Dispersion and Entropy In Social Media

On Monday, I trained it up to New Haven to meet a Yale professor named Dina Mayzlin and talk to her class. I thoroughly enjoyed talking to Dina's class as it allowed me to work on some new material in a comfortable setting. But the talk Dina and I had over breakfast before class was even more thought provoking.

Dina got her PhD at MIT's Sloan School a decade ago, before she started teach at Yale. Her thesis looked at TV shows being talked about in the social media of that time, newsgroups, IRC, Usenet, etc, etc.

What she and her colleagues found out was that volume (number of mentions) was not a good predictor of popularity. Volume was more of a trailing indicator than a leading indicator.

But Disperson, or what Dina calls Entropy, turned out to be a very reliable leading indicator of popularlity of a TV show. The wider and broader the discussion of the TV show went within online social media, the more likely the show was to become popular.

By coincidence, the material I am working on in my public talks right now is about the fragmentation of social media. And so as I talked about fragmentation with Dina's Yale class, I started to weave her work, which was still rattling around in my brain, into my fragmentation thesis.

I am totally convinced that the world of social media is not consolidating around one "winner takes all" social platform. Instead, the world of social media is fragmenting into dozens of social platforms that are best of breed for a certain kind of social engagement. If you are building a social media strategy today, you absolutely need to address Facebook, YouTube, Twitter, and Tumblr. And you should also consider Foursquare, Instagram, Pinterest and Path. If you are in the music business, you need to consider SoundCloud. If you are in the book business, you need to consider Wattpad. If you are in the TV business, you need to conside GetGlue. And so on and so forth. Many of the companies I just mentioned, but not all of them sadly, are USV portfolio companies.

That's the thesis I spent thirty minutes on in front of the Yale class. But near the end of the talk to Dina's class, it occured to me that disperson/entropy can be gained by engaging on multiple social platforms. The number of likes on Facebook or tweets on Twitter is volume and is likely to be a trailing indicator of popularity. But if you track the essential social gestures across the fragmenting landscape of social platforms, likes, tweets, tumbls, checkins, pins, etc, then you get a measure of dispersion that may well be a leading indicator of popularity or the slope of the popularity curve.

That's the theory anyway. I'll leave the research to Dina and others. I hope someone will run the numbers to see if it works.


Comments (Archived):

  1. Bill Bishop

    Wattpad is very interesting but they are not, as they say on their about page, “Today, Wattpad’s cloud-based platform is the most popular eReading community in the world connecting millions of users with over 200,000 new stories shared every month.”There is more than one Chinese eLiterature site that is much bigger. Then again, perhaps we have one world, two Internets

    1. fredwilson

      the twitter clone in China, Weibo, is bigger than Twitterwe do have two internets it seems

      1. Brandon Burns

        Despite western media’s dubbing of Sina Weibo as a Twitter clone, that’s hardly the case. It has many additional functions that Twitter doesn’t have. You also have to consider that in the Chinese language, 140 characters is equivalent to about 500 characters in a western language — that changes the usage quite a bit. Not to mention a slew of other cultural differences with how the Chinese use the Internet.Just some friendly tidbits from someone who’s lived and worked in the Internet industry in the Middle Kingdom. :o)

        1. Henry

          There is no doubt that weibo is a twitter clone just like how renren is a facebook clone. Yes, they did some adjustments for cultural reasons, but nonetheless they are clones. The basic idea, offering, and execution didn’t change. Heck, weibo didn’t even bother to change the 140 character limit idea. China even has a tumblr clone (diandian.com), pinterest clones (too many to name, a google search would get you far)- well basically, they have a clone for almost everything.You can probably attribute the lack of innovation in China to their education system. I can’t really see what I’m writing after this…so I’m going to stop…

          1. Brandon Burns

            I 100% agree that China is a clone machine, and I don’t like it any more than you do. Nevertheless, you can’t fully understand these products by comparing them to the “clones” from which they came — you have to understand where they are today, how they got there, the differences, and analyze them in their own contexts. Analogy: Facebook did things differently from MySpace to win… but that was 7 years ago. You can’t hold on to that comparison to fully understand Facebook today. – posted via Engagio

          2. Henry

            Facebook vs. myspace is totally different from renren vs. facebook. I wouldn’t call facebook a myspace clone, but I would call renren a facebook clone. Not sure if you’ve used the service before…but if you have you’ll know what I mean.I would type more and put renren and weibo in their own context and how it parallels facebook and twitter but this thing doesn’t expand so I can’t see what I’m typing…

        2. jason wright

          The Middle Kingdom. Yes.This is the cultural and historical mindset of China as it faces the rest of the world. This is its perspective on the world beyond its borders.The first character means middle or center, the second means country – the country at the center of the world. Everywhere else is at the periphery. It’s a strong message, both to itself and to everyone else.Know your enemy, and know your customer, and know your fellow human being, but not necessarily in that order.

      2. kidmercury

        and more on the way!

  2. JLM

    Many an engineer has had their studies wrecked on the hidden rocks and shoals of “entropy” and has learned to hate the Second Law of Thermodynamics.Good to see that the information intelligentsia have hijacked this terror inciting utterance.The world of information will get a bit more complex before the winners and losers are sorted out and what we think of any — every — bit of social media or platform is going to evolve in such a manner that we will not even recognize from whence we have come.There are literally no boundaries as to what is possible and therefore we can not only anticipate but we can expect — everything.

    1. Rohan

      I miss your avatar, JLM. Is this ‘file is broken’ thing intentional?  🙂

    2. ShanaC

      a bit more – how about a lot more.  Big data means more correlations which means more data.Harder to be actionable.

  3. awaldstein

    Entropy and expresso @ 6AM.My gut belief aligns with yours Fred. The amazing and inherent dichotomy of the sociall web is that while it is built on the possibility of bigger, faster, the power of ‘more’, it has driven human need and behavior to search out the individual and the contextual and the niche.But with that, whether dispersion is the metric of value? Whether cross niche and interest is the true indicator, I don’t know. I can internalize it but it is escaping me how to begin to measure this. 

  4. Dan Lewis

    “If you are building a social media strategy today, you absolutely need to address Facebook, YouTube, Twitter, and Tumblr.”Email *absolutely, positively* needs to be in this list.  It has a higher adoption rate than any of them and is relied upon as frequently as any.  It’s also been taken over by mass marketers, direct-mail style, so there’s room for disruption by simply changing how you message.  (In fact, one of the top mistakes I see organizations make — worse than ignoring email — is to treat it as direct mail.)You get someone following you via email and that connection is incredibly strong, easily stronger than any of the four you mentioned as “absolute[] needs” (which I also agree with).  All the social tools are already built in, as is the social network of the reader.  There’s a forward button and a reply button.  There are address books and the like.  And there’s always the copy/paste feature :)Email shouldn’t be overlooked.

    1. fredwilson

      Absolutely. But email isn’t social. But it is social media’s best distribution system

      1. Dan Lewis

        Email is social if you intend it to be.  Like all the other things cited, they’re all just distribution at the core — heck, the same can be said for blogs, too.

        1. William Mougayar

          But the question is- How has Email changed with social? You have to look at that. As Fred said, it’s social’s distribution option. Add to that, plugins like Rapportive that give you a social snapshot of someone you are about to email or have received an email from.What else?

          1. CJ

            Your address book is your email contact’s facebook page.  It’s where information about that person lives and gets updated in the best manner, with things you observe rather than things that are directed to you by another.  That makes it more relevant, a very stronger indicator of relationship strength and also the easiest social information source to obsolete.  As that bond starts to weaken you realize that a contact you have is no longer at the same job, or has changed phone numbers…all without notifying you.  As that bond strengthens you get the status updates that alert you to those very same things.  The address book, and email by extension, is the very best way to judge the strength of a social relationship and that makes it, at its core, the very definition of social.  

          2. Dan Lewis

            Wow, really well said. 

          3. ShanaC

            Actually, it is more that that.All social services want to talk to your email contact book.  They think it is the primary source of your friendships, your contacts, your tastegraph.Only somewhat true.  But interesting nonetheless.

          4. William Mougayar

            You may be right about the address book’s potential. But it needs to change from its current form. I know of one company that’s about to have something new related to this segment. – posted via Engagio

          5. fredwilson

            stay tuned for a post and a new service in this area. should be this month i hope

          6. Chris Phenner

            File under:- Contactually- Hashable *- Mingly* And I hope Hashable catches up.

          7. Dana Byerlee

            Well put. This is our mission at Mingly.  Entropy – this concept applies on the micro level as well.  In his seminal study, The Strength of Weak Ties, Granovetter found that individuals w/ more weak ties have greater opportunities for mobility. http://ow.ly/9etVf (This makes sense, as “strong” ties tend to be quite similar to you- i.e., most info will be old news. But “weak” ties, where individuals do not necessarily have knowledge of/ access to the same people, will be more likely to transmit novel ideas, news, and opportunities.)  The problem of valuable connections being fragmented across multiple platforms is that they are often neglected or lost. So at Mingly we see future of building relationships lies in turning currently dispersed information into meaningful action. No matter how you’ve connected, via email or social networks, Mingly keeps track of the people entering your life. You can quickly tag contacts in groups, add notes, and set keep in touch reminders. We also notify you about key events in your contacts’ lives. Here is a Mingly demo video I made the other day (shameless plug!). http://ow.ly/9eujZ  Please reach out any time, I love chatting about the social intelligence space. [email protected] 

        2. JamesHRH

          no, its not.Email is the ‘Of Record’ online communication platform, for the vast majority of people. 

          1. Emil Sotirov

            “Of Record” – compared to oral or phone exchanges, yes… but also “off record” – compared to what you post on public social media.

          2. Aaron Klein

            Explain to me the fundamental difference between sharing a link with five friends via email or via Twitter.Public != Social

          3. fredwilson


          4. Andrew Hoydich

            Using Twitter, how many times do you share a link with just 5 friends? Usually it’s with 25/50/100/400/etc. Using e-mail, how many times do you share a link with just 5 friends? A decent amount.Maybe there is no fundamental difference, the concept behind the two methods of sharing may be the same. The way they differ may be on their degree or level of social. E-mail is a social event that takes place in a private place, whereas social interaction on Facebook or Twitter happens in an open environment. One is semi-social while the other is ultra-social

          5. Aaron Klein

            Totally depends on how many followers or friends you have, and whether they’re engaged enough to follow your links on a given subject. I’d venture to guess that for the average web user (aka not someone who frequents AVC), email sharing promotes more raw clicks for the right topic.

          6. JamesHRH

            I just sent a link in an email to 3 people. Of course, none of them are on Twitter – :0- posted via Engagio

          7. Andrew Hoydich

            It definitely does depend on a lot of factors, so to try and say that one method of socializing must be better for every individual is foolish. 

          8. André DeMarre

            “Public != Social” – Right on!Conversations about what is and isn’t social are awkward because we lack consensus on what social means. The concepts of public and private are insufficient because they’re not black and white.To what size audience does a communication channel need to be broadcast to be considered public? Imagine a company-wide forum in an organization of 500 employees. Is it no longer private because of the size of its audience? Is a post on Facebook not social if it is restricted to only close friends?I keep a very loose definition of social. Social is about connecting people, and this happens simultaneously in many-to-many, one-to-many, and one-to-one relationships. For me, any communications medium is social, as is any software that connects people. I posit that version control systems are social, even before GitHub happened.Social isn’t novel anymore; hasn’t been for a long time. What is new though is the revolution of massively social applications, and this change in scale is inspiring many exciting new ideas.

          9. Aaron Klein

            Very true…

          10. leigh

            When my hubbie uses it like instant messeger with me it is email #1:  what time are you going homeemail response:  6 why?  what were you thinking for dinner?And so on……

          11. JamesHRH

            too funny

          12. fredwilson

            not public, not social

          13. leigh

            lol unknowingly you just described Peter 🙂

          14. Andrew Hoydich

            E-mail is certainly just as much of a social medium as facebook/twitter/tumblr/etc. You are being social when you interact with people via e-mail just as you are being social when you interact with people via facebook/twitter/etc. The difference is at what level of the social realm they operate. The way I understand it is that e-mailing is a form of socializing that is extremely personal and exclusive like going to a ‘members only’ club (no strangers), whereas using facebook/twitter operates partially at a surface level and maximizes your ability to interact with new people and form quick, but loose connections like going to a metaphorical watering hole of sorts.

          15. JamesHRH

            You have to judge these services by the majority of usage time when you look at the majority of users.Email is where important things get sent. Leigh’s hubbie must find dinner important or he most not be able to find his phone to text.Can you use all mediums for anything? Obv., yes. Send me an email means you are no longer public, its gotten serious. It is of record or off the record – serious, thx @eilsaritov.- posted via Engagio

          16. Drew Meyers


          17. Andrew Hoydich

            being social is not a matter of public or private. going to a friends house is being social just as going to a bar/party/orgy is being social. therefore, e-mail, even though it is a form of ‘private’ communication if you must, is still a form of socializing.  

      2. CJ

        Isn’t it social though? At work I’m on several distribution lists, those lists often receive information directed to them solely due to the nature of the group, information that isn’t shared outside of said group, it’s kinda like a Facebook group in that way. When we communicate with each other we share information and build stronger bonds based on a common or discrete interest, isn’t that social?Outside of work, I have informal groups of friends that I email, often at the same time, about certain topics.  The groups change based on the topic of conversation, some friends are privy to more personal discussions, some really grok the internet, others are all about financial matters, in every case social connections exist and we use email to strengthen and maintain them at a level outside of the networks but the ties are just as strong if not stronger.  It’s the very definition of social to me, it’s just not sexy.  

        1. Drew Meyers

          yup…agreeOne thing I will say..is that Facebook groups/lists are *starting* to take the place of SOME of my email. I sometimes share things with my “close friends” list that I would have emailed two years ago. But it’s certainly not as strong a message. I find the quality and length of conversation longer and higher if I email a message to a group of close friends rather than share it with them via a fb list. But even so, I email more selectively now than I used to.

      3. Lee Blaylock

        Whoa Cowboy, I’d beg to differ that email isn’t social.  When a company sets up a distribution list in MSFT Exchange or Notes, or if you set up one yourself, that is no different (in some respects) to Google+ Circles, FB Friend categories you set up or folks you send out a message on Twitter.Are they indexed by internet search engines, no.(but i could be snarky and say is it if you’ve been sued and in Discovery – ask Bill Gates about the email that was found that cost him $1b)Are they persistent and easily located by others in the enterprise, noCan an AI app gain sentiment analysis and other forms of intelligence, actually yes, but no easy to pull off politically in an org.but do they carry a message to a broad audience quickly?, Absolutely, plus you are _much much_ more likely to read almost your entire email “stream” than you are likely to read even a fraction of the tweets or wall posts you get.  This might beg the question to ask “what is social” but social has elements of expression of self, sharing, broadcast and advanced social is targeting those to specific messages and sharing desired.  Email absolutely has all those traits.email is pure social?  no, i wouldn’t go that far, but i would argue it is the most effective social use, at a minimum for business professionals.  email can most definitely be improved (Google Wave had early promise but fizzled), but it isn’t going away anytime soon. I’m also of the opinion that your real social network is in your smart phone and your email account.  Those are the people you trust, call, email, invite to parties, travel with and do business with.  All these social networks (FB, Linkedin, etc..) are actually synthetic networks b/c to friend or link to someone is a double opt in with political ramifications if you do or don’t accept, + even sending a friend or link request implies some political capital being spent. Yes, twitter followers and folks who follow bloggers like Fred can create a reputation in Fred’s mind about a follower who consistently makes good or poor posts, but most people in the world don’t have the platform widely read bloggers do.  Bottom line is who you store in your email contacts and smart phone are not 100% of your social networks, but it is the one that most accurately reflects your true social graph.

        1. fredwilson

          things that happen privately aren’t socialbut i agree about the power of the address bookthat’s a different thing

          1. Ali Diab

            ‘Reply all’ isn’t private, and it occurs on emailDirect messaging is private, and it occurs on FB/TwitterSo, email in the above context is social, while FB/Twitter aren’t

          2. Andrew Hoydich

            First off, e-mail IS direct messaging. That’s all it is. You choose who you want to talk to and send them a direct message. Tweeting is the exact opposite of direct messaging. You tweet something without choosing who will read it and who won’t. Second, I would say that ‘reply all’ is absolutely private. The information being sent to the recipients of the e-mail is private to all others. However, when you post a message on someone’s wall or tweet to your followers, that information is displayed for all to see. 

          3. Chris Phenner

            ‘Things that happen privately aren’t social.’I think this is worthy of its own post/thread, if only because it created a good-sized thread on AVC.  And because I think ‘private’ & ‘social’ are worth parsing further, or perhaps better-put, worthy of consideration in the fragmentation of social meeja (the post’s core topic).Let me be more specific about ‘privately social’ moments:- A ‘dontpost@’ via Hashable- A scrobbled song on Last.fm- A GPS data stream to [service]- A Klout-scored (yet anon) event- A post to idonethis.com (via email)In each of the above use cases, the ‘moment’ may be very personal or private, yet data was created and captured.  If I hear 10 songs via Rdio that are Last.fm-scrobbled and posted via FB’s NewsFeed but the songs were not individually named, did they make a sound? (bad joke at the end)My point is that the *content* of socially-collected data may remain private and personal, while the *event* having happened may be made socially, publicly available.Someone’s probably working on an Engag.io for that.

          4. fredwilson

            it’s just my beliefi am wrong all the time about stuff

      4. bernardlunn

        In my world, email is social. So is Skype. So is the phone. So is a coffee shop or bar. I created a travel blog using Posterous, for friends and family. The obvious alternative was an email list.

    2. Aaron Klein

      Completely agree, Dan, and here’s an example. In the world of investing, there are a ton of “informal social networks” built entirely on email to share investment ideas among 5-6 friends at a time. Which funds have low fees? Which stocks are feeling good? What sectors are going to climb or fall?These are 50- or 60-somethings who don’t get Twitter, and only half of them have a Facebook. It’s a far more exclusive social network than Path. You almost need the assent of every member of the group to add someone in.We’re just starting to roll out some features in Riskalyze to tap these email-based social networks and let them share those investment ideas from our platform. Will that sharing activity move off email onto our platform, and email becomes just notifications? Perhaps.But for now…email is the network. And it’s fascinating how social it is for this demographic.

      1. Dan Lewis

        A ton of demographics are like this, too.  My wife and I (more her than I), as parents, coordinate stuff with other parents in our children’s schools almost exclusively over email.  Google Wave was trying to change that — email’s not designed for that, but it’s used that way.I think Groupon and LivingSocial really understood this. They’re injecting things like yoga classes and restaurant specials into the email ecosystem in hopes that we forward them on to our pre-existing email groups. I get one, forward to friends “hey, you want to try this restaurant” and yahtzee.

        1. Aaron Klein

          I wish this was more trackable. You end up not knowing how much sharing happened, and only how many resulting actions occurred, so there’s no way to track a conversion rate.You know what would be really powerful? An < AddressBook > tag in HTML5.Imagine being able to allow users to address an email to their friends from within your web app, without having to upload their address book or connect to their Gmail account.

          1. ShanaC

            Why – isn’t that an array of essentially what should be <email> tags?

          2. Aaron Klein

            My three choices for sharing right now would be a Tweet button, a Facebook Share button and an email button.So what does the email button do?Quick and dirty, I can use Javascript to open an email message in the user’s native email client. Then their address book is immediately available to them. But I can’t track whether they chose to send the email or not, so I can’t calculate a conversion rate for this sharing.Or I can give them a web form, but they’ll have to type, or copy-and-paste, or upload a CSV file full of email addresses. Trackable but lots of work, conversion goes down.Or I can do the whole connect-to-your-Gmail-account thing, which raises huge trust issues. Trackable but conversion goes down.Imagine if you could simply code an “address book” field into your web app. The browser would have permission to search your address book wherever you keep it, and insert the email addresses you choose to insert.A much better user experience and much more fluid sharing via email.

          3. ShanaC

            Sometimes I think we’re asking too much of the browser. It may be turning into the new operating system, but it isn’t there yet- posted via Engagio

      2. ShanaC

        The most exclusive social network is the one you actually talk to.  Most people don’t realize that just because someone replies to you online, doesn’t mean that they want your real opinion

        1. Aaron Klein

          So you’re saying you don’t want my real opinion? 😉

          1. ShanaC

            I want your real opinion any time.  I just make the point that just because someone talks to me on services like facebook, doesn’t mean I’m close with that person.  Something about edgerank functionality.

          2. Aaron Klein

            I know, I was just kidding with ya. 😉

        2. awaldstein

          Change ‘opinion’ to ‘connection’ and I agree with you,How many of your FB friends cross the ‘connection’ threshold and ‘chat’ with you.Not many. Sometimes it’s a great thing, often it feels like ‘who is this stranger that is talking to me?”

          1. ShanaC

            Quite more than you think.  Actually some people I don’t have any other contact form other than facebook.  In some cases, the primary point of contact is only sms.This may be a generational thing.  I actually want to switch away to something else, but the network effect is too strong and there seems to be too much variance in technology pickup among my friendsEG: among my friends who I would shop with I’m the only one who uses pininterest, even though almost all of them like the same stores I do/etsy.(among the reasons I love my friends, I see these differences as a positive thing.)

          2. awaldstein

            Thanks.I see this as less generational and more topical.My European wine blogger community which is sizable just won’t get off of FB. Such is life. 

    3. Jordan Elpern-Waxman

      It’s amazing how debate Dan Lewis’ comment on whether email is social has created. I agree with both Dan and Fred:- Email is definitely social when using the lay definition of the word ‘social’.- Email is definitely in a different category from public sharing a la Twitter. I’m not sure whether in “social media” vernacular this is considered ‘social’ or not, so I defer to Fred on this.However if the above is true, this speaks to an ironic shortcoming with the use of the word ‘social’ in social media circles.PS Quick plug for Dan’s awesome Now I Know daily email here: http://dlewis.net/nik/. I’m a huge fan.PPS Dan, I think AVC is where I first discovered NIK. Now we can close that attribution loop 🙂

  5. jason wright

    Was your talk captured on video for future dispersal, and discussion, and popularity? I’d like to see it.Where’s ‘Fred TV’?

    1. fredwilson

      Nope. Sorry about that

      1. jason wright

        Groan.Student fees must continue to guarantee privileged access. That’s the education business model.   

        1. fredwilson

          Don’t groan too loud. I’m warming up and working on a new talk. I will give it publicly soon

          1. jason wright

            Selling tickets?

          2. fredwilson


          3. jason wright

            Yippey, but I’ll bring a fistful of Venoms just in case 🙂

  6. Rohan

    Not dissimilar from old media, I’d think. For Facebook, Twitter etc.. you had news paper, television, radio and direct mail in the past. And marketing budgets were allotted to all of these. There was no talk of 1 ‘winner’.And it’s going to be the same for social media I’d think.Different mediums, different approaches that’s all.

  7. opoeian

    Exactly! And regarding the necessity to have both a generic social platform presence and a niche social platform presence – the way I like to picture it is – at the moment most websites have a generic FB ‘like’ or Twitter ‘tweet’ button for example. But in the near future all music-based websites will have both a generic set of buttons as well as a similar button linking to a music social platform, such as Soundcloud, or all film-based websites will also have a similar button linking to a film-based social platform. A layer of niche ‘like’ buttons, if you will, linking to niche platformsI believe the future of social is an integrated structure of whole generic horizontals (FB, Twitter etc) with ‘whole niche verticals’. In fact we have designed the strategy for our own niche social platform with this very shift in mind

    1. fredwilson


    2. Josslossboss

      Like the Grimstr said (above):USERS WILL MAKE NICHE WITH TOOL FOR ALL.= a platform model ‘whole niche vertical’ – built for and by users

  8. Benjamin Kuna

    What I would add to your thesis, is that it is not that easy to generate fake traffic when it comes to social gestures across many platforms.The likes a page has are (at the beginning) are often manipulated by PR/marketing companies being in charge of profiles that shouldn’t exist, in order to anticipate a viral effect.Whereas observing the social gestures on tumblr, twitter, facebook, foursquare etc is a better way to indicate organic growth. 

    1. fredwilson

      Great point

    2. Chris Ellis

      ‘Truth is repetition’~ Kierkegaard I think. If you haven’t already look in to signalling theory. Signals are ‘hard to make, impossible to fake’. What Social Networking Services have done is made it easier for us to spot the fakes because authenticity is what endures over time and as you mention some services provide better quality signals/gestures than others.

  9. John Best

    Hrm. I saw my expansive reply appear, and then disappear on next update. I’m hoping Disqus hasn’t eaten it.

    1. jason wright

      I see it.

      1. John Best

        Curiouser and curiouser. I still can’t see it, my profile also isn’t listing it. :S

        1. jason wright

          …but now I don’t. It was a good read…while it lasted :-)Fame is fleeting, entropy is enduring…or so they say.

          1. Rohan

            You’ve got a great sense of humor, Jason! 😀

    2. awaldstein

      There are issues this morning.I’ve been posting through engagio and Disqus has both been double posting and vanishing.

      1. John Best

        Thanks, that’s reassuring. I hope it manages to find it’s way back.

    3. fredwilson

      It may have gone into spam. We will check

      1. John Best

        It’s not appearing in my disqus profile, I don’t know if that means its just gone. I’ve recreated it above anyway 🙂

  10. Dave Pinsen

    There is an interesting dynamic between Twitter and TV. Twitter gives you an incentive to watch TV live so you can follow commentary live (which has been great for the debates, in particular). That’s good for TV since there’s more chance you’ll catch some of the commercials. So maybe there’s a way Twitter can work out to get compensated for that by the networks.

    1. Tom Labus

      Good luck on getting paid by the Networks.

    2. Ciaran

      Very true. Talking to TV companies, they’ll claim that they are driving social media usage. But on the flip-side, listen to Twitter and they’ll claim that they’re driving the TV viewing.They’re both right, so Twitter should be working out how to get paid, but shouldn’t be expecting it from the TV stations, otherwise the stations would have an equally valid argument to get paid by Twitter for providing the content their audiences congregate around.

    3. fredwilson

      That would be like Wikipedia getting a piece of google search adsNot likely

      1. Dave Pinsen

        Then maybe a broadcasting company should buy them.

    4. matthughes

      It’s Super Bowl Sunday this week – they better hurry 😉

  11. Matt Pierson

    Unsurprisingly, Nielsen has done research into this (I know because I used to work there). They looked across all of social media and did find a correlation, including a predictive element. http://blog.nielsen.com/nie… also a service that developed out of the MIT scene dedicated to TV buzz in social media, http://bluefinlabs.com/ .

    1. fredwilson

      Thanks. I will check out that work

  12. William Mougayar

    I would argue that Commenting on blogs is the highest and most engaging form of social gesturing. Much higher than liking, sharing, linking. It takes 1 sec to share, but often 2-5 mins to post a comment & that shows a higher commitment from the users that are involved. You also get more out of commenting because you are interacting with others & developing relationships. That is the Engagio thesis.

    1. JimHirshfield

      Don’t forget copying. Ctrl-C is one of the strongest signals of engagement. Someone that’s copied text has read it and thought about it.

      1. William Mougayar

        Isn’t that captured under Sharing?- posted via Engagio

        1. JimHirshfield

          Not really. “Sharing” data today usually includes “likes” and “shares” initiated by users pressing social media share buttons.Copy/paste is a form of sharing, but it isn’t typically included in the sharing numbers you hear thrown around. There’s really just one company, the one I work for, that tracks and reports on this behavior.

        2. JimHirshfield

          Oh, and all copying is not sharing. People copy to search, share, or preserve.1. copy/paste into google = search2. copy/paste into email/fb/twitter/etc = sharing3. copy/paste into ms word = preserve (as in academic paper’s footnote or other research).

          1. William Mougayar

            Wow. You’re over my head now…You are the Copy master…Tynt! – posted via Engagio

      2. fredwilson

        Tumblrs bookmark let is awesome for that. I use it to do that all the time

        1. JimHirshfield

          I hadn’t thought about that. A great (passive) signal.

      3. Aaron Klein

        I still remember that the absence of this feature was my biggest frustration on the iPhone back in 2007.

        1. JimHirshfield

          And I remember everyone hailing it as a revolutionary feature when iPhone finally added copy/paste.  😀

    2. Brandon Burns

      Nice theory, but don’t forget that 1) AVC is rare, not everyone on blogs spends 2-5 minutes responding, and 2) sharing very often is coupled with a comment/explanation written by the user to go along with the post. I agree that people have deeper connections with blogs they frequent, but I’m not sure your logic is the right way to back it up.

      1. William Mougayar

        In terms of what value received, I think engagement in social communities holds true. that’s not theory. Aside from AVC, there are other vibrant communities in every field. It takes consistency and repetition. You have to show up and give your social capital before you can reap the rewards, and I don’t mean to preach to you because you know it well. – posted via Engagio

        1. Brandon Burns

          I agree that when you look at value with relation to the platform on which you found the content, commenting is valuable. It enriches the community, shows your investment which can enrich your personal experience and, even if it doesn’t link out to a social network (such as our convo right now) the poster is still more likely to be an evangelist and bring new people to the platform (such as I’m sure we both do). But when you look at value with relation to outside people, whether I leave a comment for them to read or share to my Facebook wall (with an explanation of why I posted), I’m not sure which helps convert that person into a new user more. Not exactly apples to apples. BTW, loving Engagio. If I could import my email, too, I’d dump Gmail today. Though maybe that would taint the experience? I’m sure you’ve thought about it more than I have. 

  13. John Revay

    I hope you took the Amtrak NYC>NH – much more comfortable ride than Metro North, not as many stops either, next time make a quick stop in Fairfield, we have a Shakeshack across the boarder in Westport.Interesting read – between your post earlier this week about Search vs Social, the twitter piece yesterday…Dina’s research, and your thesis about – “social media is not consolidating around one “winner takes all” social platform”  i am wondering if that has any thing to do with the timing of the FB S-1.  I thought last year they were talking Summer of 2012

    1. fredwilson

      I am not thinking about Facebook. I am thinking a lot about Twitter, Tumblr, Foursquare, SoundCloud, Wattpad, GetGlue and thirty-five other companies

      1. John Revay

        YUP -I agree w/ your thinking and focus. I was just thinking FB was accelerating their offering ……..possibly they see something in their data showing that they are peaking.. – posted via Engagio

        1. jason wright

          That sounds very plausible. The FB wave about to break on the shore. 

          1. testtest

            saw this on stocktwits:HMMM…Facebook revs: $3.2b/year, est. mkt cap $75b. $GOOGrevs $38b, cap $189b .. $AOL revs: $2.2b (but fallin’) cap $1.7b .. just sayin

          2. jason wright

            Interesting ratios. In the age of the secondary market Goldman Sachs is king-maker. Long live the king – I don’t think so.

  14. MorePancakes

    This is pretty neat. Let’s take it one step further and say that “dispersion” is the trailing indicator of “sharing acceleration.” That is, in order for something to be shared on lots of sites and platforms, certain users/services act as “accelerators” at the very beginning in order to get more people sharing it on those other platforms. A rough example of a sharing chain may look like: Hacker News–>Twitter–>Blogs–>Reddit–>Facebook–>Newspapers/Magazines. In this case, a single person sharing at the very beginning (Hacker News) can be more valuable than a person sharing at the end of the chain because they get the sharing started. It’s kind of like compound interest; it’s better to make gains at the beginning because you can leverage those going into the future. I think the lesson for all of this is to realize how important the beginning of the chain is. It’s the “accelerator” that makes dispersion possible. Any thoughts?

    1. fredwilson

      Yes. You are on it this morning!

    2. Trish Burgess-Curran

      Fully agree.  It is always hard to get the ball rolling but, once it is going, it seems to take a life of its own!  In any case, just as I mentioned in response to @jasonpwright:disqus , if you have a quality conections that promote your message, spreading the word becomes particularly simple!

  15. John Revay

    “Many of the companies I just mentioned, but not all of them sadly, are USV portfolio companies”Question about the ones that are not in the USV Portfolio – did they get away…or did some not even get reviewed/seen by you and your partners. Was it timing in certain deals ( re: gap between Flatiron and USV)

    1. fredwilson

      Lots of reasons. You can’t bat 1000

      1. John Revay

        Sorry did not mean to question your bating average…I just recall on one post when you commented about a company (possibly bump) ” that one got away”. My sense is last time you raised a fund you were over subscribed – one proxy as to how well your stock trades.- posted via Engagio

  16. JimHirshfield

    I like this theory of dispersion because it inherently involves diversity of voices and platforms. Helps solve the problem of false signals from paid likes and RTs.

    1. jason wright


      1. JimHirshfield


        1. jason wright

          Yes. I’m a twit. 

          1. JimHirshfield


  17. Tom Labus

    Look at Stocktwits.There’s going to be communities for all types of interests but real different from what we see today.

    1. fredwilson

      stocktwits is a great example

    2. Aaron Klein

      Stocktwits is indeed an awesome community for discussion about stocks, and what Howard and team have done to drive discussion quality has really worked.

      1. Tom Labus

        They started in the worst days of the 08 crisis too.

    3. JamesHRH

      It is not so much dispersion and entropy, as it is specialization and focus.Why would I organize a page on Fb when there is a branded service devoted to that area?

  18. Julien

    I remember that you once said that you considered a company a good potential investment if you heard about it from several different people.I guess that’s the same idea here. If people who are not directly related pitch you a startup, then, yes, it’s likely that this startup has a good dispersion, and you used that signal.Generally, I guess a good way to measure upcoming popularity would be to measure the unconnected-ness of mentions.

    1. fredwilson

      That’s exactly right. Never thought of that analogy

  19. David Semeria

     Many of the companies I just mentioned, but not all of them sadly, are USV portfolio companies.That’s so cool. No resting on laurels for USV then…

    1. fredwilson

      You are only as good as your last deal in this business. Resting on laurels is death

  20. jason wright

    The theory of entropy, from Dina, to Fred, to AVC, to you and me, and then on and out and beyond to others on other platforms via other methods of distribution. The theory seems to hold true. The value of the message carries the message ever onward and ever upward in popularity.

    1. Trish Burgess-Curran

      The value of the message is very important but so is the value of the messanger.  In this case, Dina was highly respected by Fred, Fred is highly respected by the AVC community…  It is about the ‘level’ of the connections that you make as much as it is about the quality of the message and the diversity of the media.

      1. jason wright

        Yes, respect, and sentiment, and reputation. Things like Klout interest me. Klout is bottling something that goes beyond the mere message. Things like Klout are very important in the fragmented social network. 

  21. Ciaran

    “If you are building a social media strategy today, you absolutely need to address Facebook, YouTube, Twitter, and Tumblr.”No, I disagree. If you are building a marketing or communications strategy, you need to address the tools, platforms and places that your audience are using/congregating around. That might include some of the above (almost certainly Facebook, at least in the ‘west’. But Tumblr? Sure, it gets billions of page views, but what does a Tumblr strategy actually look like?  I think that increasingly, having a social media strategy is going to be like having a leaflet, or cinema strategy – they’re parts of a strategy, not a whole.We need to stop obsessing over places and tools, and get back to thinking about people. We should also bear in mind that for consumers and advertisers alike, there are only so many hours in the day (and, for the latter group, dollars in the bank): so you may well need to prioritise some places and platforms over others.Of course, as most of these are really places to share/congregate around content, the most important strategy is to create content that people will want to share/congregate around, and leave the places they do that up to them.

    1. fredwilson

      I will show you a tumblr strategy when I get to a computer

      1. Ciaran

        I’m honestly fascinated. I personally think we need communications strategies and tactics for individual platforms

        1. pointsnfigures

          I am trying to figure out why to be on tumblr.  I already have three blogs I do.  Why add a fourth? 

          1. Ciaran


      2. awaldstein

        I’d love to have a guest post from the Tumblr folks on this.I’m with you but this is not well understood. And this community carries a long tail of influence out to the business world.

        1. fredwilson

          i posted this above, but i’m not sure you’ll see it arnoldhere is how urban outfitters leverages tumblr as a marketing mediumhttp://urbanoutfitters.tumb…it is native, beautiful and effective

          1. awaldstein

            Thanks Fred…Native and beautiful for certain.

    2. ShanaC

      One of the problems of the internet is we’ve partially abstracted the people out of the conversation.  We expect at some point, even though we dislike this, that people will replace us.Maybe computers will just become consumers instead of pople (meant to be sardonic)

    3. fredwilson

      as promised here is a tumblr strategy well executedhttp://urbanoutfitters.tumb…

      1. Ciaran

        I’m at danger of being accused of being nit-picky and obsessing over semantics, but I think that shows that UO have a good digital, or even more simply, a good communications strategy, and they are using Tumblr to execute that. They seem to have a tactical plan for using Tumblr which, and correct me if I’m wrong, seems to consist of ‘just’ posting pictures of their catalogue.I’m not sure it should be classed as a strategy. And without knowing more about their objectives, I think it’s hard to say much more than that.

        1. fredwilson

          they don’t execute this “campaign” elsewhereit is built custom for tumblr

          1. Ciaran

            Exactly – it’s a Tumblr campaign which is an element, I would imagine, of their overall strategy ( probably driving reach, awareness and engagement). My worry about talking about Tumblr strategies, or mobile strategies, or Twitter strategies, is that you focus on the technology, not the user. As gapingvoid put it some time ago, “Technology changes, people don’t” – this isn’t 100% true, but it is more true than it is false.Your strategy should concentrate on your target audience, their media habits and other stuff like that; your tactics and campaigns concentrate on individual media or platforms.That’s why I don’t think that this is a strategy, but an tactical execution of a strategy. Or at least, that’s what it should be.

  22. gorbachev

    I’m having much more engaging conversations about specific topics on online communities dedicated to that topic whether the topic is Internet startups/venture capital (AVC), audio visual equipment (AVS Forums) or actually any other topic.Facebook is completely useless, because the conversations there are shallow.You can’t have a conversation in 140 characters at a time, so Twitter is out.I know social networks are all the rage these days, but for me all the action is at online communities (of all shapes). They’re more social, more engaging and frankly much more interesting than the social networks.

    1. William Mougayar

      I’m 100% with you about conversations and communities being the most engaging parts of social. That’s where the most value comes back to the user who is serious about it.

    2. Aaron Klein

      Some people like to have conversations. You and I are two of them. But I’d argue that Disqus is clearly one of the premier “social networks” for conversation.Others don’t. Perhaps they don’t have the time, but deep conversations aren’t the goal. Expressing themselves or sharing cool things with their friends is.And it’s fascinating to watch the “content form factor” as it shrinks over time.- Blogger lets people write 500+ words- Facebook shrinks it to photos and 50-100 words- Twitter shrinks it to 140 characters – just right for text, imo- Pinterest shrinks it to “publish with a single click”There will always be room for both.

      1. laurie kalmanson

        Agree, plus this: Pinterest’s “boards” are galleries, which allow users to aggregate their own conten and share collections a group at a time, vs a single item.

    3. JamesHRH

      Yes, but everyone my 16 yo niece ever meets can be connected to her Facebook page. An up to date, multimedia rolodex that is not that engaging but required viewing on some repeated basis (daily, weekly, etc.).BANK.

      1. laurie kalmanson

        one that shows where all the friends are on the other platforms, too; win#aggregate_everything

  23. William Mougayar

    Do you think this fragmentation will continue with more entropy or gradually with more order? Then, it becomes segmentation, not fragmentation.I think the fragmentation is on its way to get more predictable and known. The tools you mentioned & a few others are the proxies that let us consume or be part of this fragmentation.

    1. fredwilson

      I think it evolves into segmentationI said in my talk at social media weekend that social nets may end up looking like magazines

      1. William Mougayar

        Yes…content is back…in a round-about way. But it’s content+++- posted via Engagio

      2. Irving Fain

        Segmentation is the next wave in social. The data is there and being collected. It now needs to be more effectively understood and then utilized.

  24. Jan Schultink

    In theory once network can do it, in practice not.Managing social circles is just too complicated in one social network environment.Hence we use different applications for that purpose.The different brand and UX environment switches the brain in a different mindset.

  25. Ben Popper

    For publishers like myself, that’s the power of a tool like Tweetdeck, where I can easily engage with multiple networks from a single interface. 

  26. mediaeater

    I agree,  I think the specialty networks part is dead on.  Right now, I have framed the attention across different devices and niche services (something that we measure as a co.) as marketing effectiveness that yields the coveted attention and awareness prior to launch.   This awareness may drive tune-in but the quality of the offering will be what makes it have resonance and staying power.-adding more context – @ Trendrr.TV we looked at last years pilot season, as well as this years – It is what I based the above comment on – I could pull the data if helpful but it was not a fully reliable predictive measure yet. It will become more so as we start to understand the complexity of the equation.There are many variables that contribute to success the most critical one is people actually being aware of a product, show or whatever is launching. That measure in itself is important but it’s just the first step. Measuring and looking at anticipation through NLP has helped focus down further but in the end – a hit is hit and a brick is brick – and all the awareness and measurement of anticipation will not change that. #snakesonaplane

    1. fredwilson

      The mediaeater is in the house!This man knows his stuff on this topic

  27. Harry DeMott

    Seems like you are talking about reach, frequency and engagement.The popularity of a TV show – or any piece of content that can be consumed digitally these days – can be measured to some degree by the reach of that content (thus using all of the distribution platforms mentioned above), the frequency with which that content is consumed or rebroadcast (retweeted, shared, etc…) and the engagement of the end user who has a potential viral coefficient (do I retweet, repost, reblog, like, etc… or not?)The truth is that when you start to map out a social graph, inevitably each channel is full of dead ends. Send something to me via twitter and it is a black hole, I might consume it but it will not come out.Yet send it to you – and there’s a good chance it will emerge stronger than it came in given your activity on Twitter and large number of followers who are likely to consume something you found worthy of retweeting.that amplification effect is the dispersion you are speaking of – and the more platforms you can amplify the message on, the more popular that message is likely to become.

  28. Brandon Burns

    I also hope someone will run the numbers… but how? It’s far from obvious, maybe not even possible.Can you say that 100 likes (or like equivalents) on each of Facebook, Twitter, Tumblr and YouTube is better than 400 likes on just one of those platforms? Especially when viral reach is an exponential function? Of course I want to be everywhere, not just in one place. The problem is when I look at X number of interactions, do I want them concentrated or fragmented? Simple math says you want them concentrated; media strategies have operated on that principal for over 100 years and, due to the exponential nature of reach on the Internet, this belief system is stronger now than ever. And since the way we buy media, from radio to Facebook, is based on promising clients X interactions for Y dollars, it’s imperative to look at the dispersion argument through that lens. The dispersion theory resonates with me personally. It will resonate even more with the math to back it up, which is no small feat. For a fixed number of interactions, which is currently the only way we measure the value/cost of media, the math points to concentration to get more bang for your buck. A formula that proves dispersion is better, and shows how and why, would be our generation’s e=mc2.

  29. John Best

    Post recreated, since no amount of fiddling with Disqus seemsto make it want to play ball.I was lucky enough to catch a talk by Fred on thissubject recently (and blogged about it here – http://blog.redrookdigital…..For those interested, the livestream of the talk is here: http://livestre.am/1fWj8, Fred’s talkbegins at about the 30 minute mark. You may notice some particularly incisivetweets appearing on the screen behind him.We’re seeing users dividing their social networks by intent.The people you want to share with on Facebook won’t necessarily (although theremay be some overlap) be the same people you have on Linkedin. There’s a cleardivision of appropriateness and purpose – Facebook = bulk social, Linkedin =professional. Adherents use and share on these networks differently, and sharedifferent types of information. Some networks have realised this (and I cite G+and Path on the blog) by limiting or allowing segregation of “contactlists”. I think, however that users are using multiple social networksprecisely because of that desire to separate. Its far easier to filter thetransmission of information if it’s contained within individual networks. This is why Facebook will never “win” social. Facebookis constrained by the fact that it is Facebook, and its users are Facebookusers. There are types of information which users will deem inappropriate forFacebook, but still want to share. They will consequently look for an alternatenetwork to transmit via. Social can never be a “winner takes all”scenario.Effectively the users of the networks are acting aseither the barrier or transmission vector of information (I guess I’m talkingabout memes here). The entropy/dispersion Fred and Dina refer to occurs wheninformation is interesting enough or universally appropriate enough to bypass theself-imposed filters. Indeed I suspect the true measure of how successful a TVshow/other meme is not only how easily it jumps from one network to another,but how it is enabled to do that by the user. Some information is sointeresting that the user will purposely format it to redistribute it moreeffectively on another network.I’m sure there was other stuff I’ve missed from theoriginal post, but it can’t be helped.

    1. fredwilson

      Thanks for going through the hassle of reposting. And thanks for the link to your post

      1. John Best

        No worries, it’s something I find fascinating. Not only that, but I honestly believe that the filter/barrier/segmentation concept is going to be a key thing to understanding relevance and context.

  30. Kasi Viswanathan Agilandam

    It is very obvious about dispersion… thing that gets dispersed should have sufficient stuff to start with. So a message or content gets dispersed when it is very strong.Entropy is more about dis-order than about order. What is the connection between social-media and entropy i can’t understand. There is this “social-entropy” which may somewhat fit in to the theory of thermodynamics. But what is this social-media-entropy ?? The entropy thingy itself is a little confusing stuff. Probably people are using it on social media because they are already confused about what is this social media all about… so what u do  …. you add another confusing word to confuse the confused social-media stuff. 

    1. fredwilson

      confusion for the win!

  31. Richard

    There is a useful time series statistical technique that can test the trailing vs leading indicator issue. It is known as a granger causality test. There is an ample supply of stats/quant undergrad and grad students (I’ve been studying in this area) hungry for data and who are looking for a real world hunch to test. I’ve been toying with the concept of a quora type community for data analysis that helps bridge the gap between those with the data looking for answers and those with techniques looking for applications. Thoughts?

    1. fredwilson

      There’s a service called Kaggle that is kind of like that

      1. Richard

        thanks, yes  there are several competition type sites. They all seem to formal though. I want to take a less formal more social approach, an iterative approach with input from across many fields of study (marketing, business, stats etc.).  It is  geared for university students and has a special sauce for motivating participation (networking).

  32. Seth Godin

    Fred, there’s something big here, and I hope you’ll write on it again.I’m confused about something though, which is the distinction between a medium (a channel of communication) and the diversity in the way people use it.For example, the telephone is used for conversations between best friends, between drug dealers, between telemarketers and their victims, etc. But no one talks about a “phone strategy” as if there’s a monolithic way the phone is used.I think the same is true with Twitter or Facebook. YES, volume on facebook might very well be a trailing indicator of popularity, but what about diversity within the likes, diversity within the comments? If 10 people from 100 social circles end up getting involved, that’s significantly different from 1000 people in the same circle, no?In other words, I’m not sure a diversity of channels is what she’s talking about. I think it’s a diversity of micro-audiences, which may or may not end up using the very same channel. Other than from a traffic/investor point of view, then, what matters isn’t the medium, it’s the messengers and their tribes.

    1. fredwilson

      I have to keep working on this Seth. This is a half baked notion that jumped into my head when I was talking to Dina’s class. It needs more work. Which is why I blogged it!

    2. Luke Chamberlin

      Replace “phone” with email or print and you do hear people talking about those kinds of strategies. An email strategy is different from a print strategy regardless of the target audience.In the end the audience and message are what’s important, but from a strategic point of view you plan a TV campaign or Twitter campaign in different ways. Or maybe I’m confusing strategy with tactics?

    3. Mark Essel

      And those micro audiences or social clusters are dynamic so tracking them is a tough problem. Tough unless you’re Hunch which already clusters folks by interests in a variety of orthogonal areas.

    4. Kasi Viswanathan Agilandam

      when social entropy matures it gets into anarchy …. messenger and tribe…

    5. Boris Wertz

      Except that specialized channels are better suited for the conversations of those micro-audiences as those platforms are public mediums (not private ones like the telephone) – so the public nature of the conversation should lead to a diversity of channels.

    6. testtest

      each medium has a gestalt. the one side, or few sides, of identity which are visible on a given platform is influenced heavily by the medium.whilst micro-audiences may be from different groups the gestalt shines through. the medium and the culture it creates affects the message.imho 

    7. propjoe

      Seth, you are right, entropy is the measurement of how dispersed the volume is across different “types of people” with different interests. I suspected this was what she was talking about when I first read this article, and after scanning through her model in the paper, this seems to be the case.The underlying diversity isn’t about the platforms but about the people. They measured this diversity in one platform, Usenet, across different newsgroups which represented different groups of people with different interests. One can argue that different platforms also represent different groups of people, and this maybe true for niche platforms. But since Facebook has representatives from perhaps all different corners of the earth (probably much more so than Usenet), measurement of dispersion across social groups in Facebook should work as a leading indicator of success.There are a couple of caveats, I think. First, different people may use different platforms differently, so even though I am on Facebook, Twitter, and Google+, I tend to use Twitter to talk about entertainment stuff. So if people in my “social group” and some other groups disproportionately use Twitter for this task, an entropy measurement on Facebook will not be accurate. Then, there’s a case to tie different social networks together (a tough task) and measure the dispersion of the total volume across these networks.Second, a more interesting problem is how to separate/define the “social groups” across which we need to measure the volume dispersion. Dina used newsgroups as a rough measure of diversity, what’s the equivalence today? Lists? Fan Pages? Or can we do even better because we have a much more nuanced understanding of each user than 10 years ago. Grouping by age, socioeconomic, education, location, industry? This may create even better grouping and better measurement of entropy and better prediction of future success. Someone should do this. I can see a company here 🙂

    8. jason wright

      “…it’s the messengers and their tribes.”The medium is the transmission channel (the transmission medium) of the messenger/ message to the ‘tribe’. The bongo drums aren’t the message, but when they ‘play’ the tribe listens. Here everyone has bongo drums. 



      1. fredwilson

        pincus was a decade ahead of himself with tribe.net

        1. FAKE GRIMLOCK


    10. Prokofy

       The reason nobody talks about “a phone strategy” is because no company would pay anybody to talk about a “phone strategy” because they know it would be patently ridiculous. You get on the phone, you make cold calls to the leads, end of story.The reason people talk about a “Facebook strategy” or a “Twitter strategy” is because it is still new enough that you can still scare people with it and make them pay you to understand what it is.But it boils down to the same thing, where they have to go on Facebook and Twitter and make cold walls using leads.

      1. SubstrateUndertow

        Are you saying that a social media strategy is by definition limited only to commercial endeavours  ?

    11. leigh

      I think the confusion lies in the desire to create a distinction between the medium and the way people use it.  They are interdependent are they not?  The dynamics of the telephone directly impacts the diversity in the way people use it.I think what’s missing is what she means by ‘wider’ and ‘broader’ — the context of how diversity is being defined —  And i’m hoping the net net isn’t that the medium doesn’t matter bc i just fundamentally disagree with that so completely and I think there is also something to the notion of entropy ……

  33. aweissman

    Most interesting to me is that this fragmentation/segmentation reflects human behavior.  People like lots of things in lots of different places in lots of different ways.  We tried to do the opposite at AOL – everything under one roof, organized by channels (the “portal”).  It didn’t work.   Google came along and perfected a service that reflected how users, people, want to browse and consume – all over the place.

  34. JamesHRH

    Fred, you are developing your case using basic branding theory.Consumer internet is a brand marketplace. Traditional development chain for any consumer brand requires you to connect 1-2-3:1) core attribute / positioning2) concept / brand promise3) core usage scenarios / required UX performanceSo, just like the toothpaste aisle – the hoariest of branding examples – there is going to be 7 or 8 monster social mediums and lots of big ones.Pinterest is a great example of a service that knows its core customer exceedingly well and delivers on its core attribute (basically, easy sharing of visual artifacts),You could argue that Pinterest is to Tumblr as Facebook is to MySpace (which, I was fairly shocked to learn, is in the top 5 social media sights, according to recent post by Om).So, you will end up buying ads across platforms, just like big ad budgets bought across media categories.The real shake out is coming from the buy side – they might be big and slow, but they know who has the money.

    1. fredwilson

      Pinterest is like Tumblr in some ways but very different in others. I think they will coexist and compliment each other more than compete

      1. JamesHRH

        Totally agree – customer overlap is v close to zero.

        1. fredwilson

          bullshit. i follow 142 tumblogsvery few, if any of them, are on pinterest

          1. JamesHRH

            Misunderstanding.I am saying the same thing, that close to 0% of people who tumbl also Pinterest.

          2. fredwilson

            ah. then i take my heated reply back!

          3. JamesHRH

            @fredwilson:disqus interesting to see Kate’s statement that she could ee crossover between the two, as they mature.They brand purist in me doesn’t like that idea, but, given the core ‘visual / interest / sharing dynamic’, maybe it has legs.I tried to calculate the time of day that you posted and thought – that’s BEFORE 5 AM, he mis-read something……….I’m all for being heated every now and then, otherwise you don’t care enough! 

      2. Kate Greer

        I would love to hear how you think that will be the case.  It’s something I’ve been thinking about a lot as I, a long-time Tumblr user, explore Pinterest, which has consumed my childhood friends from the Midwest.Ps excuse my attached photo. I’m (obviously) a first time poster!

        1. fredwilson

          welcome Kate!

        2. JamesHRH

          Hi, Kate – I can tell you why I think that is so.Your friends did not leave the Midwest, for one. This may seem silly, but most great brands are centred on the customer, not the product.So, the daily routines, exposures and even shared aspirations of people are often the foundation for a great brand.People who move from the middle of NA to the edges are actually quite different from people who don’t. They left home, family, old friends and known routines for none of the above: work, adventure, love.This is just one slice of it: the psychographic slice. Other things like demographic, economic, cultural focal points can be used as an axis to build a great brand. I am sure some other folks may chime in – my list is far from exhaustive.For me, and I am only a limited Tumblr user (not an insider, which @fredwilson:disqus is), Tumbrl is more urban and arts / creativity oriented. It is geared towards getting exposed to entirely new experiences.Pinterest is geared towards a suburban mindset – and for lack of a better term, arts / crafts oriented – the exposure is more variations on a theme.Its easy to dismiss Pinterest as a place to swap recipes, but if you like to cook and have responsibility to cook for a family of 4, 5 or 6, variations on a theme are pretty important – especially if they are time savers and routine changers.its also a huge market.To be transparent, I am still awaiting my invite to Pinterest, but have spoken to many people who are on the service and done a little digging around. I am not an expert on the service and just giving you a guess on the brand experience, based on reporting I have at the moment.

          1. Kate Greer

            Thanks for the response James.  I like your pyschographic take and I agree that it plays a huge part in who makes up the early adopters of each.  Also, in the differences of content on each.  I’ve had the chance to play around on Pinterest a little.  Although, I prefer the content my Tumblr followers post (and that, for the most part, they create  – a huge differentiator), I find the boards on Pinterest to be really efficient for getting to what I’m interested in learning more about.To @fredwilson:disqus ‘s point (and thanks for the welcome!), Pinterest has generated a significant amount of traffic to my Tumblr. Not from recipes or because I’m originally from the Midwest, but because we are all interested in what it’s like outside our lives… at least from what I can tell.  After my first hour on Pinterest, I sent an email to my photographer and artist friends recommending that they explore Pinterest. When I search “slam-dunk,” my sports photographer friend’s great content should come up. What a fantastic way to make your name synonymous with good sports photography!I’m wondering/excited for how Tumblr and Pinterest will interact once its dispersed past the early adopters. 

  35. Carl J. Mistlebauer

    I cannot speak of anything other than the results we generated from our own social media campaign over the course of the last five years.  Again, we were attempting to “engage” (I think in any conversation of social media this word needs to be defined) a particular consumer.  I defined “engage” two ways:1.  Get consumer email address and or address.2.  Get consumer to order.We experimented by running campaigns on just one site, different campaigns on different sites, same campaigns on one or two sites, campaigns on social media along with email campaigns, direct mail and social media campaigns, oh, and we had a blog.  We found that consumers were more likely to order if they ran into a “mention” (either an ad or a mention of some sort) across more than one social media venue.Thus, I came to believe that to be successful at engaging consumers you have to promote your product or service following the philosophy of Phil Spector’s “Wall Of Sound” (http://en.wikipedia.org/wik…One final note, we found that putting a landing page between whatever we did and our website was the best way to seal the deal.  If you only have 2 or 3 seconds to make your point, its best to make the point as easily and simply as possible.  This also lets you tailor your message better.I imagine Arnold is shaking his head wondering, “Wall of Sound?”, remember I am not a marketing or advertising expert….all I do is attempt to understand by experimenting and testing….

    1. fredwilson

      The “wall of sound” theory. It has a name now!

      1. Guest

        Of course Phil Spector gets a royalty for that ;)http://en.wikipedia.org/wik…

    2. Cynthia Schames

      Wall of Sound technique = brilliant name. 

    3. kirklove

      Of course Phil Spector gets a royalty for that ;)(meant for this to go under your comment Carl). Great point. Silly Disqus.

      1. jason wright

        Not that he can spend it.

  36. Dina Mayzlin

    First, thank you Fred for your post.  I am very flattered that my research with David Godes (http://bit.ly/wEDVum) overlapped so well with your own ideas.  I have been a Fred Wilson fan for a long time, so it was an amazing experience for me (and for my students) to have you come to our class.  I also am really enjoying reading the comments on your post.I wanted to share some resources with your and your readers:1) The paper is available at http://bit.ly/wO2KFs 2) Here is a video where I talk about the paper http://bit.ly/AkB2Hn            

    1. fredwilson

      Thanks for sharing those links Dina and thanks for having me in your class

    2. testtest

      fascinating; thanks

    3. ShanaC

      Thank you for linking to the paper, Dr!

    4. leigh

      Tnx for the link.  Lots of goodness in there.Interesting talking about ethnical questions of community — if you ever end up doing work on this, I have an exceptional amount of examples that you might find interesting.

  37. Bruce Warila

    I agree that dispersion is a solid leading indicator.  What I don’t really buy is the notion that dispersion can be individually accelerated or controlled.  For example, examine the top 1,000 people that have the most followers on Twitter (each has more than 500,000 followers), there’s a lot of people on that list that can’t even move the needle on YouTube views when releasing a new song or video.  What I am suggesting is that for creators, investing a bunch of time into a myriad of platforms is nearly pointless.Dispersion is something organic that happens to remarkable content (fans do it for you http://j.mp/cJzJyl).  Mass-market exposure and strategic promotion accelerate dispersion, but the notion that a single person can light enough sparks to start a massive brush fire is social media bunk that rarely happens.

  38. mattellsworth

    Can’t believe you left Turntable out of the music business conversation.  From a social perspective, more artists hanging out in Turntable connecting with their fans would be really powerful.

    1. fredwilson

      That was a mistake. I will make it up to Billy

  39. Alex L

    Couldn’t agree more with this. But I do have a question I’d like both Fred and the community to answer. If we need a different site for each type of network, why did Ning fail so badly and why hasn’t a successful Ning replacement come out yet? I have my own answer (and solution) but I’d like to hear from you first 🙂

    1. fredwilson

      they went all the way to the other side. they went too far

  40. Ashwin

    Fred,If Facebook successfully achieves what it set out to do at F8 last year, that is to have all other conversations, engagements, interactions between people , between brands and people,  eventually end up on FB,  why split the attention? 

    1. fredwilson

      because the only way to get engagement is to engage on the platform where the conversation is happening

  41. David Landsman

    If you watch Deb Roy’s astounding TED talk “The Birth of a Word” http://www.ted.com/talks/de… you will see something amazing. During the back half of the video, Deb & his team at MIT more or less developed a real time measuring stick for the social graph. (and it seems accidentally) The application of his discovery as a leading indicator or a trailing one remains to be seen, but it is truly amazing and extremely relevant to today’s discussion.   

    1. jason wright

      David, a really great link – ‘Link of the Week’ no doubt.   

    2. fredwilson

      thanks. fantastic link!

  42. gleslie

    A very thought provoking post. This research is very exciting because I think it will give a lot of new insight into pre-social media social marketing theory.”The number of likes on Facebook or tweets on Twitter is volume and is likely to be a trailing indicator of popularity. But if you track the essential social gestures across the fragmenting landscape of social platforms, likes, tweets, tumbls, checkins, pins, etc”The challenge here is accounting for the variable that each user (and demographic of users) engages with social media to a different degree. My mother for example, might hear about a movie she wants to see on facebook, but rather than “like”, “comment”, or “tweet” about it, she will pick up the phone and call her closest friends to let them know that she wants to see it. She doesn’t get the same reaction as I would if I posted “Want to see X movie this weekend.” This is essentially social media interaction that won’t show up in the data. Depending on a user’s level of engagement, they will take the dispersion process off the grid.

  43. Noah Waisberg

    Don’t forget that there is a distinction between indicators and drivers. So Professor Mayzlin may have found that dispersion is a better indicator of popularity than mentions, but it is not the same as finding that hitting multiple platforms drives popularity–maybe the more popular shows were just more entertaining. Still, interesting work and an interesting post. Though you may wish to develop that causal link more.

  44. gleslie

    “I am totally convinced that the world of social media is not consolidating around one “winner takes all” social platform.”Couldn’t agree with this more. It’s relatively intuitive and I think we’re already seeing it with facebook. You simply can’t offer a product that is everything to everyone. Facebook has become ‘something’ to nearly everyone, but have you ever seen one of those posts that say “I hate when people have long conversations on facebook! Clogging up my newsfeed!” Enter disqus. Social networks inevitably imply some kind of social compact, a consensus on how the service should be used. Any time a user breaks that, an opportunity for another network emerges.

    1. gleslie

      Spotify app is breaking it… I see frequent, “I don’t care what people are listening to on spotify” posts.

      1. fredwilson

        spotify in 2012 is what zynga was in 2008spam

      2. jason wright

        Spotify used to be good back in the good old days, but now it permits only a limited number of free plays for each song and then it’s not possible to listen again without paying – I’m cheap. 

  45. Alexis Madrigal

    I couldn’t agree more with you, Fred. In fact, I wrote up my own version of this with the niche in the ecosystem metaphor last month: http://theatln.tc/t0NRiT The takeaway is that there may not be a “next Facebook.”One dynamic that I didn’t get to explore in that post was the way that these social networks build atop each other’s infrastructures. Having the base presence of Facebook let’s Twitter get away with minimal profile pages. Having the networks created on Twitter lets other networks leverage that to jumpstart their own products. And on and on, backwards and forwards in time.



      1. jason wright

        I expect a retreat from FB la la land as FBers grow up and move on to more rewarding and socio economically engaging fragmented networks. The next generation will skip the FB stage entirely. 

        1. FAKE GRIMLOCK


          1. jason wright

            I’m sure there’s a very strict regime inside the inner circle of FB about what kinds of analysis and statistical projections get written down in official company notation and what kinds do not – shredding may not be necessary.I have a wild hunch that at some point in the future there will be a  shareholder class action against FB and GS for financial misconduct.

          2. autom

            an intriguing, wild hunch indeed

          3. another cultural landslide

            The real question about Facebook is not how many users they have – it’s how many Active users they have. Anecdotally, Just looking at Wendy’s “friend” list of 200 users reveals only 20 users that post at least once every three days; another 7 that have posted in the last 10 days; and another 6 that have posted within the last month. And Wen says her FB activity has been dropping steadily since late November.Thing is, a year ago it was much, much more active.My gut is telling me to stick with my initial impression: Facebook today is what AOL was 12 years ago.And look at where AOL is now.FG is dead-on-balls-accurate: Facebook is only a tool. And old tools get replaced by newer tools… & on the net, that happens very, very quickly.The one sadness I feel is for all the suckers who buy into Facebook’s IPO. That money train left the station a long time ago.

        2. Douglas Crets

          I really doubt that is going to happen. Maybe we should place a wager for 2020.  

          1. jason wright

            I don’t gamble. I have 2020… vision :-)FB is like primordial social soup. Many thinks will evolve out of it, but FB will always be… soup.

  46. William Mougayar

    Dina has posted her article relating to this post:http://faculty.som.yale.edu…  I should add…it’s not for the faint of heart, reading-wise.

    1. LE

      Study (as Fred mentioned) dates back 10 years. The conclusions I have to say make intuitive sense and certainly seem correct.But I’m wary of a document so complex that you can take data and essentially lead people to any conclusion you want.  Just one scan of it picks up this “Our WOMdata are drawn from Usenet newsgroups.” How representative of society is Usenet? I’ll leave the formulas and statistics to @sigmaalegbra:twitter  to vet.For anyone not wanting to read the complete paper the following is relevant:While we have taken an important first step in several directions, we acknowledge that our approach is burdened with several limitations. We have focused on a single product category, TV shows. While we believe the results to be relatively general, it would be important to replicate these results in other categories characterized by different types of consumer actions.The decision to watch a new TV show is a relatively low-cost and low-risk decision. It would be interesting  to investigate the role of WOMon the adoption of new technologies or the purchase of higher ticket items, for example. It would also be important to identify the underlying category factors that make dispersion more important than volume or the decline in the effect of WOM to be particularly steep. This would have an impact on both measurement and management strategies. Econometrically, our approach leaves open the question of sample selection bias. One benefit of the truncated sample approach we focus on is that it minimizes the potential for such a problem because most— though not all—shows survive at least four or five episodes. Our investigation of dynamics that uses all the data is potentially prone to sample selection bias. Finally, we have not been able to control for potential important factors in the model. For example, we cannot rule out that at least some of the WOM or ratings we observe may be generated because of advertising or positive critical acclaim. To demonstrate causality between WOM and subsequent sales, future research will either need to include advertising data or to control for such exogenous factors in other ways.



      2. fredwilson

        the callout to siggy is funny

    2. fredwilson

      with academics, it never iscarlota perez and yochai benkler are proof points. they are must reads but they are so damn painful.now clay christensen is the exception to the rule. i love the way he writes

  47. kidmercury

    fragmentation is where it’s at, although for propaganda reasons i prefer the word niche. fragmentation gets the apple slaves i mean fanboys all up in arms demanding their fascist world government. so i’ll use niche, although the ideas are very similar. once you go niche a few things happen:1. depth of wallet instead of breadth of wallet: meaning business models that capture $1,000 from 10 customers are more viable and better than businesses that capture $10 from 1,000 customers. this is a huge point with many implications and makes most of the current hyped stuff unviable for those looking out years in advance.  2. multiple internets, alternate DNS, niche devices….this is where apple actually kinda sorta has the right idea, if only they were to focus a bit more, and understood how to use open source production to reduce costs. amazon has the right idea here. it’s bezos’ world, most people just haven’t figured that out yet. of course the real story here is that niche stuff has a cap on the profit potential. that’s why pouring 2+ billion into facebook is such a poor idea, now they have to take over the world to even dream of making a good enough return on that. same with twitter, tumblr, airbnb, foursquare, and the rest of the kids in bubble 2.0. creating a federation of niche things is the investment opportunity. anyway, explore this thinking long enough, and you’ll end up wearing 9/11 truth t-shirts talking about the end of the nation-state. be careful. it could happen to you.

    1. Aaron Klein

      On point 1, why is that model “better” per se?Great businesses have customers all the way across that value chain.But the businesses with real staying power don’t lose 30% of their revenue if they lose only 3 customers.

      1. kidmercury

        as for why depth beats breadth, think of why we’re having this conversation here on fred’s blog. i mean we could have it on facebook, right? but here on fred’s blog we get maximum grade (quality of discussion) and optimum density (concentration). to have this discussion on facebook we’d have lower grade and lower density, because it’s a more diverse community.accordingly, avc has earned our trust, so when we make our biggest purchases in this sector, we’ll be more inclined to come to avc.   and this principle constnatly applies. yes, we coudl buy shoes on facebook, but why not go to our local shoe community? real estate? video games? etc. meanwhile, facebook bears the cost of supporting an infrastructure for many people.twitter has the same situation, i’d argue even worse. they are all the same, facebook is actually the best of the one worlders.   as for businesses with few customers, think of fred’s business, union square ventures. if USV evolves to a ycombo type model, they’ll have at most a 1,000 companies in their portfolio. yet it will still be a lot better than many businesses with a much greater number of customers. that type of economics is what will drive the web going forward in my opinion. 

        1. Aaron Klein

          I think we’re talking about two different things.We completely agree about depth of community. I’m just not sure that translates to business models and the quantity vs. size of your customer mix.At Riskalyze, we’re going to have a number of institutional customers who pay us large amounts of money, and a much larger number of users from whom we earn small amounts of money.I have the next ten years mapped out in my head and I think that mix of revenue models is going to allow us to build an incredibly sustainable, defensible and profitable business that stays independent and focused on the long term.

    2. CJ

      “creating a federation of niche things is the investment opportunity.”  – Well said!



        1. fredwilson

          that’s taking it too far

          1. FAKE GRIMLOCK


    3. ShanaC

      Depth of wallet creates way too many issues long run – lose a customer, lose your business. It is good to be a bit diverse

      1. kidmercury

        diversify too much, though, and you lose all your customers to more focused businesses.

        1. ShanaC

          It depends on the kind of business. – posted via Engagio

        1. kidmercury

          as usual, people over-reacting. let’s see how much damage this actually costs. is it going to be perfect? no. is it going to cause some damage? yes. it is it going to cause less damage per unit of energy generated than anything else? yes. by this definition, is it better for the environment and more economically viable? yes. the facts are the facts, which is why barring some dramatic breakthrough or massive change in human behavior currently not in sight, nuclear power is inevitable.but i’ve conceded its imperfections, and hope that regulation and innovation can minimize the downside. now tell me what’s better for a world with 7+ billion people living increasingly in cities…..

          1. jason wright

            Nuclear is fine…but off the planet please :-). We don’t need to replicate nuclear fission… here. It’s being done for us 90 million miles away. We just need to capture it…here. It’s not rocket surgery.

          2. kidmercury

            oh but it is rocket science! they’ve been trying to capture solar energy for a long, long, long time. tons of money, from the private sector and governments, and lots of smart people have tried figuring this out. maybe they’ll figure it out some day, although i don’t see any progress aside from dinky little things like solar powered laptops and e-readers (which is great and very useful, but we live in cities and have greater needs beyond that). in the mean time, china, india, korea, france, and even japan and the US are moving more to nuclear. because it’s either nuclear power or blackouts. that’s what germany faces if they don’t wake up and start getting down with nuclear.

          3. jason wright

            In reality your rocket science hits the target, the ICBM target. Imagine a planet stripped of its civil nuclear reactors, which of course are the fig leaves hiding the truth, military nuclear reactors producing weapons grade output with only one possible use. Populations would revolt, governments would fall, orders would become disordered, and MAD would be replaced by sanity.

          4. kidmercury

            your comments, while i don’t agree with them, are besides the point. it boils down to this: do you want the world to have affordable electricity or not? if so, nuclear is the answer (natural gas gets an honorable mention, and oil too if we fracking can make oil drilling much cheaper as it has for natural gas). as for all the war stuff let us not blame the technology for the moral failings of humanity. they’ll use solar to create war if they need to (US military is already largest consumer of solar power). and then they’ll try to block the sun from getting to their opponents. so should we ban solar technology? no, obviously that’s silly. the same applies for nuclear. at the end of the day this boils down to physics and economics. nothing beats nuclear in generating electricity en masse inexpensively for the whole world. it’s very simple. and the internet community as a whole should rally around nuclear, because the energy disruption has to come first in order for lots of cool internet stuff to really kick in across the world.

  48. Irving Fain

    Fantastic post Fred! With the understanding that this is still very much a work in progress, I think one of the critical factors that has to be considered is not just the breadth of engagement, but also the depth of engagement. As many people say, all “likes” aren’t created equal nor are all “followers” equal. Some clicked the button ages ago and pay no attention while others are heavily engaged.My assumption (untested of course) would be that breadth along with depth combined would be a true measure of entropy as people are more deeply engaged with subjects they care about and have passion for.



      1. William Mougayar

        ENGAGEMENT > ACTION too. 

        1. FAKE GRIMLOCK


    2. fredwilson


  49. J Nicholas Gross

    There is lots of research on word of mouth as the mechanism for spreading ideas, including product adoptions. In general though the mechanism is opposite to what you might suspect: the likely success of a product in a population for example is better measured by the appearance of pockets of rapid adoption in small concentrated areas, rather than the existence of dispersed or diffuse areas throughout the population.  The former is attributable to word of mouth (viral) effects, while the latter is usually due to traditional advertising/promotion. So in terms of future success, it is better to see massive adoption in a small area in a short time as opposed to a less concentrated result across the spectrum as a result of broader marketing.You can see some representative research herehttp://w4.stern.nyu.edu…

  50. Jeff Mason

    How does Klout relate to your idea? Is it measuring something different than dispersion?

    1. fredwilson

      it is measuring popularity

  51. ErikSchwartz

    Volume is easy to manipulate, dispersion is much harder to manipulate.

    1. fredwilson


  52. sigmaalgebra

    Yes, “Sometimes we find that a good question is more important than a good answer.” (R. Bellman). Or, in original work, an early challenge is the problem formulation, i.e., what problem are we trying to solve? And it has long been recognized that one of the keys to eventual good results from original work is initial good problem selection.Fragmentation? Quite broadly, fragmentation — that is, cases of specialization, customization, partitioning — is enormously powerful. E.g., at least by analogy, recall the Riemann integral of classic calculus: For curves of quite high generality (actually, as shown in W. Rudin’s ‘Principles’, continuous everywhere except on a set of area zero), can get the area under the curve from partitioning the variable, making the partitions small and in each partition just using a constant with no additional effort to follow the curve. And there are other, more powerful arguments for partitioning.Or, no single partition accounts for much of the area, but in total the partitioning grabs all the area.It appears that considering TV show “popularity” and social media “fragmentation” together is mixing apples and oranges: A TV show had little choice but to pursue the goal of popularity, but an instance of fragmented media needs only do well in its selected partition of the whole audience. So, in principle and maybe significantly in practice, fragmented media in total can dominate and take all the Internet audience even if no one instance is very “popular”. Moreover, with fragmented solutions taking nearly all the audience, an unfragmented solution has essentially no chance.If the content is so highly fragmented, then the search function needs to ‘get at’ both the ‘meaning’ of the fragmented instances and also, to have information on the desired ‘meaning’, do well on ‘personalization’ for each user. How to do such things? Hmm ….There is the issue of “social platform”, e.g., Disqus, Hacker News, SAI BI comments, Tumblr, eGullet.org, and “a certain kind of social engagement”, e.g., AVC.com on Disqus. meetups via Hacker News, food media on eGullet.org. So the “social platform” is not the same as a “kind of social engagement”.Yes, we need to understand better how to fragment, specialize, customize, partition, etc., what people want in social media, the platforms, and the kinds of engagement.To so understand, I suspect we have to return to old observations about human concerns and wants: An old, standard observation is that the central problem in life is addressing the anxiety from the realization that alone we are vulnerable to the hostile forces of nature and society and that and for this anxiety three of the standard wants are membership in groups, love of spouse, and love of God (quoted loosely from E. Fromm). Of these three, the most relevant for Internet social media is likely membership in groups, and there people may seek connections, praise, acceptance, and approval. Might also consider the Maslow hierarchy, etc.My guess at the next step would be to pursue the interest graph instead of just the social graph. That is, for people, interests have quite fundamental importance and promise to continue to.So, for a platform, want it to let people pursue their interests. Then have some platform features that facilitate wanted cases of engagement. So, in principle there is a chance for one platform for many interests.

  53. Marc Brodeur

    From a branding perspective, historically, things never converge. Each brand/group/audience creates their own brand niche and brands diverge.Just look around now: Beer/food/restaurants. Consumer products. Blogs and communities. Cars/airlines/travel.All sorts of brands and categories splinter as niches break off.So I agree with the post 🙂

  54. rohit_mod

    interesting point to ponder – higher entropy = good as a starting condition – then must induce axis of crystallization for seeding/growing popularity curves. high disorder alone wont suffice and measurements of slope will be misleading – need to measure second derivative of the slope.

  55. Dave W Baldwin

    Noticed the mentioning of e-mail as a network, I would say a category reflecting Collaborative Networks/Platforms need to be added. The bigger fish to catch is working from other side of all the Social Platforms where it is easier for individuals to slide into several macro/micro options in the easiest, yet controlled fashion.

  56. Peter G

    Dina’s research holds true with our experience, but interestingly, we find a lot of companies are afraid of dispersion or entropy. We regularly encounter clients who are at first excited by social product distribution, but then later ask “Can I control which sites and platforms where this is shared and who shares it?” It seems that a lot of companies seek comfort from a tunnel vision focus on their social media strategy. They want to know all about either Twitter Followers or Facebook Likes, but when we look at our success statistics, we see source traffic spread across a range of social platforms, blogs, etc. Metaphorically, the conversation inevitably spills out of the bar, into the street, and into the next neighbourhood.In the social web, Dispersion/Entropy are chaotic and completely unavoidable for a successful campaign, but that success-yielding chaos is also scary for companies that try to over-manage social engagement for their products. I’ve been wondering why online word-of-mouth scares so many executives, since traditional word-of-mouth is similarly chaotic. I suppose it may be persistence. While chatter on the streets is fleeting, social communication online is perpetually discoverable. I’m curious if others experience similar reticence to the chaotic and dispersive nature of social media and if any other reasons come up?

    1. fredwilson

      chaos is what you wantwe found that in the stop sopa movementit was totally uncoordinatedand it worked

      1. Peter G

        Definitely, chaos is what you want. Which is why it surprises me every time a client asks us to remove the chaos from our viral movie distribution system. Every hit we have has its own unique uncoordinated success pattern.Our experience is that if you try to stay out of the chaos, the audience will find their own way, and in our industry, that usually means piracy (as you pointed out a few weeks ago in your article about scarcity). The film industry model of scarcity and windowed releases is an effort to fight the chaos they should be embracing…

  57. Conrad Ross Schulman

    Top tier VC’s are always interested in web companies who track user engagement and display the level of engagement to the public to show status of that brand. 

  58. AVillageOfHockey

    I think of it as ‘point of feeling’. When I feel a need to access friends, I open facebook. I feel a need to access interesting reading material, i open twitter. Something funny? YoutubeThe instant a thought comes into my head, I have one or another channel to express it online and those channels are being taken up one by one. That is partly why it’s hard for one company to do multiple channels because I associate a name with a feeling and I want instant access (not 2 steps)Location based services have not matured because when I ‘want to go out’ I can’t think of a single company that fits that feeling

  59. Otto

    J’ai le cœur Dina.I heart Dina.

  60. matthughes

    Fascinating stuff.Can’t wait to see the talk at some point. 

  61. testtest

    Some thoughts on social media marketing:– test all channels. allocate most of your resources to the most effective channel(s). there’s pareto like effects– look for emergent behaviour. there’s little to compensate for thinking for yourself– the most effective things will be what no one else, or few, are doing. in a sense it’s zero-sum– keep an eye on analytics. it’s your eyes and eyes on the internet– always be testing– cross-promote when you can– if someone takes an action ask them to take another. there’s compliance momentum– there’s always something you can do that’s extremely effective. always. magnitudes better than anything else. you need to figure it out– if something doesn’t work try it again. the first time may be an outlier– differentiation is inherently viral– fuck freetards– don’t mess around with volume. if you haven’t got it pay for it. allowing you to test– the worst reaction is indifference– if you know you’ve done the marketing right it may be your product, market, or business model (change if necessary) — the low barrier to entry on social networks creates saturation — social media net-net is an inefficient marketing platform. unless you’re at serious scale. migrating users onto your own platform is more effective– learning to write copy gives you an edge. Do it NOW!   Some thoughts on the fragmented market:– the internet is inherently fragmented. from the protocols all the way up the stake. it’s fractal– with the inherent fragmentation identity is also fragmented– the value online is with extropy– platforms are the best way to capture value with fragmentation– web first, mobile second, is more effective in capturing users when you’re small. due to the fragmentation on the web and the consolidation on the app-internet– technological advances drive the fragmentation

    1. ShanaC

      Thank god I can write copy then :-pSeriously, it sounds like beyond coding, we need to teach people communication skills in order to handle the fragmentation.

      1. testtest

        yep. and i think communication skills are important just to get in the game.and coding: for startups.in the world of information workers the bar is rising. but we’ve got access to a ton of knowledge now, so that’s all good.i also think time is becoming more valuable, as it’s more easily turned into intellectual capital  

        1. ShanaC

          Very true about time. I often wonder about the relationship with that bar rising to time. If it takes more time to gain knowledge while the bar keeps going up, then what?- posted via Engagio

    2. fredwilson

      seth godin taught me this almost 20 years ago”test, measure, test, measure, test, measure”rinserepeat

  62. Adam Schwartz

    Put simply, width matters more than length.



      1. kidmercury


  63. Kyle Comeau

    Pinterest is a monster. If my newsfeed is any indication, Facebook has been reduced to a place where girls go to comment about how addicted they are to Pinterest.

    1. fredwilson


  64. Jon Lawrence

    Great observation.  The most interesting question to me is how to create service/platforms that can quickly and easily measure these interactions about the media and turn them into meaningful and sustainable conversations that enable content creators and audiences to forge better connections.We think that’s a question worth pursuing with great vigor and passion:)

  65. saumilnanavati

    Volume: The definition of “popular” is rather vague. It really depends on target audience and the niche (if at all) its trying to target. Take this scenario: say the movie Social Network was watched by 99% of techies but < 1% of general public-would this be considered less “popular” vs say Mission Impossible that was watched 20% of entire US audience?Winner Takes All: Niche/Tribalism/Specialization is a natural (final) evolutionary phase in every open environment. We have banks specialize in FS industry, law firms do the same and consumers tend to choose and engage based on best fit for their needs. Therefore, any new vertical created on digital platform should see similar evolutionary path, including social media. Its imp to choose the right (social) platform for the audience one is try ing to harvest.IMO, capital liquidity speeds up the evolutionary pace on digital platforms and specialization or in some cases are being “managed” by incumbents thru acquisitions.



  66. leigh

    Network thinking is not something that comes naturally to some people.  We work in old world structures, old world media models and old world marketing approaches.Now we are asked to think like a network – approach everything with an ecosystem approach – and what do we do?Look at volume, reach, and try to use old measures to look at this new way of approaching the world.The fact that she’s thinking in terms of entropy is very interesting.  Duncan Watts amazing work on big seed marketing comes to mind.  Would love to see how her research and his might intersect.  



      1. leigh

        The Internet has started to retrive tribe but I’m afraid my experience has been the undoing of hundreds of years of top down social, cultural and religious ways — well, that may take longer then we’d like……

        1. FAKE GRIMLOCK


    2. fredwilson

      funny. we talked about duncan watts. dina has read all of his work. thinks some of it is great. other not so much.

      1. leigh

        Be interested in which was which…..

  67. Jeremy Toeman

    love this post Fred, only comment is that I think if you are in TV you “need” to consider Facebook, nothing else is relevant yet… 

    1. fredwilson

      wrongthere are more tweets about a live TV show that facebook messages

  68. C-jones

    Very interesting post…it occurs to me that any analysis of dispersion or “Entropy” among social media platforms will be complicated by the fact that many users “link” their accounts, so a message will appear simultaneously on Twitter, Facebook, and Linked-in for instance…



  69. Adam Feldman

    Many people talking about one thing in many places > a few people screaming about it in one place.Dispersion is definitely a better indicator than volume. I feel like bit.ly or another link shortener would have good insight on how content spreads and is amplified through multiple networks.

  70. Jim Christie

    All communication is “social” independent of media but when I think about social in this context I would exclude email and chat, regardless of whether it is passed along. The distinction is that social media is an “activity” shared among many close and more importantly, not so close connections both in real time and as an asynchronous record. The word public is problematic. I don’t think public maximizes dispersion unless your network has the right “shape.” What is interesting about these discussions is how they return to concepts described in the real world in books like “the diffusion of innovations” and even “crossing the chasm.” These are more concerned with adoption rather than just informing but it is a given that in social networks everything is faster and steeper. Rodgers spoke about this when he mentioned connectors having the the greatest variety of connections which mirrored the sheer number of their connections as well. It would be my guess that the dispersion effect mentioned would most affect rate and #s of early informed and laggard informed but hard to figure out because it seems like the importance of the information would have a swamping effect on that of dispersion.

  71. Jennifer McFadden

    Uggh. I’m an SOM grad and live only 20 minutes away…had I known you were going to be there, I would have totally dropped in on the class! Hope you enjoyed the ‘Have.

    1. fredwilson

      did you take a class with Dina when you were at SOM?

      1. Jennifer McFadden

        Unfortunately, no. I was heavy on finance, not marketing. Also, I don’t think that there was even a social media class offered at the time (I graduated in 2008). Glad to hear that they are finally covering the topic–it’s a good sign. There are several of us who are recent grads and are interested in seeing the school incorporate more tech/entrepreneurship into the curriculum. It is still virtually non-existent. And, as I’ve argued to the Dean, technological innovation that leads to whole industry disruption is going to be THE management issue of the next 5-10 years. It has already occurred in the newspaper industry (my world) and we are bound to see other industries impacted as well (education included). The entrepreneurship program at Yale is largely nonexistent and the Dean’s strategy is to align more closely with the Yale Entrepreneurial Institute. YEI is great (I was one of the founders), but that is not the answer. There is no reason why Yale couldn’t create some sort of semester-long program similar to what is being offered at Harvard with Eric Ries. MBA programs are already essentially trade schools for finance–I don’t understand why they can’t be the same thing for start-ups. Instead of “Training the Street”, student’s could spend a weekend (or, couple of weekends) learning to code and build a product–even if it is lightweight. Outsource the curriculum to some org like General Assembly or Code Academy. It would be great practical experience. This is similar to what we are doing with the Entrepreneurial Journalism students at CUNY this year–reaching out to the tech community to give them the practical skills that they actually need to execute. It will be interesting to see how it all plays out.So, a bit off-topic, apologies. But, glad to hear that you enjoyed and that they are pushing things in this space a bit!

  72. pointsnfigures

    agree on the fragmentation.  was doing a blogpost instructing older people on how to use social media and came to yours.  I still think there will be dominant platforms.  Facebook, LinkedIn and Twitter aren’t going anywhere.  A lot of platforms use them to get users.  What would Zynga be without FB? What will be really interesting is as the fragmentation happens, I predict your online personality will meld with your real one.  You won’t be able to hide behind pseudonames etc.  

  73. M. Edward (Ed) Borasky

    I’d be willing to bet that some of the “big names” in social media monitoring and analytics already have this nailed down. In fact, it wouldn’t surprise me if HP Labs, Google and Microsoft already have “skunk works” projects of this nature.

  74. Douglas Crets

    I’ll put on my where is media going hat. I think we’re approaching a moment in media evolution where demographics are determined by what platform you are more than what city you live in or what race you are. I think traditional broadcast news media, for instance, has it all wrong. they peddle a watered down information, and what people really need is context. What better way to deliver context as news than to base the derivatino of the context from the platform you deliver it to?  Twitter news comes from Twitter, based on what people do in Twitter, which also has some basis in what happens in the real world.What will happen with TV news, I think, is that it will point to where news is happening, but won’t report it. It will basically be more like a billboard advertisement on the highway. 

    1. fredwilson

      we aren’t there yet but we may get there

    2. Carl Rahn Griffith

      TV is just a console – no longer a medium.

  75. William Mougayar

    Speaking of social fragmentation, it reminded me of this attached Infographic I did 6 months which just showed that, without calling it fragmentation. Look at the section called “Social gestures” for e.g.” . I posted it on Tumblr then http://wmougayar.tumblr.com…

  76. Youssef Rahoui

    Wether email is social or not (I think it is), I also think you can’t do without it if you plan a social media strategy. One must not also forget Google or Yahoo! Groups. It is not glamour, for sure, but it is popular and I know many valuable ones even in the tech industry. I also use them: it the easiest way to gather people online around a topic with simple privacy settings.People are not using tools for the sake of them or because they are fashionable, they use tools to get their job done.

    1. fredwilson

      i think email is so important that you need an email strategy and a social strategythey are not one and the same

  77. vruz

    Jeremiah Owyang predicted all these things back in 2007  (I know because I made some translations of it into other languages)”The Irrelevant Corporate Website”http://www.web-strategist.c… Video:http://www.youtube.com/watc… 

  78. Abbo

    This is a very interesting read, I may have to return to continue reading these comments. It would appear everyone is looking for a measurable yard stick and formula for how all the social networking can benefit everyone to be profitable in one easy post or thesis. I would like to mention that the speed things move, as soon as you have the answer you think is correct, things have moved on. I’m afraid patience, interaction honesty and adaptability seem to be the watchwords for success in this world of social interaction, and a firm understanding that No one has ultimate control over what may become successful/viral. It should be an enjoyable experience. companies should employ people who enjoy this format of communication, and not force it upon people to come up with results in the boardroom meetings.In my humble opinion.

  79. Prokofy

    I’m puzzled why she is using the term “Entropy” — doesn’t that mean something dying out? When the dispersion is something that continues to move in waves. While it may have an “end point” at some juncture, maybe it doesn’t? And keeps reviving over the life of the TV show, etc.I just ran across Pinterest yesterday and it seemed cool and all but fragmentation also means that the time for adapting to One More Thing just starts to run out. You didn’t mention Scoop.it which is my favourite tool right now curating news.I’m wondering if there is actually so much fragmentation for *the individual*. That is, sure, there are all these platforms in all these fields. But most people knit themselves together in one place. It might be their blog, or Facebook or Twitter. Foursquare became uninteresting to me after a day. But I’m fascinated to see how *religiously* all the people in the Russian protest movements now are using it, and using it to keep up with each other and track if people are jailed or lost etc. It’s one thing if you are merely checking into a restaurant in NYC where you have no hope of being “mayor”; it’s another if you are checking into the MVD (police) building in Moscow and they’re not letting you check out…Don’t you think most people use a cluster of these things that sort of make a cloud around them?

    1. jason wright

      The user perspective and the investor perspective – the same perspective?

  80. FB Apps

    Social media, its a super saturated Entropy for anything..

  81. Raji Bedi

    Great post. What do you think about the term “social neutrality” to describe the democratization and adoption of a variety of interest-based social exchanges? More on how that term popped into my head: http://bit.ly/dIC6Ba

  82. Douglas Crets

    This conversation reminds me of something I said before at AVC.  New apps and platforms need to be careful of falling into the Yahoo! pit — trying to do all things for all people, creating feature creep rather than doing one thing very well and making that part of the brand.I have said before that I don’t want Foursquare that can do many things and that can grow over time.  I would rather see Foursquare the company create different apps that do those things, rather than use Foursquare to suck in what other apps do well. 

    1. fredwilson

      i agree

      1. Douglas Crets

        Thanks. That actually means a lot to me and made my day! – posted via Engagio

    2. Andrew Hoydich

      For sure. If we look at the development of society, one thing that can’t be ignored is that jobs/occupations have become incredibly specialized. The dozens of types of doctors we have today all have a common ancestor, the shaman/witch doctor/priest. There is no such thing as a one stop shop anymore…accept for walmart.

  83. johndodds

    Technologies see a network of nodes.Humans see an array of subsets of other people.Technologies prescribe behaviour.Humans seek the method best suited to the relevant subset.Technological connections are optional.Human connections aren’t..

  84. Carl Rahn Griffith

    Apropos nothing at all. but somewhat related to this topic – and pretty much everything we discuss on A VC – I have just discovered that the BBC documentary/reflective on Steve Jobs is available in full, it seems, on YouTube – so, everyone anywhere can watch it, hope.I feel I have to stress to everyone here to try and make time to watch this program – there is something about it that makes it very special and inspiring – now a few weeks have passed by since his passing away it resonates more than ever. I have watched it several times since it was aired just before Christmas – on our home digibox it is set to ‘Keep’ status so it can’t be deleted by mistake – a rare accolade.Anyway, here’s the link – hope you can access it, wherever you are – enjoy…http://www.youtube.com/watc

  85. Douglas Crets

    We just did a podcast about something very similar to this. It’s about 23 minutes long and it’s about discovering that people use different platforms to complete very different jobs they need done in their lives. http://www.therewiredgroup…. In essence: media is not fragmenting. It’s creating. People are drafting up and completing different media platforms because there are so many individual jobs to be done, by the user, in each person’s life.  

  86. Ray

    @leigh:disqus … my buddy’s step-mother made $7,360 last month. she works on the internet and drives a Cadillac Escalade. All she did was get lucky and follow the information you can find here..tiny.cc/wxw1y

  87. danshafer

    What, no mention of LinkedIn?Otherwise, a nice tickler article. I’d love to see more details and data. Pointers?

    1. fredwilson

      linkedin is a utilityit is not anywhere near social media

      1. danshafer

        Interesting perspective, Mr. Wilson. Can’t say I agree with it and it’s certainly not shared by many others I read, but it’s interesting nonetheless.

  88. MogulAzam

    I think this reminds me of the textile industry or the jeans.  Levis was the category killer with the largest market share.  Next came the designers  with their own niche brands: Jordache, Guess, Calvin Klein, Diesel, etc.  Just like fashion brands aim to fulfill the need  of individual to express themselves. Social Networking tools ie facebook, twitter, instagram, path, pinterest, tumblr,etc  are used by individuals to express conversations/content in their own way through different channels to others.  I think another point is that we are going to need to follow and study all these different tools with streams  to get a collective understanding of the content and context.

  89. ninetwentyeight

    After reading about it here a few times, it’s great to have stumbled into SoundCloud whilst working on a project.  Checking out Sound Engineer portfolios for a new venture (inspired no less by this blog making me watch Waiting for Superman a year or so back) and almost all of them seem to have profiles on SoundCloud 🙂  Looks like the music business has found it’s social network.

  90. Betsy Kent - BeVisible

    In reference to your comment about fragmentation…I’m seeing it daily in the work I do for my clients. For example, my CPG clients get their best engagement on Facebook. A celebrity photographer has thousands of followers on Google+ while his Twitter and Tumblr fan base is idle. An e-commerce client is getting massive traffic from Pinterest (which we are analyzing to see how valuable it is to actual sales) but has trouble getting engagement on Twitter; their Facebook Page struggling.Facebook may have the most users, but with their algorithm making it almost impossible for any business without a large ad budget to engage with fans, I think we are going to see more and more movement off the platform, especially for smaller brands who are going to have many more options to reach their market where they actually hang out. Right now we have to jump on all of them and see what sticks. In addition, there is some indication (my own non-scientific observation- I have a 22 year old son) that Facebook may not be where 20 somethings want to hang out anymore. So don’t discount the next “Social Media Platform of the Week”…they may all be able to  co-exist peacefully. We shall see.

  91. Anne Helmond

    We did a project last year where we “ran the numbers” by tracking the popularity of an article across different social media spaces: https://wiki.digitalmethods…



  93. fredwilson

    i hope last night was better to you

  94. fredwilson