same way. From there, it is a short but fatal leap to conclude that a particular brand of unifying description will therefore be broadly and swiftly adopted (the “this will work because it would be good if it did” fallacy.)Any attempt at a global ontology is doomed to fail, because meta-data describes a worldview. The designers of the Soviet library’s cataloging system were making an assertion about the world when they made the first category of books “Works of the classical authors of Marxism-Leninism.” Melvyl Dewey was making an assertion about the world when he lumped all books about non-Christian religions into a single category, listed last among books about religion. It is not possible to neatly map these two systems onto one another, or onto other classification schemes — they describe different kinds of worlds.Because meta-data describes a worldview, incompatibility is an inevitable by-product of vigorous argument. It would be relatively easy, for example, to encode a description of genes in XML, but it would be impossible to get a universal standard for such a description, because biologists are still arguing about what a gene actually is. There are several competing standards for describing genetic information, and the semantic divergence is an artifact of a real conversation among biologists. You can’t get a standard til you have an agreement, and you can’t force an agreement to exist where none actually does.</quote>

    4. Dan Ostermayer

      Which ultimatly ends up making multiple graphs.

  16. Akshay Mishra

    Your points are correct about requiring different social networks for different purposes and not letting them merge.But, as far as Facebook’s intent is concerned, it’s all about collecting data about users and placing ads strategically – which is how Facebook would make its money. It’s not about the social graph at all! It’s about having one passport, one identity to rule them all and using that to track your activity across the web.(Don’t be conerned about privacy – Google scans our emails and places ads adjacent to them – we live with it)They are trying to get into the DNA of the web – the “partner” websites they have hooked up with for the “Like” button now “know” that you are here. It’s not about socializing, it’s about being the ultimate content portal. Even certain Wikipedia pages are now part of the Facebook architecture.Remember – The Open Directory project? People fretted back then – about not finding enough people to categorize the web. Well, here you are – you’ve got 200mil+ people out there now who would do that happily for you. FB Connect was a master-stroke – and so is this.I think it’s a big threat to Google.

  17. Sebastian Wain

    I think it’s obvious that we need to fight against social graph monopolizations. Sadly, in the early times of Internet nobody implemented a social graph based on a RFC 🙂 These are the kind of things that go against Internet neutrality too.Big players like Facebook, Twitter, Google (OpenSocial), etc face an innovation dilemma, if future social graphs technologies/adoption are more open and distributed they’ll loose big part of their assets.

    1. Mark Essel

      I’d argue the social graph was never their asset in the first place. The pen maker doesn’t own our collective writing.

      1. Sebastian Wain

        Since it’s not yet possible to move your social network from one place to another, there is a big barrier of entry and exit.When you use a pen maker you don’t depend of others to enjoy your writing experience.

      2. Michael Jung

        Do I own my data? My contacts? Can I export it and delete everything on NO! I get a headache thinking about this.

  18. tko

    good post & comments discussion– thanks

  19. CeB

    Facebook is largely a 2 dimensional graph: person x person. Tremendously powerful, but is like saying the world is flat. In the “real” world, the graph is multidimensional, e.g. person x entity x person, person x group x person. The opportunity is getting the entities (read enterprise) setup to leverage the graphs connected beyond marketing functions.Once you add another dimension, however, complexity explodes to take into account the legal, regulatory, etc. requirements of the entity. And because of that complexity, most widget-focused VC’s won’t go there.As an example in your world, Fred, the winner is a platform that allows you to invite people that are part of your business graph behind your firewall to privately communicate and share the proprietary, high value information (with the requisite compliance, legal and regulatory support).I seriously doubt anyone would ever attempt this with Facebook. Opportunity knocks.

  20. ash bhoopathy

    Each new decision regarding privacy and the move to become the “central social graph of the internet” has initiated me to remove more information from Facebook.When building web services, one thing I learned last year was to think about an equal exchange of value that you give to the user for everything that you take back. I certainly don’t perceive that the value has increased, while the potential risk of using my data in ways that’s perverse and unintended has increased. Just to check out the new like feature yesterday, I visited some shopping sites with the integrated social features. Putting my “user testing” hat on, I felt really weird about having people I’m connected to on Facebook show up on this page with their birthdays next to the product I was looking at. I understand the intent, but ultimately, I felt turned off by it. So, I did what I suspect many people will do, changed my privacy settings to the highest level I could.When I did that, I realized what you did, Fred. the average mainstream user *doesn’t* want to manage a permissioning system.

    1. fredwilson

      i am going to reblog that last part on fredwilson.vcit is so right

    2. Mark Essel

      Right on! I’d like to add web services need to give users more value than they take, and make up the difference with the enormous value of network effects and information.

    3. Laurent Boncenne

      That facebook like button is just a new way for people to become the new spammers of today. And a massive bunch of spammers….

  21. rajjr_tx

    I think you’re right on. For me, they’re forgetting the intimacy issue. I follow a lot of people I want to keep at arms length. I don’t want to include everyone in my inner circle, or even think about ‘managing’ the relationship. Interested to see how this plays out.

  22. Shyam Subramanyan

    Facebook is my personal (close friends, family) social graph and Twitter is my professional social graph where I meet new folks that have similar interest. I think this is true for most people.I think Facebook has been trying hard to tweak their privacy rules to encourage people to friend people beyond close friends and family (one graph to rule them all). I resist, but many folks don’t. It will be interesting to see if Facebook is able to make it happen, but in that process they might create an environment with too much noise pushing everyone back to email to communicate with the subset of folks they want to communicate.

  23. Michael Weiksner

    I agree that there will be multiple social graphs. However, what will make FB central is that has identity warranting. Any service that requires you to authenticate as a real person or works better with a set of your friends, will be better with FB. And that’s an aweful lot of stuff.It will be interesting to see where Foursquare, plancast, and others that have built significant core functionality on top of FB draw the line. I suspect that the competitive pressure to provide everything to FB will be too strong to resist over time.

    1. fredwilson

      i think we will see the danger in that play out in the coming months and years

  24. akharris

    It’s really interesting to me that, the more facebook grows, the more I see people locking down their profiles to all but an inner circle. When the service was founded, everyone was fine keeping pretty much everything open. That was a function of two things:1) It was just us Harvard kids, then a few select other schools using it2) It didn’t really have all that much information on itWhen they started putting up pictures, I remember the hubbub over various risque shots that showed up. That scared some people into tighter privacy. Then, when they opened the network to everyone, it scared more. Facebook used to be a village where you could trust everyone. Now, it’s a bit more like a glass megalopolis, and people are not, I don’t think, comfortable with that kind of exposure (I know it makes me uncomfortable).The thing is, for every person that locks themselves down, 1,000 more sign up with full access.So, that said, I don’t think that current notions of privacy (even as it extends to who you want seeing your 4sq checkins), those that were formed and hardened in an earlier time are going to survive. Whether or not facebook wins as the single social graph, or simply the hub of them all will depend on what the upcoming generation sees as the necessity for privacy. It’s unclear to me what will actually be treated as private vs. not. If you want to check out an intriguing look at that concept (albeit from a pre-facebook and very different perspective) check out “The Light of Other Days” by Arthur C. Clarke and Stephen Baxter. In it, the average person acquires technology that lets them see anything, anywhere, anywhen. Society has to shift in order to deal with this total abrogation of privacy. It’s an interesting read.

  25. bijan

    couldn’t agree more!last fall i wrote about the notion of a winner takes it all – this was & remains my feeling about a one graph scenario:”Facebook is the market leader for sure. But its not winner takes all because its just too big of a category for innovation & creativity. Right now my social network for music isn’t on facebook (instead it’s on hypemachine and tumblr). My photo social net is on tumblr and flickr. My social net for television is Boxee. My information social net is Twitter. FourSquare and Twitter are my social net for places & events. The list goes on.”

    1. fredwilson

      i remember that post Bijan. you were spot on then and still are

  26. robertverwaayen

    Fred,With all due respect, if this is your great insight and “a very important topic for those of us who invest in the social web” than you should know there is a whole body of academic work out there, which I recommend you and your readers give a go to reach a far deeper understanding about social networking.Particularly on the issue you adress of network structure and its affect on behavior as well as the question of network formation. This stuff has been around for almost a 100 years.Here’s an interesting overview piece:

    1. fredwilson

      thanks robert. i think i may have read that a few years back but i will give it another read.

  27. andyswan

    I think facebook is trying to REVEAL the social graph, not create it.Your example graphs all had one thing in common: you. Facebook’s strategy should be (and likely is) to own the ultimate internet profile.By revealing, enabling, and connecting all of the social graphs….they build the value of that asset and can eventually insert themselves into a massive number of transactions, which is where I believe the true dollar potential of the social web lies.

    1. awaldstein

      Thanks for this.Really well said!

    2. David Semeria

      …but once you manage all that social data, the temptation to take a peek at it will be overwhelming.There will be a big market for services which manage your graph on the understanding that they will never use or look inside it.

    3. whitneymcn

      I agree except that I think that Facebook is trying to get us to *recreate* within FB the graphs that we’ve already revealed/discovered via other services.It seems to me that Facebook has come to terms with the fact that the binary “you’re friends or you aren’t” model ends up missing some pretty important aspects and shadings of a relationship, and they’re going to make a good run at kicking out the companies that have occupied these odd little social corners.

      1. andyswan

        Good point

      2. ShanaC

        Except they haven’t. They’re trying. Hard.

      3. Guest

        i think facebook realizes that it needs to coexist with other services to add that color/personalization though. that means not only twitter, foursquare, tumblr, etc, but also the app economy built on top of it, like zynga, causes, and rockyou.the fact that it isn’t a good channel for the long-tail interests is precisely what makes it the ultimate network. the niche interests revealed/discovered via other services add color to our online identities, but the canvas of the painting has become, by default, facebook. it’s slowly becoming our digital “driver’s license”, and i believe its mobile growth is a big part in making that happen –

    4. andreaitis

      “Facebook’s strategy should be (and likely is) to own the ultimate internet profile.”i think this is right. aol used to hold onto consumers in part by owning their email address – people did not want to leave and give up that identifier, that core connector. facebook demolishes that concept in so many ways. i don’t need to remember or store anyone’s personal info anymore. i want to reach out to someone? i just go to facebook. they are the new phonebook, the new 411. if you’re not in facebook you’re a digital hermit.

      1. ShanaC

        I know a guy (or two) like that- and actually, I value trying to get in touch with that guy all that much more. It isn’t like he doesn’t understand Facebook-he tried to get me to read the likes of Cory Doctorow.He truly understands the following:1) How powerful face to face interaction is. He gets to be much more in control of his presence by extremely limiting his web presence. He’s basically almost impossible to track on the web. Unless you knew facts about him already, you couldn’t really be sure something on the web was him. He essentially is in control of his own creation.2)That much like what andy says- without him, no transaction, social, monetary, or other can take place. By removing himself from large chunks of the internet- he gets to choose what those transactions will be, rather than have other people choose for him, and he gets an almost darwinian edge.Now if only he knew how to program and perhaps had a mentor, he’d be set for life. He truly understands things I don’t about living in public.

    5. kidmercury

      revealing leads to controlling… is just like government. first they want to keep track of gun ownership, just for public safety, of course. next thing you know guns are banned.

  28. im2b_dl

    because it is not about a social graph… that is where everyone is missing the point. It is about a launch point…a dashboard that acts like a platform /foyer for each individual piece of content.And we need to start thinking in two different directions…the intake valve..and the out take valve. This is a place where the cognitive web/application will start to come in handy. Using the term social graph… …would you use that for your office, party, neighborhood?… Graphs map things…they do not “channel” or “carry”… imho

  29. philiphotchkiss

    I very much agree. No one anything will rule everything on the Web or anywhere else for that matter.The dynamics around these social graphs and their many intersections will be dynamic, self-organizing and amazing to both participate in and watch unfold.

  30. andreaitis

    we’re experiencing a pivotal shift. as a society, we have long been private by default; that’s now shifting to public by default. if google’s mission is to organize the world’s information, it seems facebook’s mission is to organize people’s information. in both of those scenarios, context matters. there’s tremendous opportunity in how to create that context, how to group and sort and filter the open web to create useful social graphs. how will we apply brian ascher’s concept of ‘right-time web’ vs. real-time web — and how do we do it so it’s easy for the mainstream, for the 90% who never touch default settings?as for the facebook like button, i can imagine one useful application here in the avc community (a social graph in and of itself). fred – you’d asked a while back about ideas for your redesign. i’d like to see what people in the avc community are reading, what they’ve ‘liked.’ i can imagine a module that shows headlines and number of likes from people here, to foster ideas and conversations within avc and on other sites as well. that would allow this social graph to become a group graph,moving across the web together. taking side trips to see and learn and engage but always returning home to share with avc.

    1. fredwilson

      i’ll see if we can use to do thati’m not putting a FB like button on AVC for the same reasons i don’t use an iPhoneit’s a political statement

  31. Mark

    I see a facebook browser on the horizon.

  32. Mark Essel

    10/10 on the LotR title.Opportunity comes in the strangest forms. Facebook is building the adaptive web that I covet ;). Fortunately for me they aren’t doing it in a web friendly format due to privacy concerns.Private communities have numerous options for selective growth. What I’m most interested is self selecting public networks, and smooth data flow into, between, and out of the networks. The code that supports and empowers focused communities will have a fully interoperable data flow model. It’s the depth of connectivity that allows us to identify with communities.Take for instance the thousands of comments we make here on AVC. The data network is connected to other blogs by Disqus, but as long as Disqus embraces open standards anyone can move data in or out of the network, in a web like way. Everyone and no one owns comments, just like social data.This goes well with my post today on open design. Click my name above if you’re curious, cut and paste on mobiles into Disqus is a futile effort.

    1. fredwilson

      i was inspired on the LOTR thing by tim o’reilly who uses it more broadly to talk about the web

  33. bsiscovick

    Regardless of whether or not Facebook is the only social graph or one of many, FB clearly has the one of, if not the, richest graphs.Overlying explicit *interest* data with the largest social graph is brilliant and incredibly powerful. A great comment earlier talked about equal exchange of value to incentivize participation – well I think the Like feature creates value for all participants:- Users get to share interests with friends. Users also get to see interests of friends driving them to more relevant/useful content (this is very Twitter-esq) – Content producers and ecommerce vendors get ridiculous viral distribution- FB gets a stream of the richest, most monetizable individualized interest data (short of Google’s immediate intent data)This is definitely not a winner-take-all scenario – many graphs exist and will flourish. But FB is making serious inroads in the all so important area of the *interest* graph (where Twitter is king), and that is a big freakin deal.

  34. Publiseek

    At f8 Facebook focused on the Open Graph and not the Social Graph. The Open Graph, as various comments suggest is a platform where anything can be represented as a node and connected to anything else using a labeled edge that represents the relationship between those two objects.So if you tend to use Google Docs to collaborate with your business partners, then Google can start with your Facebook profile node, and look at which edges labeled ‘business_partner’ link to other node that are also profile nodes. Then you can be presented a list of people, on Facebook or outside of Facebook that you are connected do on the graph that you can authorize to access your document.

  35. ErikSchwartz

    The thing that makes FB so scary powerful to me is that they have so much data they can dynamically create many graphs.”Show me Erik’s graph of people he went to school with who clicked on articles he shared about technology.””Show me Erik’s graph of current husbands of his ex girlfriends”This is why facebook is still cheap.

    1. kidmercury

      “show ads of baseball bats while erik is viewing profiles of current husbands of his ex-girlfriends”

      1. ErikSchwartz

        LOL. They’re “ex” for a reason… I’m glad they’re happy now.

      2. ShanaC

        Don’t even start kid…I’ve been shown the weirdest ads on that thing…

  36. ceonyc

    This is why I never liked FriendFeed. It stripped away the context of the app itself and the nuances of how I want to share.*However* there’s still a gap in my web experience–finding out whether I have professional acquaintances that like the same music that I do (LinkedIn/ mashup) or whether there are product managers that also frequent the Ace Hotel (LinkedIn/Foursquare mashup). How about being at a bar and finding the 6 other people who want to ask them to change the music as much as I do? (Foursquare/ There are so many interesting combinations that could be realized if each individual company didn’t have to crawl each individual site and normalize all the data individually.

    1. ShanaC

      Although Charlie, I love this comment- I also fear this comment. What happens if this comment came true? Part of the joy and sorrow of life is figuring out these commonalities in all sorts of complicated ways.If you were told all these facts about me before meeting me- wouldn’t I be boring? How would you learn to get to know me, to trust me, to work with me?* One of the larger vaguer issues we’re all struggling with here is how much is too much. In fact, I would much rather only know of only parts of your web presence- so that I could earn the right to know the rest, or have you tell me about the rest, and then earn the right to see the rest. it’s just how do you build a system that is simple to use and provides meaningful action for this sort of information.*Note, it doesn’t have to be me per say- just a someone. I’m using myself as an example.

      1. ceonyc

        If you are boring past a small commonality, that would be sad. 🙁

        1. ShanaC

          I never said I was or wasn’t. The problem is with more information outthere- you can create a lot of small commonalities and list them all. Itturns out it could be a huge list. And the paper list makes the person seemboring or perfect.Often what makes people interesting are their insights- which isn’t going toturn up on a list. I have friends who you wouldn’t expect me to be friendswith at all- it’s purely a friendship chemistry thing. We found ourcommonalities later.I have a certain fear that paper lists will make me “want to like you”rather than just accept the situation at hand.

    2. kidmercury

      yup. IMHO we need businesses that agree to standards (technical standards, standards on intellectual property, on ownership of user data, etc) in order to enable mashups. this also facilitates acquisitions by reducing the cost of integrating another company’s assets.

      1. ceonyc

        And as we’ve seen in the OpenID world… “open souce” open standards justdon’t seem to gather as much traction as when a big guy pushes it…unfortunately. If you want everyone to standardize, you need Zuck and acattleprod.

    3. fredwilson

      great point about context. that’s another big part of why the smaller nets are so powerful.

  37. Greekpunk

    IMHO, the larger the social graph, the less likely one will grant various permission levels to walled gardens of information. Some people have 1500 friends comprising many different(and overlapping) real world communities. Too much effort will be required to set up(and maintain…because online associates can often transition to real world contacts over time) these lists and will elicit the default response of simply ratcheting up privacy settings for all. The irony of it all is that those with the most Facebook friends are the largest influencers on the Net with the most to gain(and be gained from) from a well-oiled social graph.Fascinating debate.

  38. Jennifer Johnson

    Totally agree. The Twitter/tumblr main features are simple and public. Twitter has the added value of unique fresh content that is made to search. Facebook is complicated – eg. creating subgroupings at Facebook for private/public/work seems be a tad too much of a hassle for many users. That said, Facebook by size alone will reap a lot of value, particularly as a platform for games, networking, etc. Still privacy issues are limiting. For instance, adding Google Buzz to gmail was an invasion of privacy, as are Facebook moves between opt-in and opt-out privacy.The two “socials” will likely diverge, but I would bet on the bigger revenue markets ultimately comprising public information – checkins, updates, comments, marketing. Facebook still seems to have one homepage, albeit a changing one, which has echos of Yahoo, a model that got a lot of traffic, but added too much friction from making choices and navigation. I use Twitter equally if not more for search, which follows the primary Google business model. That’s one feature, without navigation options. Should be interesting to see how it all plays out.

    1. awaldstein

      I agree…And the size and breadth of the FB platform, makes it interesting a a jump off point for social business innovation. Fan Pages, Facebook’s ‘Main Street’ are a new frontier that sits on top of the growing user population. More people. More information. More interest to businesses.

    2. fredwilson

      i think FB really wants to become public by default

  39. Dave Pinsen

    I am trying to use a free steak dinner now to expand my site’s social graph. It’s hard to build a commenting community from scratch. I wonder how many other sites are attempting my Pavlovian method: red meat for intelligent comments.Facebook is the ADM of social media, I guess. My site is one of the little Midwestern farms Willie Nelson used to do those benefit concerts for.

  40. Terry Heaton

    As a consultant working with Media Companies (who are discovering the value of Facebook), let me add that Facebook has SERIOUS issues that must be overcome before the business community fully embraces it. At the top is the absurd notion that the name and email address of the person who starts a fan page remains with the page in perpetuity. You cannot imagine the problems this creates. How does one, for example, “purchase” the assets of a page started by someone else? What happens if a business is sold? What happens if the employee who started the page moves on? The problem here, for me, is that it is impossible for someone representing a smallish sort of business to get to anybody with any clout who can address these issues, and it remains my biggest gripe with Zuck’s company. If he seriously wishes to be the identity center of the Web, he’s got to get off his pedestal long enough to deal with matters such as the above. There. I feel better.

    1. fredwilson

      i’m happy to oblige your need to rant Terry. you should stop by here and do it more often!

    2. Mark Essel

      Terry, Facebook is putting the features of fan pages into widgets which businesses can place on their corporate sites. I think fan pages were a temporary inside solution, and social widgets are the way Facebook provides social connectivity to external sites in exchange for social data (Likes).

  41. Mihai Badoiu

    In absolute terms you are totally right. There are multiple graphs, and most likely it’d be very hard to have all of them managed from the same place.However, the average user is not Fred Wilson. The average user may know only 1 graph. The average user may only understand simple connections, and maybe only the idea of groups, which facebook has. The young user may start with schoolmates, and then he may expand to friends outside school, and before you know it, that’s the place where he goes for all social interactions. The average user may be lazy enough not to build any other graphs.In other words, the question should be about what the average user wants, and not power users. The first time I watched average users use Google Maps, I was shocked to discover that what comes natural for power-users may be very difficult for the mainstream — they may get some things, yet miss others which I would have considered more trivial concepts. Getting inside the psyche of the average user is not trivial and may require elaborate testing. You may be right on a case by case basis. I don’t think you can make general statements about lots of small networks becoming the status quo. Facebook may very well encompass most of the networks into one big blob. In any case, I think using facebook connect is a smart move for most small services.

    1. fredwilson

      the question about the average user is will they prefer to join a new network for a different use case or manage their permissioning on FB to do that?

      1. Tereza

        I don’t think it’ll work much if it’s a discrete, conscious user choice.The likely scenario is they’ll get lured in based into creating the custom network for the use case specific to the service they’re eager to play with.

      2. Firoze Lafeer

        I think it’s certainly easier to join new networks, and that’s much more inline with what we’re used to in the real world. I wrote something about this last week:…The problem with one network is bigger than just managing your privacy setting properly (hard enough as it is), the bigger issue is everyone else also has to manage their settings well or the whole thing becomes way too noisy.The interesting work, IMO, will be acknowledging that multiple networks will exist and then finding ways to convey or coalesce content across them at the right times.

  42. Turph

    All of this thrash is why I want my relationship control at the address book level. Managing settings across sites/services is a pain.

  43. Scott Asher

    Fred, you nailed it. On my post at, I pointed out that our FB persona is not the one we want representing us throughout the web — much like you say there are many social graphs out there, some for fun, some for work, some for interests, some for a mix of all, etc.However, I think the issue is that FB has made a grab for controlling ALL of it, and even if FB grants us very granular control over what sites use our FB data, most people don’t want to spend their time doing that type of granular permissioning.

  44. paramendra

    Wenger is making a key point. It explains why Buzz did not take off.

  45. Leonardo Aranda

    I couldn’t agree more.I am not interested in all of what all of my facebook friends do.I don’t want all of my facebook friends to see all of what I do.I don’t want to friend (on facebook) all of the people that I interact with in all other sites.

  46. zackmansfield

    A couple points that almost everyone agrees on: 1) facebook has incredible reach and is the default primary social network for a vast majority of internet users; 2) navigating the current facebook privacy settings jungle is unmitigated disaster at this point in time.I believe facebook *has* to make the play they are making precisely because they are the largest social network with the most penetration among users. Getting as much integration of external information into the facebook graph as possible (and thereby mapped/attached to the graph for monetization purposes) is the right play. Now whether or not Facebook is the “one graph to rule them all” is partially semantics and partially hyperbole – can any service ever truly “rule them all”? I think rationally we’d all say no. But facebook can and likely will become the majority owner of the graph – or in Fred’s nomenclature the “center” of all social graphs and driving the activity.But by being in the center Facebook will need to be vigilant about how it handles the privacy issue, which is perhaps one of the few things that could really hamper long term sustainable growth. There are some here in the comments who take a “facebook is evil because they make the privacy setting so hard to navigate angle”. I actually don’t think that’s the case – rather, it’s the result of iterative changes over time from the days when facebook was open only to college classmates. On a short term basis it may be beneficial to have privacy navigation complex as it allows for more data in and open; but long term value will be tied directly to the amount of engagement and data being shared in the service and will be driven largely by how comfortable people feel sharing this data. This comfort is aligned with the ability to easily manage permissions. I expect facebook to drastically change how users manage this in the future, something that looks altogether different than what we see today.For a final point, I think it’s important to remember that Facebook didn’t create the social graph – the service simply enables us to see the graph(s) that already exist. For many of us that means that facebook is an amalgamation of multiple subsets of our relationships – high school friends, college friends, neighbors, friends of friends, people we met on trips, etc. It’s the personal rolodex of everyone we know and have known, which is useful in some capacity for sure. But for high value usage, it makes sense to segment off our interests – which is why niche services can and will continue to exist. We already do this in our offline lives – we have our Christmas card list which is large and our dinner groups who we get together with one every few months and the closest friends we talk to on the phone weekly (and this doesn’t even count organizations like our kids’ sports teams or church/synagogue or PTA, etc). This is not a zero sum game where there’s one winner – our own offline relationships should tell us that.

    1. fredwilson

      great comment zack. do users want to explicity manage permissions?

      1. zackmansfield

        thanks – I don’t think users want to explicitly manage permissions – if by that you mean navigating the so-called jungle that is permissions/privacy control today. I think Fbook needs to think/engineer a privacy/permission infrastructure which is totally different and looks and feels easier to navigate. Nailing this will be key. As much as users hate explicitly managing permissions, they absolutely want to have the right/power to restrict what can and cannot be seen. A new nomenclature may be required – something beyond just friends – but it needs to be clean and simple.A separate macro-level movement in facebook’s favor is that people are today increasingly comfortable with sharing personal information with their online world. Remember, it wasn’t too long ago that most screennames and email addresses were crazy aliases like surfergirl67 to create a masked online presence instead of an actual identity. Facebook’s contribution to the comfort people have in assuming their *real* online identity cannot be overstated. As facebook becomes a de facto online identity across the web I think all web services are richer, and thus, our online world has greater overall utility.

  47. RichardF

    Integrity, trust, respect for the individual and the privacy of their data are what is required if any one company is going to provide the long term service of “one social graph” and that is not Facebook in it’s current format and nor will it ever be if they continue with their current strategies.I appreciate they are massive right now and they are going to make a ton of money but they are not the future they are just another step along the way.

  48. Mihai Badoiu

    Is it possible that there’s a more important graph than the social graph? A friend of mine claims the “local” graph is going to be the next big thing and that mobile is going to make it big very fast.

  49. howardlindzon

    early dayzzzz

    1. fredwilson

      it’s hard to remember that sometimes but i think you are right howard

  50. Danielito

    This privacy issue seems to be the biggest problem to solve on the Internet today like search was a few years ago.And I think that relying on users explicitely setting up privacy controls is a wrong approach as maintaining a people powered directory like Yahoo is wrong and then one day google came…But on the other side they’ve shown that they can be very quick at exposing features to third parties to solve their problems : what if they’d automatically maintain a list of my foursquare friends inside of Facebook, or expose something so that groups and list can be created and maintained outside of facebook…

    1. fredwilson

      right, as i said “If it is to empower the creation of many social graphs for various activities and to be in the center of that activity and driving it”

  51. ShanaC

    I’m thinking that one of the reasons we have multiple webs is that we have multiple kinds of trust. These trusts changes, grow, and fall apart over time (that’s natural, we change and grow and shift over time). I question largely groups like Facebook, and even some of the ethos put up here (though not all) about how we will manage trust. Although we display initial layers of trust easily, we tend to build up into deeper layers of trust. I guess the idea of having all of my “data” (I mean it is data, yet it is also more than that, it’s bits of displays of my personality, my psyche, my hopes and dreams) out there and having it flattened, packaged, and easily accessible to anyone takes away some of the more simple joys of getting to know someone.And while I love David’s hope for a chmod for “normals” (though I think everyone here is normal, if I did a survey about your hopes and dreams for the near future y’all would probably sound pretty normal), I think largely we need a change in our ethos around this data and what it means to live in public. Our unconfortableness about these issues is a sign that perhaps we want something more- the ability to relate to each others as humans beyond the super-mediation.I think people will pay for niche communities, and they will pay for the ability to hide some of the data. Just so that they have a chance to experience some freedom away from the data and the burden it can be to live in public. Sometimes, being an anonymous guy in the street is better, especially when you want to just get to know someone.

  52. AndreaF

    A few thoughts: 1 – I agree that there are multiple social graphs and the reality is at one end even more complex than what Albert describes but on the other end, the average web user probably spends 80% or more of his/her time on FB anyway so at least with regards to his/her web based social graph FB can get to a good approximation;2 – I think that it is fine to have multiple social graphs and that these are based by macro sectors; i.e. think travel/social vs financial vsmedical for example; each social graph is good enough (with good approximation) to provide me with relevant info within that sector; so, if I am looking for travel inspiration it’s simply important that my travel social graph be accurate and available on whoever is going to build that travel solution; the financial platform I use doesn’t benefit too much by knowing about my travels; or rather, it does but only to refine the details of my queries;3 – most apps and websites will gradually develop on the FB platform and although that is scary to think about, it may give FB even more power it has now and may even benefit the user – it is really scary though;

    1. fredwilson

      medical and financial are great exampleswill FB be the place we choose to network in those areas?i think not

  53. jeremystein

    i think theyre trying to monetize a lucrative protocol.its possible that twitter might displace sms at some point. but twitter can never replace the native social graph on my phone– my phonebook. facebook isnt trying to create one graph to rule them all. they want to monetize the infrastructure they have in place. they are building the sms of the internet.

    1. fredwilson

      what’s cool about android phones is they recognize your social nets and link them to you contacts in your phone (which are in google contacts in the cloud)i’m very bullish on android because google understands the cloud in a way that apple has not shown it does

  54. Ric

    This is of the mindset of not doing one thing well. FB needs to step back, rethink and determine HOW far they can stray from core competency and still serve its user base. I like to think of an analogy in the words of Jeff Goldblum from Jurassic Park: The inventors get so busy trying figure out How to do a certain thing without ever considering whether they SHOULD do a thing. FB is squarely in this zone.

  55. michaelmegaw

    This seems like one of the greatest opportunities over the next 10 to 15 years. While facebook has capitalized on the “social graph” there is a reason that twitter has become so successful despite much of the same functionality exists in the facebook environment. There needs to be multiple outlets in order to express yourself on multiple levels.

  56. Ivan Kirigin

    Fred, you have ~100 facebook friends and about 300 foursquare friends. I’m having trouble understanding your comments in that context. Are your facebook settings more public?

    1. fredwilson

      mine are not. i cut back facebook to only people i would invite to a family wedding or bar/bat mitzvah.but most people i know have 1000 or more facebook friends

      1. Tereza

        There’s a really important point in here.I think an activity, in a vacuum, in which you sort and inventory all your friends/people/contacts is a pretty tedious thing.Some people are disciplined enough to do it, but most are not. (I am not).And early in life, the categories are less clear than they become over time, when life has you natively doing things where you have to do the slicing and dicing.You’re throwing a wedding, a bar/bat mitzvah, sending out holiday cards (2 sets — home and business), a 40th birthday girls’ weekend, lists of thank you cards you need to write. These are native life events where you can’t help but choose people.Once you create any such list, that is personal intelligence and that would be a great resource of what Facebook or other nets should draw off of.In our family, we started with our wedding list, it turned into our Christmas card list and has split off in a few lists. Each year we review, prune, delete, add, and get the contact info perfect. It’s an Excel file with 10+ years of our joint social data, and it is the cleanest data set we have. And it is a really long list.But it’s totally disaggregated from any of our Facebook, Linked In, Address Book or Google Contacts lists. And that is silly. But at the same time, I don’t have the time and inclination to set aside time to get it all synced up.Are we outliers? It just feels rude for a soc net to not try harder to leverage the activities we’re already doing, when and how we’re doing them.

  57. terrycojones

    Hi FredThere’s an infinite number of graphs, you’re right. E.g., the cycling graph (you’re connected to anyone you’ve ever gone cycling with), the pizza graph (anyone you’ve shared a pizza with), the handshake graph, the Erdos number graph, etc. To the extent these are about people, they’re all simultaneously overlaid on the set of all people and co-exist simultaneously. Facebook have just made a play for a simple and prominent (informally speaking) graph, based just on like.I agree people don’t want to manage permissions systems. But you need permissions somewhere in the setup if you’re to manage these graphs computationally (in a shared environment). The perms can be hidden by an app, can have good defaults, can be easy to understand – but they need to be there.And…. (you knew it was coming, right?) FluidDB can do all this. I.e., it allows the creation of an infinite number of graphs, you can can search across them (because the graph info is on the same objects – where it naturally belongs, where it’s most valuable), it has a dead-simple perms system, no-one has to anticipate what graphs someone or some app might want to build, you don’t have to ask permission to create a graph.This all falls out of FluidDB quite simply, due to 1) objects not having owners, and 2) there’s an object for everyone, and 3) you can build graphs just by tagging objects (that doesn’t lead to being able to do sophisticated & fast graph operations, but it is extremely flexible).Apologies for posting yet another FluidDB-can-do-that-too comment. I hope you don’t get sick of them. I have another one for you too, re Twitter annotations 🙂 In lieu of that though, see…Terry

    1. ShanaC

      I have to ask this Terry- what kills a tag? Like how do I prevent someone from tagging me with cursewords per say?

      1. terrycojones

        Hi ShanaYou can’t. In many ways FluidDB reflects how we work with information in the real world. You can’t stop me from having an opinion about you – possible private, possibly shared with friends, or written on my blog, etc. Of course you can take measures against such things, if you’re aware of them, but these may not be successful. A FluidDB object is like a concept in some ways – because it doesn’t have an owner, no-one needs permission or to have their needs anticipated. If I want to think “New York” is a sunny friendly place, I’m perfectly free to do so. Someone else can think the exact opposite, and they’re free to too. By making things by-default writable, you let people (and apps) put information where it’s most useful. But there’s no architectural attempt to ensure accuracy, consistency, freedom from cursing or slander, etc. That creates the potential for a mess (and for evolution), which to me is a positive thing (consider e.g., the web or wikipedia as things with a huge potential for mess & evolution), though many computer scientists are outright alarmed by the idea of a database with those properties.

      2. terrycojones

        Another quick comment – FluidDB has a tag for each tag and for each user. So it’s possible for you to tag somone else as unreliable, a liar, etc. It’s possible for you to tag their tags: you could indicate that you trust & use the fred/startup-opinion tag but that you don’t really like the fred/music-rating tag. Because of this, FluidDB itself, and its users, can show others what sorts of things in the system are in heavy use, what’s trusted, what your friends know & trust, etc. Evolution of reputation and trust is also important in a social data architecture. By providing a tag for every tag and user, FluidDB gives a locus for that kind of reputation/trust information to accumulate & be found. It will take a little while until we reach a point where that sort of thing enters the mind of the average user, but we’ll get there.

        1. ShanaC

          Life is about to get extremely complicated. How do you redeem yourself insuch a world? What labels are trustworthy? If it works, how do you controlthe deluge of data?

          1. terrycojones

            Life is already complicated. How do you redeem yourself in real life? How do you decide what to trust in real life?Sorry to reply to your questions with questions, but I’m not really trying to provide answers with FluidDB. I’m trying to provide an information architecture that allows us to work with information in a more natural way. Along with that come the various problems (and advantages!) of the real world. Likely they’ll have to be solved in the kinds of ways we solve them in the real world: by communication, by friends networks, by being discriminating, by the legal system. Semantics has no place in a data architecture, IMO: the architecture should be neutral with respect to content, including not distinguishing between data and metadata, but that’s another subject.What may seem more constructive (I was being constructive above, though it may not feel like it): in response to the deluge question, I am pretty sure an important part of the answer is “more information” and that the more information should be able to be put in the same place as the original information. So it’s not “more information elsewhere” – which contributes to the deluge and makes matters more complex (other APIs, other apps, other permissions, etc) – it’s more information in the same place (same API, same perms system, same query language, ability to filter/search/personalize as you wish). All of that points to the need for a shared *writable* underlying architecture.And I’ll leave you with a quote from David Weinberger’s wonderful book Everything is Miscellaneous: “The answer to too much information is more information.” I’d add: “in the right place”, where right = suitably actionable.I hope that makes sense! Thanks as ever for your time & energy in thinking about FluidDB.

          2. ShanaC

            No it makes sense- I’m just not sure which philosophical point I agree withmore1) a centralized system2) a decentralized systemI also recognize that there is a large linguistic issue over semantics thatspreads into how APIs are constructed. It’s one of the reasons they are amess right now- no one can agree what they should say and what they shouldreveal.So I understand the point- I just wonder, that with a centralized system, doI lose the possibility of forgetting, which is one of the ways humans work-I mean, do I really want people to go to the same website I went to when Iwas 12, which I don’t know if it exists anymore?And while I agree on suitably actionable- how do I let people vote withtheir feet on it.I do think for the vast majority of what you are saying though is verytruthful. This is just something to ponder.:)

    2. fredwilson

      keem ’em coming terry

  58. cgarb

    The issue here isn’t whether people will have one social graph, people technically only have one graph of connections to other people. The issues are:1. Will people want to digitally store all of their relational connections in one place?2. Will people want to interact with all of these digital connections in one place? And can one website amply provide a diverse enough experience to fulfill the different types of interactions people want to have with one another online?Because Social Networking Sites (SNS) are not just digital rolodexes and provide an environment and content for people to interact in and around, we currently see hesitancy to store all of one’s contacts in one place. No SNS will ever hold all of one’s graph – the only way this would be possible is if different large scale SNS completely opened their APIs in a way they could easily match connections one has on another network, which has already created the need for user controls since people don’t want this.Where I disagree with some of the comments here is the argument that people want different graphs for different activities, which isn’t exactly true. If a user signs up for Pinyadda, Blippy, or Foursquare, they don’t necessarily care about creating a “new graph” – they care about making the right connections for the environment and interactions that happen on those sites, which certainly isn’t all of their Facebook, Linkedin or Twitter connections, but definitely some of them. What this points to is a large potential for new sites to succeed by providing a great experience around a specific environment/interactions that are not core to existing SNS, and leveraging user’s existing graphs in a way they can easily choose people they want to interact with in this new environment from the existing connections on other SNS. A great example of this is a SNS grows to critical mass, it becomes increasingly more difficult to redesign or iterate the existing graph structure. Facebook tried to iterate on their reciprocal edge structure unsuccessfully. They have also tried to create ways to assign more extensive values to relation types within their graph, such as lists, and remember way back in the day when they would ask you what type of connection you had to a “friend” – i.e. took a class together, hooked-up, etc? This is incredibly valuable data, but it is requires too much work on the part of the user and will be hard for them to change behavior in ways to get said data.Regarding overall graph designs, we are simply at the tip of the iceberg. In the next 10, 20, 30 years we will begin to see more complex graph designs evolve as different SNS emerge by successfully leveraging people’s existing graphs and build a differentiated and gratifying experience in seperate environments from other SNS.We have already began seeing SNS emerge with directional relationships (Twitter), bi-modal network sets (meaning two sets of actors, i.e. People and Sites, like on Pinyadda). We will start seeing more advanced graph design around multivariate structures, meaning more diverse sets of relationship types between actors, and weighted relationships, meaning systems will be able to assign values to different connections – i.e. levels of “friendship”, or more realistically, of the people I follow for say “Red Sox” content, who I listen to the most.SNS sites hosting interactions will also begin to evolve in my mind similarly to countries, with governance becoming a more prominent issue and clustering based on activity/beliefs.Either way, its an awesome time to be in this industry, lots and lots of potential. Digital SNS are really only about a decade old, if that.

    1. Matt A. Myers


    2. fredwilson

      as charlie said in an earlier comment, context matters a lot

  59. Matt A. Myers

    I watched parts of F8.I see what they’re doing and it’s brilliant, however, from a business perspective it doesn’t make me want to jump right in.As I commented in another post of thread regarding Twitter why I wouldn’t currently want to integrate my business fully with Twitter.The same with Facebook comes to mind.Why would I want to spend my marketing dollars to make Facebook new users?And this is clearly what they’re doing: To gain new previously unreachable individuals (who are in niche social graphs) who won’t find Facebook specifically useful, and then solidify their usefulness by being linked directly into the niche graphs. Their primary objective is gaining the new user, and they will get a flow of the new users from the niche social graphs they are connected to will generate. This will help build and solidify the usefulness and quality of Facebook “and network”.But is Facebook needed? I’m going to make high-quality sites, that will be profitable, and they can grow a userbase on their own. Always have to factor in competition too: Anyone who’s smart enough to understand these dynamics will realize they can do it on their own, without feeding Facebook or Twitter their userbase from their growing traffic.There are positive examples though. Gaps that could be filled. And Facebook’s social graph had a big gap to fill — high-quality addictive games. Zynga was the first to go after that, and did it well. There are others now entering the the high-quality addictive games market on Facebook and seem to be having the same rapid growth of users; First to market was probably important there, but that’s something I’m watching for. Was it important because it allowed to grab the biggest market share for the cheapest? Is it’s fulfilling a void / need of users that can’t be filled by a competitor? The viralty that exists is the linking between friends.. and I’m not sure how many gaps there are to fill.There are pros and cons of linking into Twitter and Facebook … I need to put more thought into it before I’d decide if I’d integrate with Facebook with my plans. I have some ideas of what might happen..But first I need to keep working on my presentation, mockups, and layout of my plans to continue to finding bigger and bigger chunks of money. 🙂 I’ll be ready soon though.

  60. Zvi

    The problem with “One Social Graph to Rule Them All” is the same as w Semantic Web – who define taxonomies? Folksonomy/tagging doesn’t work

  61. Jan Schultink

    I think it will be practically impossible to manage fragmented social graphs, you end up with 3:1) Assymetrical following: Twitter, your blog2) Symmetrical friends, people you allow in your facebook friend list3) Tiny application-specific networks (your stock trades as an example)It’s just too difficult to put people in buckets all the time:- I loosen my privacy standards a bit, allowing “strangers” in- I cut back on the things I post on facebook

    1. fredwilson

      application specific networks are going to be big

  62. Guest

    Completely agree.

  63. Geoffrey Vitt

    Could not agree more with this post. One thing that wasn’t noted in the post (I’m sure it was in the comments…) is email. For me, the customized/extremely direct communication from email, is often times the most comfortable way to communicate and create my own social graph.All that said, the pervasiveness of this service and related data could make their ads and ad networks an extremely compelling service. I believe Venturebeat had an awesome article on this

  64. johnfurst

    I agree, one graph won’t do it.Why do I use more than one social network, because of the different crowds I meet there AND because of the different tools that support specific interests/tasks/goals, whatever. A couple of days ago someone sent me a request for “liking” his facebook page and in the same sentence that person wrote, “You can hide me if you want.” Ridiculous isn’t it.Personally I will tighten my networks. No interest to participate in the race for ever increasing numbers.By the way: I liked the Hunch questions.

  65. defrag_Ami

    The f8 announcements and discussion got me thinking the last few days about sort of the flip side – what’s the future going to look like for our private communication? Surely it is not just email or DM but will incorporate enhanced functionality from open social apps. We might want controls to discourage sharing or forwarding of certain communication, for example. Or, a universal cone of silence mode that disconnects us from all lifestreaming tracking and sharing. Anyone else thinking about this?

    1. RichardF

      yep 😉

    2. fredwilson

      i think about the cone of silence all the time

  66. timcohn

    Without further “personalization” Facebook’s annual revenue per user – $1 Billion / 500 million users = $2.00 a year – will continue to languish @ just $0.005 a day from each user.

  67. Devin Chasanoff

    I agree with Albert- while I may be saying the same thing in a different way… all of these websites that we use are tools, or extensions of ourselves, and I like to think that while they are on the internet, all of these social networking sites are meant to mimic and enhance our natural human behavior.With that being said, I am a part of many different cliques and groups. I have my best friends from college, high school, etc. I have a group of people who I like to get together with regularly to play basketball. And, I obviously have a strictly business relationship with a ton of people. While this is a very simplified version of my social life, I think you get the point.The problem I’ve always had with Facebook is that despite its ability to connect me with such a far reaching, diverse group of people, I don’t necessarily want everything I do broadcast to the world. For example, when I comment on the most recent episode of Lost on Facebook, I don’t need my friends calling me a nerd for doing so… but I love taking part in that conversation.The point I’m trying to make is that we have many different aspects of ourselves that we share with different people. That is why being a part of multiple social networking sites, like Albert said, is useful.However, everyone has that one voice inside their head that belongs only to them. And, this is where I think Albert came up just a bit short in his statement. I love Facebook’s newsfeed- my homepage- because it acts to mimic what I comprehend as me. Facebook is on track to provide the most personalized and useful internet experience we have ever seen in the history of the internet (if it hasn’t already) due to this Open Graph.For that reason, I’d love to go around the entire internet clicking, “Like, Like, Like” on every single piece of information I find interesting. However, like I said before, I do NOT want that information broadcast to everyone I know.The solution, which I doubt Facebook- despite having the means to do so- or another up and coming company will do is to have an incoming and outgoing social networking and internet hub for each individual. In order to get the customized Newsfeed possible, you can click “Like” for anything you want, but choose which social networks and groups you broadcast each piece of information out to. It’s like how Pandora operates in that I create a personalized hub for music, except it is for information and has a broader means to communicate that information back out to the community.

  68. LIAD

    online norms mirror offline norms.the people I play poker with aren’t the same people my wife and I invite round for dinner parties and they aren’t the same people I have business meetings with.Whilst Facebook may own the largest social graph, they do not and cannot have the definitive one as it’s just too fractured and diverse.When it comes to social graphs I’m not sure that absolute size matters. The riches are in the niches

    1. Matt A. Myers

      Facebook just wants to know what niches you like … so they can make money.. so being tied into other niche social graphs lets them know just that. 🙂

    2. fredwilson

      why am in not surprised you are a poker player LIAD 🙂

  69. Jason Keramidas

    The question really is what to do with the social graph, to what end does it satisfy a business need? If you just want to create a graph to show that people have relationships well, I guess that’s interesting, but there is a lot of valuable information there (trends, demographics, etc.) that a whole lot of companies might find useful. If you want to create an overarching social graph, it has to be useful to the end user to the extent that they can a) control permissions to reveal what networks they participate in or not and b) make it interactive rather than static so that you can actually navigate the network rather than stare at it. I wrote a little post of my own ( a while back looking at one company that offers some technology that can do just that (I have no stake in it) and if Facebook is forward thinking enough to really integrate interactive data visualization tools, then they really might have something

  70. Joe Siewert

    Wow, 222 comments as of my writing this. What’s the record for most comments on a post at AVC? So much great conversation here.

    1. fredwilson

      we’ve gotten to 400 beforethankfully not every dayit’s a lot to engage and stay engaged in a thread this big

      1. Joe Siewert

        Yes, rather daunting to jump into that many comments, but a good problem tohave.

  71. sigmaalgebra

    Broadly Albert is correct, but he is touching only a little of the crucial issues.WHAT a STENCH. Hold my nose. Choke. Gag. Lost it: UPchuck. And that’s a sheared, 100% wool carpet.Sure, E. Fromm explained that people are highly motivated to get security from membership in groups, in particular, to make ‘friends’. So, to provide a computerized way to connect with ‘friends’, Facebook serves a lot of Web pages per day to a lot of people. And now Facebook has some software to permit third party app developers to access and process some of the Facebook data.”Come, free data! Get your free data!”.This data is a ‘social graph’. It’s coming again, choke, …. Got to the kitchen sink just in time. Whew, that was close.Four points:(1) Do we have a description of this graph? That is, a ‘graph’ consists of arcs and nodes, and each arc starts and stops at a node. So, in this case, what are the nodes and the arcs?(2) How are we going to process the data in the graph? A LOT is known about how to process graph data.(3) What valuable results do we want from processing the data in the graph?(4) For the results, why start with a graph at all? That is, in the sense of, say, ‘data modeling’ as in relational data base schemata or entity- attribute- relationships (EAR), why a graph?For the graph and the results, in the ‘social web’, desired from processing the data, I can guess. My next guess is that a ‘social graph’ is not nearly the best ‘data model’ as a source of data for getting high quality on such results.In particular, not nearly all the data available via the ‘social web’ is a graph.For one data model to rule them all, hmm …. Well, I wouldn’t pick a social graph.For the processing, some advice: Eschew intuitive guesses, heuristics, ‘machine learning’, neural networks, and artificial intelligence. Instead, pursue processing that has a solid rational foundation so that you know in advance that you are doing something powerful that has high promise of giving you something valuable for the results you have in mind.Or data is the raw material for information like beach sand is the raw material for Intel’s chips, and in both cases the processing is important. Also, Intel, Boeing, Cisco, NASA, etc. very much want to know well in advance that the processing has a solid rational foundation; just toss the data into a bucket and stir briskly clockwise will not be taken seriously.Finally, if I get a Board and at a Board meeting get a suggestion that we should change our direction to do things with ‘the social graph’, then I hope that the table and floor are washable. Not a happy thought.

    1. fredwilson

      very colorful imagery

  72. Eric Leebow

    My belief is that one graph is not accurate, or a true assessment of one’s social activity. A CEO of a popular Internet dating site recently said at a conference that he believed there will be one online dating site that rules them all, and it’s my belief that this cannot hold true, as there are plenty of ways people can be matched in online dating. This holds true in social networking as well. There are multiple ways to think about this, and it’s not going to ever happen due to free market society.People have a choice, and the reality is that some sites are doing things that distance the human networking in real life, whereas others increase the networking in real life (ie. FourSquare,, and others are exclusively for online communications. Because of the increased options, there is no correct way, and this leads to inconsistency in a social graph, as soon as someone joins a new service that is not part of a service that is related to the previous graph. The person who follows you or subscribes to your updates on one network, may not do the same on the other (ie. your blog followers are different from your phone contacts), nor do you care to be associated with those who might follow you on a social network (ie. Some guy follows you from the other end of the world, you might not meet).As we learned early in grade school, there are multiple ways to solve a math problem, and likewise this works with networking and social networking problems as well. The challenge is that not all sites that became social networking sites were optimized to be social networking sites for every person’s networking style (ie. If you are reading blogs, then Tumblr might be the best network for you, yet if you’re going out on the town, then FourSquare would better fit your behavior), and hence this leads to no perfect solution. Likewise, who is to say that one site does it all right or even has the capability to do everything right?Just because one site may be considered the largest social graph, does not mean it’s the most reflective social graph. You must ask yourself if this is an accurate social graph or a true assessment of human interests on the Web and the real world. If you can say that this is not an accurate assessment, then you cannot conclude that this there is an ultimate network that will rule them all. However, what you can conclude is that the Web itself is just that, a virtual environment filled with layered networks.The reason for this challenge to conclude that one graph would rule them all, is that some networks belongs to network optimization and there is no cohesive standard. Although the sites once may have thought they may have been increasing the greater networking good in some ways, they actually may have caused a greater people and communication challenge, which created a barrier for networking. Hence, there are different ways to do things, and not all of them are perfect. At the end of the day, all of these services are to be used as tools, and the real reflection of a social graph might be the combination of mobile, your email, your social networks, and blogs.

  73. Suzanne Lainson

    Here’s my Twitter version of how I view this:The more Facebook tries to tell me every like of every person I am connected with, the fewer people I will have time to be connected to.A lot of the “likes” many of us have aren’t IMPORTANT likes. So Facebook, let’s not read too much into any of them.

  74. William Mougayar

    The Social Graph is like DNA. Nobody will own my Social Graph except me. Slicing, dicing, tagging, filtering that graph is becoming important and not an easy task. We need better open data interchanges that allow us as individuals to paste together an ongoing view of our Social Graph. Even better, we shld be able to put it to use for searching, discovering both content and people. There’s a new feature that Google has been selectively testing; when you search, it shows you results related to your Buzz friends. And when I tried it, these results were more relevant than the general-purpose search. So it’s like steering Google around your Social Graph- that’s powerful!

    1. Tereza

      +1I like that idea of network as personal DNA. If someone cracks the code on that, then maybe even the Wall Streeters will participate in soc nets. Most of the ones I know from B-school that are quite senior now are incredibly underinvolved in social networks and view it as just a fad. None of them allow FB at work so net result is they’re residing in a different world.Maybe it’s a protective mechanism to keep them from facing the music of how deeply people are pissed off at them. LOL. Alternative universe.Last week I was talking to a friend who has a large LinkedIn network. We were trying to help each other out with introductions but were crushed under our networks.We were bemoaning that LinkedIn is not very good on tools to help you sort/curate people who are already in your network if your network is large. And frankly I consider my professional network to be “medium”, not that large. But it was nonetheless brutally painful — 2 hours of scrolling through profiles.Where are developers to create services on top of them? Or are they not open API. What a huge miss.

  75. David Fishman

    I believe social can thrive inside retail. I do not believe retail can thrive inside social. Facebook will NOT own this graph. To quote @jasonbillingsley Selling into social is like going to a Cowboys game & trying to sell Eagles jerseys. You’ll sell a few, but mostly ppl will just hate you.

  76. Dan Ostermayer

    It’s becoming increasingly obvious the information shared on the social web is a large part for the benefit of the company owning the network and not the friends of the person sharing. The common folk of facebook do not realize this fact.

    1. David Ashwood

      That’s the current model Dan – generally companies are rewarded (by investors or stockholders) their amount of ownership. It doesn’t mean it’s the only model however – and we’re likely to see the ownership model dissolve with time as people desire more open sharing of the characteristics of their network. If the current companies don’t evolve & adapt – then somebody will take their place.

  77. Nathan Leehman

    I am still very interested that there is no differentiation here re: user type by generational bias. Specifically, can there be only a single graph with so many views/biases included in the overall user group? What I like, you may not, and what freaks me out (or Fred, or Alan, or Mark, or…) may not be the same thing that gets other people’s knickers in a twist. I am absolutely convinced that this is a significant divide between net natives, technically savvy users, the technically literate, and the technophobic (as posted in response to an earlier post on avc and on my blog), but would suggest that it becomes even more nuanced when hoping for truly widespread adoption.Remember when Google introduced gmail and wanted to search your email? It didn’t even occur to them that storing 1G of your personal email on their servers and indexing every word for targeted advertising would scare people, but by the time it was rolled out there were a number of options for ad-free use. Was this huge oversight in the original announcement just a function of S&L’s myopia, or was it based on the fact that due to their complete submersion in all things web that it didn’t even occur to them that some people would have a big issue with their personal information being scanned? My supposition is that as we go more and more native, there will be a large sub-group that simply stops caring about how this information is collected, stored or used, as long as they get the results that they are looking for: ie. a value that is worth the price of admission (which they may or may not fully understand).Last thought: most of us are comfortable with the cloud and its benefits/implications at this point, but take a minute and try to ask your parents about it. Or the police officer who pulls you over for talking on your cell without a hands free device (no disrespect meant to your parents or the local constable). You will probably find that the idea of information “floating around” without a physical lock and key puts them off quite a bit, no matter how you describe the safeguards.So one graph? Only if you are OK with losing the outliers and only if those outliers are not numerous enough to throw off the curve. The latter implies widespread general acceptance, and I’m not confident that we are there yet: I continue to segment my web persona pretty significantly. Fred notes that he does as well, and my guess is that this is a widespread occurrence for folks who continue to adapt to the growing digital world. Net natives may be ready for this, but it will take time before the world is ready to hitch their wagon to a single experience. Until then, mass interaction and unlimited user customization will remain a stretch goal for those interested in total market dominance.

    1. fredwilson

      my daughters are totally obsessed with facebook as any 17-19 year old wouldbebut they have their own blogs and have experimented with twitter andfoursquare tooi think that “segmenting my web persona” as you so aptly put it will becomemore mainstream soon

  78. Keith B. Nowak

    This is sort of where Meebo sits with the Meebo Bar –…. Facebook has released the version of a social toolbar but the main drawback is that it assumes a single social net for all of your sharing while, like you point you, this is not what people do. Different stuff is shared on different social graphs.

  79. Michael Wu PhD

    I agree totally. We as human beings in a complex society have many different kinds of relationships, and we value these relationships differently. Family, siblings, school mates, colleagues, etc. These different relationships comprise our complete social network. See my post onSocial Network Analysis 101.Facebook is really just a social graph that reveals one kind of relationship. Before facebook can allow us to enumerate the various different relationships. There will always be another graph that accentuate the various different relationships we have. For example FB for friends/acquaintances, LinkedIn for colleagues, business partners, twitter for casual chatter, etc.

  80. team national

    Thanks for the information.It’s really a very informative and innovative post.Keep it up.Keep blogging.

  81. hanxi

    ┈ ﹎ ┈ ┈ .o ○ ﹎┈﹎ ● ○。。!▂▃▅▆┈ ┈ /█/▓ ﹎ ┈ ﹎﹎ ┈ ﹎ ““““““““““““““““““““The ample capacity of the Hampstead is perfect for everyday use. Carried on the shoulder, it comes in Damier canvas andfeatures the Louis Vuitton Inventeur metal plate.Damier canvas with smooth leather trimShiny golden brass hardwareAdjustable flat leather shoulder strapsGolden “Louis Vuitton Inventeur” plaque on frontGusseted sides with snap closures to increase bag capacityOpen top with clasp closureRed textile liningInterior patch pocket, cell phone pocket, and two large zipped pocketsMade In France…

  82. Scott Germaise

    “I am not sure if Facebook’s ambition is to create the one social graph to rule them all but if it is, I don’t think they will succeed with that.”Probably not.Yes, in one sense we all have one social graph. If for no other reason that you can draw such a thing with the bag of protoplasm, DNA, fingerprints and favorite desserts that is “You” at the center. But as a practical day-to-day matter, there’s many fully separate graphs that have no need to be brought together.The thing is… most people neither want nor need such a thing. While humans are communicators right down to our very core, none of the zeal with which we’re throwing out our online tendrils eliminates the idea that we still have various roles that are naturally separate. They don’t need to be brought together. And for marketers, even behavioral tracking doesn’t really require the sameness / differences to be understood. (Although ok, some similarity functions might depend on this.) Besides different roles, graph connections are complicated and rendered at least incomplete by some odd asymmetries in the adoption level of our various connections as well. For example, a personal policy of mine is to have a fairly clear separation of friends on Facebook vs. Contacts on LinkedIn. So far so good. But let’s take Farmville for a moment… (Yes, I admit it. When not playing Ghost Recon, I’m growing virtual wheat.) Not all of my close friends play. In fact, it’s mostly my ‘edge’ friends. I play Farmville with a friend’s ex-wife. (Ideally no one will think it’s wrong that I fertilized her crops the other night with my wife sitting right next to me.) Another neighbor is the estranged daughter of a good friend, and others are friends I’ve not actually spoken with, much less had BBQ or a beer with, in years. Anyone doing a touch point analysis concluding these are my closest buddies would be all the way wrong. Twitter? For me it’s mostly a semi-used business thing I play with on occasion. I’ll just leave aside any Meetup groups or wholly offline groups I belong to.For netheads, this may be fun to think about. I share this interest. But for most normal people, there’s neither need nor desire to bring all these together. In fact, there’s some good sense in keeping them apart.

  83. hanxi

    (●?????) (●?????)/█ ?/█╔╗ * . * . + * .* .+║║╔═╦╦╦═╗+ . *. * +║╚╣║║║║╩╣* * . + *╚═╩═╩═╩═The ample capacity of the Hampstead is perfect for everyday use. Carried on the shoulder, it comes in Damier canvas andfeatures the Louis Vuitton Inventeur metal plate.Damier canvas with smooth leather trimShiny golden brass hardwareAdjustable flat leather shoulder strapsGolden “Louis Vuitton Inventeur” plaque on frontGusseted sides with snap closures to increase bag capacityOpen top with clasp closureRed textile liningInterior patch pocket, cell phone pocket, and two large zipped pocketsMade In France…

  84. Elmira Gazizova

    “The access to public user data (social graph) should remain public and shouldn’t be guarded or controlled by any non public body (corporation, private company and etc.) ” Enough said.