The Implicit Social Graph

John Battelle wrote a gushing post about Color and what it means for mobile/social/local/realtime, augmented reality, and more. There are most certainly some big ideas in the Color app. I've never put a mobile photo app on my phone but I put Color on it last night. I don't have any of my family  on Color yet, but I hope to get them all on it today (I'm on spring break with my family and some of our kids' friends). Then we'll see what all the buzz is about.

Regular vistors to AVC know that I am in the "many social graphs" camp and most certainly not in the "one graph to rule them all" camp. I believe we will have at least dozens of social graphs in our lives. But even more, I believe that we will have social graphs that come and go and that are formed implicitly not explicitly.

My first experience with this sort of implicit social graph came almost six years ago via my musical neighbors graph at I don't think I actually know any of these people in real life, but they are the users who have the closest taste to mine in music, right now. That right now is important because my musical neighbors graph looked differently last year and will look differently next year. uses my listening history to create an implicit social graph in real time. Color uses my location to do the same for photo sharing. There are a bunch of social news discovery services out there that use my current reading history to determine a "news/interest graph" implicitly in real time.

This is the next frontier in social networking for a bunch of reasons. First, curating social graphs is a pain. It takes work. And simply importing your Facebook or Twitter graph is suboptimal for most social services. You then need to add and delete to get the right graph for the right app. And second graphs change over time. Who has time to constantly manage their social graphs. So they get stale and one day you say "why I am following this person?" or "why is this person a friend on Facebook?" And maybe most imporantly, sometimes you only want a social graph for a weekend, a day, an hour, or a minute. The only way to make that work is to construct it implicitly.

So I don't know if Color will turn out to be a big deal or not. I don't care that they raised $41mm. Seems like most anyone can do that these days. But I do care that they are pushing the envelope in social graph construction in an important direction. That's why Color is on my phone this morning and why I'm going to get my wife and my kids and their friends on it today and see what happens.


Comments (Archived):

  1. Aaron Klein

    What I find amazing is that on Twitter, it never feels like I’m “managing” my social graph. It’s just natural flow. I get tremendous value out of reading tweets, and along the way, I find new people to add to the flow, or others who aren’t giving value any more.The idea of implicit social graphs is very interesting. They get more interesting when I’m in a bit of a lull and am more interested in meeting new people or seeing new things than accomplishing what I’ve already got going.And the very limited time I have for that impulse is rarely not satisfied by Twitter.

    1. fredwilson

      twitter makes it fairly easy to add via WTF and to unfollow because their isno social cost

  2. JimHirshfield

    Could be a pretty cool collage of your family holiday – everybody’s different perspective on the same thing. Interesting that your experience with this app starts with the closest/strongest social graph there is, your family. Don’t need proximity to imply it’s existence.

    1. fredwilson

      yeah, but my kids’ friends being with us adds an interesting dimension

  3. Brad

    Anyone can raise $41M? Wow, I did not realize how easy it is.I will have to check out color to see what $41M buys you today.

    1. fredwilson

      well that was “tongue in cheek” but the whole funding mania is making mesick and has been for the past year.

      1. Brad

        Fred – Here is my problem with this deal. How are they going to recoup $50 let alone $41mm?I have a fancy MBA from a fancy school on the border of Harlem and Manhattan, I have ran a very successful business for the past seven years and I know that it is very very difficult to create cash flows. So my question to you about this is…..why do VC’s invest in a business like this that may or may not have a cash creation plan?I love social media, I think it is an awesome and free way to communicate with friends, customers and family. However, I can not wrap my head around how color is going to create cash? Through a sell of the company? To whom?Why don’t vc’s give an equity bonus (or some other fashion) for not using all the money from the first round?Let’s say you gave me $1M for a certain stake in and I created cash and a sustainable business and we did not have to go raise more money. I then returned to you $550k, and we decided not to go get more money. Isn’t that better than sequoia giving $41M and immediately begin working on series 2 financing?I have raised money, and on my last deal I returned half the money in 2 months because we did not need it and we started creating enough cash to sustain ourselves.I think part of the problem is that VC’s get as excited about raising money as the founders do and somewhere in between the idea of returning capital gets lost.This is my two cents, but it sure seems we are creating a bunch of free applications that do not have a way to create money. Just perceived pre and post money valuations. Tell me where I am wrong….

        1. fredwilson

          we wouldn’t do a $41mm first round either Bradi can’t answer your questioni know that Facebook is making a lot of cash flow now as is Zyngaand i am confident that Twitter will as wellbut you raise a good point

          1. Brad

            good to hear that their is some sanity out there in the VC world, and I agree that FB and Zynga are doing pretty well. I would also like to see twitter get there, in fact I want a lot of businesses to succeed because the country needs it.I would like to see VC get innovate and think outside the box, just like they expect their portfolio companies to do. I believe that Color will burn through whatever investment they have and hopefully in the process get to a break even or even profitability. I would like to see some restraint and prudent spending by founding teams and a reward for doing so.Your blog is fantastic, and I appreciate you doing it. By far the most interesting and thought provoking thing I read all day. Keep it up!

          2. fredwilson

            i would LOVE to see VC innovateand it is happeninglook at Y Combinator and Angel List and the explosion of microVC fundswe have changed so much of how we do things at USV since we got started in2003innovation is happening, but not everywhere

          3. sigmaalgebra

            Innovation in VC? I have no more data than you do, and likely less, but maybe we are seeing some ‘innovation’ in the Color deal.(1) Just the announcement of the $41 million got them going with users. They will have decent ComScore numbers like maybe tomorrow.(2) They have, what, 27 people? So, that is less than $2 million per person. Someone, not really me, might argue that’s like two guys in a garage getting $3 million? Also, what does it really take these days to create one job in ‘high tech’? What would a big company spend on such a project?(3) Maybe Color has in mind something specific and important to do with much of the money. E.g., Silicon Valley Bank is in, but have to doubt for just pure equity. So, maybe some of the money will go for some assets with some reasonably long lasting value, although for just what I don’t have a clue.(4) They are doing this in Silicon Valley. Okay: A house out there costs, what? Seven figures and up? So, maybe they will be paying their people fairly well. Maybe not Ferraris but maybe private schools for the kids; not usual for a ‘startup’ but okay in a nice place to work. Besides, what might Google, Facebook, etc. out there offer such staff? Since the equity funding is less than $2 million per founder, and since not all of the money will go just to ‘headcount’, if add in everything, office space, computing, Internet access, plane travel, company chef, winter outings to Vale, the less than $2 million per person won’t last many years before the company will need either some good revenue (say, $1 million a year per employee) or enough progress for another big round of funding. Net, it looks like they’ve got about 18 months of ‘runway’ before they need to be ‘off the ground’.(5) Mostly they will be stuck with just the 27 headcount until they have done some quite good things: For more headcount, tough to recruit, offer, get actually in on the first day of work, become useful, AND deliver value within that 18 month window. So, really, it’s just the present 27 until ‘off the ground’ within the 18 months.(6) Somewhere in here they might have to build an in-house server farm. Maybe with their plans, they need the server farm up with high scale soon; maybe with the publicity today they need that farm up this evening?(7) Their pitch ‘foil deck’ might have evidence that they actually have a lot of work done which justifies going directly for the real thing now and not waiting for a lot of slowly growing ‘traction’ and ‘milestones’ as intermediate ‘stages’.(8) Maybe Sequoia, for whatever they put in, maybe only $15 million, thought that, it’s a good target so get out the big gun, aim, fire, and intend to hit the target fully hard in one shot, NOW. Maybe they thought that it was an extra $10 million to significantly increase the chances of getting to $1 billion company value quickly, at which time everyone will forget about each part of the $41 million and the total.(9) It’s tough to ignore the value of Google and the estimates for Groupon, FB, Twitter. If such values are to be had, then TAKE them. “Strike while the iron is hot”. The $41 million is big bucks for a first round for a startup, but it’s not for a big business. Heck, that amount might be the bonus for each of a few dozen guys on Wall Street, and a really successful hedge fund guy could drop that for a nice dinner party. It’s also about what one of the corporate jets for Page and Brin cost.For what I’ve seen so far, the $41 looks a bit high, but I haven’t seen even their foil deck.You wanted ‘innovation’ in VC; maybe you are seeing some. Else, “no guts, no blue chips”. Net, I’m not totally convinced that the $41 million was too much.

      2. kidmercury

        imagine the poor ass fools who have to pay $4 for gas so zuck can enjoy $85 bn valuation…..or the 44 million additional people who have been thrown into abject poverty since central bank inflation went into high gear so that color could get 41 million…..

        1. fredwilson

          yes, i am not nearly as sick as they arebut i still feel sick

          1. kidmercury

            well, only rebuilding the world’s money supply solves the problem.

    2. Tereza

      Yeah you beat me to the punch, the hair stood up on the back of my neck on that one.Most people I know are begging — begging — for $10k or $25k checks. And are not getting invited to dinner anymore because their friends don’t want to be asked.Can and should this thing, no matter how cool, efficiently absorb $41m of seed?This is more then entire *funds* aiming to support ‘alternative investments’ such as women.The skew in order of magnitude here is insane. This is like a corporate investment. I know you’re going to say Sequoia needed to do that, to win the deal. And i don’t care! They should take half of that and spread it around. It would support a shitload of innovation.

      1. fredwilson

        don’t get to upset TerezaBill is a serial entrepreneur and this is about his sixth or seventh deal. ialmost invested in him once back in the mid 90s. i met him through markpincus.

        1. Tereza

          Don’t worry I’ll be fineI’m sure he’s awesomeIt’s just a crrrrrazy amount of moneySo I wanted to say what I KNOW other people are thinking

          1. fredwilson

            i did not say he is awesome

          2. kidmercury


        2. Tereza

          I do hope that he’s paying it forward in a thoughtful way. He very well may be.It’s that ‘to whom much is given, much is expected’ thingy.

      2. Aaron Klein

        The only business that I can think of that needs $41MM of seed is an airline.(tongue firmly in cheek, but still…)

        1. Tereza


        2. Brad

          LOL…I am still baffled by this deal. I even downloaded the app, and it keeps crashing…..

  4. ErikSchwartz

    I’m assuming that there’s more to Color than a nifty photo app. I hope they’ve developed some very cool engine for the dynamic creation and maintenance of ad hoc networks based on various, and even arbitrary parameters. If those tools rock, I hope they do more interesting things with them than letting you see pictures taken by other people at the same place I am right now (which frankly seems to me to be a solution looking for a problem).On the giant raise; The thing about raising 40MM pre launch is that it skews your adoption numbers. They will get to 1 million installs based purely on the hype of Sequoia investing so much money. In a week they’ll make a press release saying they’re the fastest app ever to a million installs and that won’t actually mean a damn thing.

    1. Aaron Klein

      I agree with you. I have no idea why a web startup would need or want 40MM pre launch.If someone offered to write me that check today, I would say “I’ll take 2MM” and even that would be way more than I need right now.

      1. ErikSchwartz

        Taking so much is very limiting as to what businesses you can get into.Apparently (according to TechCrunch) there are 27 employees at color. That seems like a lot.

        1. Aaron Klein

          Holy cow. $41MM and 27 employees. There’s got to be more to this than photos tied to location.If not, we’re getting well beyond frothy and into ridiculous.We did a nice small seed round and are focused on keeping a very conservative capital structure. Maybe I could have gotten a better valuation, but now I have plenty of room to grow it with even just a little bit of user traction later this year.

          1. ShanaC

            I don’t know why people aren’t more conservative. Money is inherently risky if you like turning it into oil for a tech fire…And I keep thinking we’re pouring money on a tech fire…

          2. Aaron Klein

            I’m not sure they understand the consequences if they’re not.The finances of my State of California look a lot like a startup with an overgrown valuation in relation to their market traction… πŸ™‚

      2. Donna Brewington White

        I have to think there is more to this than meets the eye.And, if not, well…that’s scary.But, in reading a chunk of “Founders at Work” on the plane last night, it was clear that there often is more than meets the eye in the initial funding and that the initial product may be an on-ramp rather than a destination.BTW, sounds like you’ve got a great example of a lean startup, Aaron. Looking forward to hearing more as this develops.

        1. Aaron Klein

          I agree completely. That just has to be the case…And all I will say about our startup is…you definitely don’t need $41MM to have “more than meets the eye.”The core technology we’re building, and the engaged community of users that we’re going to build from our products, have the potential to disintermediate three massively huge markets.Enough that I’m convinced we probably can’t do it all ourselves, but time will tell how we choose to get there.

      3. eyesparky

        Seems to me that the VCs have seen something in the medium to long game and want to corner the market on it, effectively funding several rounds at once to free the team to get on with developing the product free from focusing on funding.Really surprised how unpolished their launch apps have been though, which really has the potential to put many off the product short term. Actually love the concept of fluid, proximity based networks but feel that Color jumped the gun a bit with their release.

        1. Aaron Klein

          Don’t get me wrong, I love the concept. I just can’t believe the numbers. I wish them nothing but the best and hope they’re wildly successful.Still don’t get why this wasn’t launched a few weeks ago at SXSW. That’s the perfect spot for this app to shine.

          1. eyesparky

            Totally agree about SXSW but having played with the Android app today and talked to a friend on the iPhone app, it feels like they dropped them too early. We could have had a PR announcement today around the funding and a teaser about what was coming soon, with more polished apps and website hitting a little down the line. Scoble had a point in his Cinch cast about first impressions and with the PR for this launch reaching pretty wide there may be a fair few unforgiving early adopters.There is no doubt that the team is a strong one and that they are working on a bigger vision. It just feels like early traction may be hit a tad. Will be fascinating to see how things unfold.

    2. gzino

      Agree what is more interesting are the parts that don’t appear to be known:1. Curation and signal to noise mgmt? Mix of crowdsourced, algorithms, hired humans?2. Business model? Pls don’t decrease my signal to noise ratio by injecting $ funded pics/videos to top of my stream.3. APIs? Any third party app can build on top of stream, adding in their own curation etc? Can user post to all his social nets and Color at same time, and/or auto-synch between Color and personal social nets?

    3. ErikSchwartz

      There’s a story about color on CNN now. That’s a lot of hype.

    4. markslater

      i just tried it. WTF?

  5. OurielOhayon

    Fred, out of curiosity do you think that apps will have their implicit social graph, like music, media or games? does that make sense to you?

    1. fredwilson

      yes, very much so

  6. Clay Schossow

    Couldn’t agree more about multiple social graphs. I think this is true for both individuals and companies. For me, I’ve met (and drank with) 99% of my Facebook friends, but I’ve probably only met 5% of my Twitter friends. Similarly, I think people who Like a company’s Facebook page are looking for a fundamentally different (more interactive) experience than those who follow them on Twitter.I don’t see any day where there is just one graph — it just doesn’t make logical sense. Think about it in real life as well, we all have different graphs of friends, associates, mentors, etc. There’s a certain appeal to being able to have separate graphs that are unique and not intertwined.Color looks very cool and I think it could fit well into another service, but I don’t think it’ll be the graph that rules all as Battelle seems to argue.Enjoy Spring Break and thanks for continuing to post!

    1. fredwilson

      posting is thinking outloud for me. i don’t stop thinking when i am onvacation. if anything, i think way more.

      1. Clay Schossow

        Good point. Bet it gets tricky when you think in the shower, though…

        1. fredwilson

          the water protected iPad! i am joking but half serious. i’ve blogged onsubways, treadmills, ski lifts. why not showers?

          1. Mark

            Was just at a hot spring in China, where I saw a 60-something yo man using an iPad in a spring.It was jarring.

    2. sigmaalgebra

      You will find some at least moderately thoughtful discussion of your points about “different graphs of friends, associates, mentors, etc.” in…This was from a guy at Google. My guess is that he spent a few months putting this together. For a first cut, it’s okay. My guess is also that someone at Google wanted some ‘clarity’ on just what a person’s ‘social graph’ is in principle (separate from computing), that is, what categories of ‘friends’ they might have and how many friends they might have in each category, as a foundation for more efforts by Google in ‘social’ whatever.To give you a preview, for ‘good friends’, say, to keep up with closely, looking at only half a dozen or so. For various categories, say, work, school alumni, School Board, neighbors, looking at some quite separate sets. In total, for the number of ‘friends’ can keep track of in any very meaningful sense, looking at max ballpark only 150. For more and better, follow the URL and read the ‘foils’.Social’s important. Sure. People are “social animals”, and the market cap of Facebook, Twitter, etc. provide good evidence. Maybe the ‘social Web’ is actually changing people and civilization, getting rid of the 150 limit, etc. If not, then there are some fundamental constraints on just how seriously we should take the ‘social Web’.Then one guess is, once some of the fad of Facebook, etc. dies down, people will find that they have plenty of ways to keep up with friends in various categories, e.g., fact to face, telephone, letters, e-mail, fora, blogs, ‘gated blogs’, along with FB, Twitter, etc.I would propose: When people spend a lot of time on FB, Hacker News, etc. there is something they are looking, hope to find, partly find, but mostly don’t find. More progress would be to provide more accurately what people are looking for, after we identify that!My thinking is, compared with the purely social, people have more direct, pressing, important, serious, and fundamental concerns, e.g., careers, family, entertainment, that, yes, also support their ‘social’ interests. So, the current, first cut, ‘graph oriented’ way to get at these concerns is the ‘interest graph’.

  7. Raj

    The implicit social graph is the disengaged social graph. And we know there’s little value in that.

    1. fredwilson

      to date that has been true. and a great point Raj. but i think that will notbe true forever.

      1. Raj

        Passive services which lack engagement are not very valuable. Been there, done that.They typically fall into the category of “problems that don’t need to be solved.”There are always exceptions, but in this case there are very few.Interested in hearing more about why you don’t think that will be the case forever.

    2. tomazstolfa

      I tend to disagree. There is a lot of value in connecting “tribes” and this is what does well.Color might be a different story, but I can imagine a great use case – the SuperBowl or another huge event where you would get several thousand snapshots, auto-tagged and grouped around the same context from people that fit in the same “tribe”.

      1. Raj is all about ego. They aim to be about discovery, but it’s not the case. There aren’t millions of people using to find music. They use it to see their play count increase. It’s all about vanity and ego.I stopped using it when I realized that. It adds little to no value to my life.

    3. Aaron Klein

      Won’t be true forever if you can dip in and out of it for the value you want, and it’s enjoyable to engage when you’re there.But I do think Twitter is what ties it all together. It can be a great way to drive people into a sub-graph, and a great tie that binds after you leave it.(Obviously, not for all the people you meet there, but for a subset you really enjoyed getting to know.)

  8. Darren Herman

    I want to start a company around social graph maintenance. It’s annoying, so thus, an arbitrage opportunity.

    1. fredwilson

      yup. but i don’t think straight up is the way to go. back door seems better.

      1. Darren Herman

        and for the record, i was joking πŸ™‚

        1. fredwilson

          i am not. i think there is a play to be made there

          1. Fraser

            Agreed. My pain with social graph maintenance / ‘one graph to rule to them all’ is accelerating, which means there’s an opportunity here.I wonder if a (light) maintenance service could be the thin-slice way to get traction in the identify market and challenge FB.On the pain front, here’s a short exchange from Twitter I had with my friend Vinny:Vinny: Don’t like blogs using Facebook comments. Wrong graph for sharing my comments. Wish Disqus had comment graph.Fraser: agree. this is a perfect example of the need for different graphs. tough problem but one with an increasing need.Vinny: gives hope to startups that Facebook won’t dominate everything

          2. Aaron Klein

            I have yet to comment on a site that uses Facebook Comments. It’s just not the kind of thing I want to broadcast on my Facebook circle, so that’s the number one way to lose me as a commenter.I hope that Disqus has nothing to fear from Facebook comments. They certainly have me locked up as a loyal user.

          3. fredwilson

            i refuse to use FB comments. there have been at least a dozen posts at TC iwanted to weigh in on since the cutover. but i haven’t.

          4. Aaron Klein

            Really dumb move on their part. The comment streams look bad, feel bad and they lost a ton of engagement they had going with Disqus.

          5. ErikSchwartz

            I HATE Facebook comments. Most of my friends (real life friends) do not give a rat’s ass about technology. They have no interest in seeing me get into a long comment thread on TechCrunch about the nuts and bolts of how an HFC cable plant works.

          6. fredwilson


          7. Ro Gupta

            Can you describe more about what you and Vinny mean when you say you wish Disqus had a comment graph – i.e. What exactly would that look like to you, what specifically would you like to be able to do with it.

          8. Fraser

            Ro,David sort of touches on it here:…There are people I want to see blog comments from, and it’s not my highschool friends who are commenting on ESPN’s Sports Guy or Perez.Every day I come across a tonne of interesting articles related to tech startups. I’m sure there are consistent commenters across these posts. Implicitly connect me to these people. I may not want to directly converse with them right away but like neighbors, it can be a great way to drive content discovery.

          9. Pete

            @ro – there is a long thread on the topic of social interaction on Disqus, buried in the middle of an unrelated string of comments on an (also unrelated) avc post a few months back:…What do you think?Pete

          10. Ro Gupta

            @Pete thanks – hadn’t seen that extended thread. some of the basics can be found at but stay tuned πŸ™‚

      2. thewalrus

        agree. compelling use case ALWAYS comes before platform. many people tried to offer the distributed social platform of the web (ex. ning)…. but facebook provided the instant gratification of sharing drunken college photos.identity will evolve the same way… openID will happen… but through a compelling use case, not by trying to sell the benefits of an open platform (its just too abstract for anyone outside the tech industry).

        1. fredwilson

          so true. i want to write and think a lot more about this. it is soimportant.

          1. thewalrus

            i’ll be in NYC in a couple weeks… i’d be happy to think out loud with you πŸ™‚

    2. falicon

      dealing with some of this should naturally come out of the feature set down the road (ie. if we do a good job of pointing out the stuff you missed, and the stuff you probably want to pay attention too…then we can also do a good job of the reverse and eventually point out the people you should start/stop paying attention to)…so all joking aside, it is at least on our minds and in our longish vision =D

      1. fredwilson

        i found the link to color last night on

        1. falicon

          So awesome! Thanks.

          1. Mark Essel


      2. Bart Denny

        great ideas. just signed up!

  9. OurielOhayon

    Fred i think the app is probably promising but it has a serious non instant reward problem because it requires other people to have it to enjoy it. Building an implicit geo network works fine when you already have a mass of users (something foursquare can afford – and that agora app for example has done with its API)But doing it from the get go is really a hard jobps: btw ratings are terrible for now

    1. fredwilson

      it’s not a great out of the box experience

      1. PhilipSugar

        Which is such a big challenge if you raise that much money.Everybody gets to see your alpha and how much it sucks.I am with you when I see $41M out of the box it puts a pit in my stomach because I know the higher you get the pendulum to swing on the upside means the higher it swings on the downside.

      2. Rocky Agrawal

        they need to seed it with other data (e.g. flickr/Yelp photos) and have their own photos have persistence until there’s a denser user base. as it is, it’s going to be tough to have enough liquidity.if i keep looking around and keep seeing nothing, i stop using the app. also potentially bootstrap with notifications to encourage picture taking.otherwise you run into the sorts of issues google had with wave rollout.

    2. Renee

      Yeah. At first I thought some of that was just snarking on the funding announcement, but I actually can’t make the app work on my 3GS. Keeps crashing after I take a picture of myself, which is kind of funny.

  10. Dave Pinsen

    When you blogged about Hashable last week, I felt like I was in-the-loop a little, since I’d heard about it before reading about it here. This time, with Color, I heard about the parody before reading about the real thing here: “Color.XXX: people. colors. apps. mobile. social. pivot”.

    1. fredwilson


    2. sigmaalgebra

      Cute parody.BTW, the Siamese cat in that slide show, except for the bacon, looks almost exactly like a cat I saved from the unusually severe winter we just had in NY — some snow still on the ground. Cat was on my back porch eating old bread I’d tossed out for the crows — hungry cat. I kept putting out food, then a little enclosure it used, and finally when the weather was about -15 F brought the cat in.I have a fantastic kitty cat, a gray tiger striped boy about 6 years old, playful, comes when called, understands several hand signals, understands “No!”, tugs at me when he wants something, follows me room to room, likes to sleep on the blankets over my feet, sits beside my right arm when I use my computer, etc., and I shouldn’t have a second cat.If anyone wants the Siamese, it’s ‘totally available’.It’s affectionate, curious, very lonely, somewhat ‘vocal’, sensitive, perfect in its litter box, and, still, somewhat scared but calming down. It’s built very much as in the picture. The eye colors are about the same.

  11. Greg Neufeld

    I played with Color last night. It’s clean and has a great UI, but in such close proximity to neighbors in the east village, I started to see some interesting pictures of people that reminded me very much of chat roulette (not in THAT way, but just in the sense that there were pictures of strangers in dim lighting popping up on my phone). I can see the value when you’re out with a group or on vacation, but I’m sure people will also stumble upon some pictures that they didn’t want to see.If Color is Yobongo for photos, then it’s not true that a picture is worth a thousand words. It’s more like 30.37 words, as both announced funding yesterday.

    1. fredwilson

      that’s great input. density matters, a lot

      1. JimHirshfield

        Which is why someone on Hacker News made the intelligent comment that perhaps Color should have launched in one or two cities a la FourSquare.

        1. lawrence coburn

          It’s a good point Jim. But I think this is where the $41M comes in.

          1. JimHirshfield

            I guess so…that’s a lot of feet on the street marketing. Or similar local marketing.

  12. andyswan

    I’m looking forward to the day where I no longer have to discover anything. Where all of my relationships, likes and dislikes are all mapped on graph paper, digitally.If done properly, 95% of the people that “know” me won’t know I’m dead until long after I’m gone….when they realize that they never knew me at all.

    1. fredwilson

      don’t laugh

    2. JimHirshfield

      Can you prove that you’re not? How do I know your comment wasn’t written by a bot?

      1. andyswan

        You don’t. My twitter-powered tombstone is in production, and will be in place shortly. We’re goin Huck Finn on the social media world.

        1. fredwilson

          Huck Finn and Snowcrash is all you need to read to understand USV’sinvestment strategy

          1. Dave Pinsen

            Shapeways seems like it was inspired by The Diamond Age.

          2. fredwilson

            OK, we’ll add that to the list!

          3. cyanbane

            Please tell me at least one of you comes to work with a samurai sword and pizza delivery bag.

          4. fredwilson

            no but we should

          5. ShanaC

            Ok why Huck Finn. Snowcrash I get, though I think you should add “Super Sad True Love Story” as well for a more distopian view….

          6. fredwilson

            The way tom sawyer got everyone else to paint the fence

        2. JimHirshfield

          Is that JSON or XML feed from your Tombstone? Hmm…now THERE’S an implicit social graph: all the corpses within your proximity.

        3. orenraboy

          sell it on an app-store and setup a foundation from the proceeds.

    3. Elia Freedman

      Thank you for summing up my cynicism so succinctly.

    4. ShanaC

      I’m not. Life is a journey. If I am that well known to others, how will I discover the new in me?

      1. andyswan

        See if you really knew me you’d know that was sarcasm. πŸ™‚

        1. Mark Essel

          Imagine a virtual digital assistant that “knows” your tastes, interests. It’d be a great time saver except when it screws up and pokes someone’s wife on your favorite social network.Overall a transient history of our tastes in a digital format would open up new channels of communication and exchange. I can see implicit digital data being very helpful to your mission.

          1. andyswan

            Imma get a real life one.

          2. Mark Essel

            You and your wetware πŸ™‚

        2. ShanaC

          πŸ˜› I should tell you that every once in a while you hear a wooshing soundpast my head – it is the sound of me missing sarcasm

    5. Donna Brewington White

      This is scary funny.

  13. falicon

    Speaking of posts from six years ago…I wanted a quick and easy offline way to catch up on all the avc posts…so I wrote a script that pulled them all down and dumped them into a PDF…it’s not perfect, but it does the job I wanted…anyone else that wants a copy can grab it from here ->

    1. fredwilson

      every single one of them???

      1. falicon

        well a little over 5,000 of them…btw sorry about nailing your server during the download process =)The PDF includes just text (so no pictures, links, music, or comments etc.)…but each page does have a link to the actual post so people can jump over and engage with it properly there…BTW – so entertaining to go back to 2003 and read your thoughts…amazing how many things you’ve invested in these days are directly related to the things you were talking about even then (just more clear now that it’s all in hindsight) =DOh and I also did the same thing with in case anyone wants that PDF too…they are really both so much better than most books on the markert today.

        1. fredwilson

          i feel like you downloaded my brain!and i’m flattered anyone would do that

          1. JimHirshfield

            An API for Fred’s brain? Cool.

          2. fredwilson

            it gets rate limited in the mid afternoon when i need a nap and after 10pmbest time to hit the api is 5am eastern

          3. JimHirshfield

            Surely a cloud-based, load balancing, auto-failover DB would solve that problem. Talk to your buds at MongoDB

          4. Robert Holtz

            Talk about the graphs that would come out of THAT — the “Fred’s Brain API.”Sign me up for the SDK right now!I’m willing to pay good money for that.:-D

          5. falicon

            Now to just find the time to actually read all 6,282 pages (it’s a 16MB PDF)…and then I’ll get started on the cloning process! =D

          6. fredwilson

            and people tell me i should write a book!i’ve done it already 10x over and i’m still writing

          7. Mark Essel

            Pro move, will grab it post work.

          8. peterarmstrong

            Fred, at Leanpub we did this as a proof of concept for MBA Mondays at, and we’d be happy to do it for your entire blog. (That’s what we did with Eric Ries’ blog and Venture Hacks.) The only reason we haven’t done it for your whole blog is that it’s your copyright… (I pinged you regarding the MBA Mondays book earlier, but I assume you missed it.)Anyway, we’d be happy to import your blog and make a Leanpub book. We could sell both MBA Mondays and the full blog, and donate all the proceeds to charity.Can you let me know if you’re open to that? My email is [email protected].

          9. falicon

            funny thing is that…even though I don’t need to now…I would *still* buy this if you did it.Would be even more valuable though if you do a limited run, print them out, get Fred to sign them…and then offer it up as one of the ‘rewards’ for the annual Donors Choose challenge Fred does…Cash in on the 1,000 true fans…wave of the future πŸ˜‰

        2. Greg Neufeld

          congrats Fred, you’ve reached Jay-Z status (when Bill Maher presented him with a book of all of his rhymes)

          1. fredwilson

            does that mean gotham gal is beyonce?

          2. Greg Neufeld

            No, I’m pretty sure yours wasn’t an arranged marriage πŸ˜‰

          3. fredwilson

            that is for sure

        3. kidmercury

          seriously awesome man. you deserve a badge for this.

          1. falicon

            Thanks! If they ever come out with a “I hacked Fred Wilson’s blog” badge…I *totally* want it. =D

          2. Mark Essel

            A touch of gimp/adobe and the “award” could be yours… so little time so much awesome in the world.

    2. JimHirshfield

      You want $41M for that?

    3. RichardF

      awesome Kevin! That is going straight to Kindle – thanks very much

      1. falicon

        Let me know how it looks on the Kindle…I really only spent about an hour hacking this all together…so no real formating or anything special…just the basics to get what I wanted done…the true joy of being a programmer! =)

        1. RichardF

          Kevin it looks great on the Kindle. Just need an index and Disqus but it’s still great having avc on my Kindle to dip into.First page is Fred’s first post and it reads like a book, last paragraph of first chapter:”I read blogs a lot. And i think they are great. So i am starting a blog. I have no idea if i’ll write a lot in my blog or rarely. I hope its a lot, because i have a lot to say. But we’ll see about that.”say no more….

          1. falicon

            So cool…I hacked the project together for selfish reasons/needs, but it’s very exciting to see so many other people getting value out of it too…It does make me think I probably should have spent a little more than two minutes on the layout/design though…ah well, maybe in a future version =)If anyone has a good idea on a proper way to integrate the comments, I would do that…but right now, it would turn a 6,000+ page PDF into 20,000+ pages easily…I would probably have to make the PDF generation on-demand and allow you to pick the specific set of ‘thread starters’ you wanted comments included from…I think that *could* be really interesting…

          2. RichardF

            I’ve just run the pdf through Calibre and created a table of contents based on the urls of the posts. Once I’m happy with it I’ll post up the ebook

          3. Donna Brewington White

            Richard, you rock!There truly is a “dream team” in the midst of this community.

    4. ShanaC

      thank you!

      1. falicon

        My pleasure! We still need to catch up for a cheeseburger club lunch one day soon too!

        1. ShanaC

          Well, what are your free moments before Passover? Shoot me an email @ shanadot carp at gmail dot com

    5. Dan Epstein

      Thanks Kevin. This is awesome.

    6. Mark Essel

      Great work Kevin, I wanted to do something like this for my blog but the formatting or embeds always get messed up, forcing me to manually review all the posts. Dynamically constructing pdf bundles out of tags would be interesting…

    7. markslater

      awesome! – who want my 700 comments? ……crickets…..

    8. Donna Brewington White

      My dream come true. Fred Wilson: The Book. hahaThanks, Kevin.

  14. David Semeria

    I think Disqus should push this angle more.When people congregate around a whole blog or a single post, that’s a strong implicit network indicating a shared interest.Best of all, it takes no settting-up, it decays over time and can be highly relevant from a real-time POV.

    1. fredwilson

      yes and daniel knows this.

      1. gbattle

        There’s a whole lot of interesting things that can be done when you already have a line of asynchronous Javascript on someone’s page … I can name dozens, and shared a few with Ro and Dan last year.

    2. awaldstein

      agree completely on this David

    3. Tereza

      So true. BTW would love to see video/audio integrated into the experience. FUN! A camera roll on the right.Remember that day when Fred posted a pic of himself ROTFL-ing? And how we jumped on that? It is so engaging. But convo around it is critical, it’s the glue

      1. David Semeria

        I do remember Tereza – that was a blast.But something tells me moments like that will become ever more rare as our host’s public profile continues to rise.C’est la guerre, I suppose.

        1. fredwilson

          i should do it more, not less

          1. Donna Brewington White


      2. Donna Brewington White

        Truly an inspired moment…within the life of an online community and in the context of that interaction.

        1. Tereza

          Kicked off by YOU, btw!

          1. Donna Brewington White


      3. Universal_Mind

        What Disqus has right now is perfect. I say leave it alone.

        1. Tereza

          Kinda like overdoing it at the all-u-can eat buffet? πŸ™‚

        2. Trevis Thirdgill

          I think so too Disqus has the best commenting UX. Color should incorporate this commenting structure into there app for all the spontaneous conversations that emerge around various content. I think you could actually incorporate the entire internet into this concept. Article posts from authors really aren’t different then any other comment its just a public message from an unconfirmed contact.Put color, pulse RSS feed reader, Yobongo and Qwiki into a blender and you have the future post pc operating system. Right now I think its a little too photo centric. You should also be able to post just text or text and pics/video and have a thumbnail version show up like what pulse does with pics and headlines. This gives more context. Thats just the front end. Algorithms like IBM Watson, Gravity’s interest graph etc. should be running in the back end. Ultimately, Implicit social graphs, recommendation engines that push multi media content (i.e people, photos, stories, etc.) to your feed merged with Qwiki’s information experience concept is the way of the future.

          1. Trevis Thirdgill

            extrapolating this further, imagine instead of explicitly having to take pictures you lifecasted from your always on sensory network — not just cameras to capture what you see but some kind of neuromorphic nanotechnology in your brain that captured your entire gestalt i.e. everything you see, taste, touch, smell, hear, say, feel do, etc. that others could tap into. I think that is the future of the hive mind. Color is a step closer to this.

          2. Trevis Thirdgill

            heres how it would work automatically with lifecasting; your full embedded sensor network is always broadcasting your experiences 24/7 365, but due to sophisticated computer vision and robust image recognition capabilities, it is smart enough to recognize when you are in the bathroom or doing other private acts, undressing, having sex, etc. and will autopause the livestream and rebroadcast when you are decent again.These automatic Privacy/content filters will be customizable for each person as opposed to the current forced one size fits all approach. shared porn experiences may even become a popular option that pops up in the feeds of the more sexual exhibitionists among us. Meanwhile, The more conservative user will only see things they’ve deemed appropriate — sexual content would not show up in there feed.You can also pause the stream manually at any time simply by saying “pause”. People however will be less inclined to be on pause too often because While on pause, not only will other people not be able to access your live stream but you will not be able to access their’s either.In the background, the system would simultaneously auto edit a running highlight reel of your more interesting experience clips, which cuts out all the boring mundane parts of everyday life (i.e sleeping, zoning out while waiting for something, etc.).By default raw experience streams would always be available to the collective, (excluding the aforementioned inappropriate/private parts) but users could switch to the ‘highlight view’ which only recommends interesting/relevant experience clips that you can tap into via seeing what they see, hearing what they hear tasting, touching, smelling, feeling what they feel or all of the above.

    4. Rocky Agrawal

      definitely.. and not just on this blog but others on the disqus network. i’ve long wanted to be able to select certain people and have their comments automatically highlighted whether i’m here or at cdixon or msuster.

    5. Gregory Magarshak

      Disqus is awesome. It’s on the distributed web (wordpress ain’t a centralized social network) and has ~200 million users (I think). It could gradually become the awesome social network that facebook wants to be on the web.

    6. Donna Brewington White

      Really appreciate this observation, David.(Completely coincidental that this is the second time I’ve responded to a comment from you today — I’m not intentionally following you around the blogosphere. Although, with Disqus, I COULD if I wanted to!)I recently noticed that on the Disqus dashboard, they title the list of blogs I have commented on as “communities.” While this is not always the case with the blogs listed, I really like where they are going with this.

      1. David Semeria

        Don’t worry Donna, I’m used to ladies frantically chasing me around the web ;-)Seriously though, it reminds of a game Chris Dixon invented (and which Mark Essel and I had a stab at) in which people had to secretly communicate through open services on the web.I thought it was great idea. Someone should pick that ball up.

    7. Chris McCoy

      Communities are built around content. Simply put. Graphs are real identities of people, connected to things (personalities, places, events, knowledge, payments, etc.). Social graph are your connection to those people and things. So yes, Disqus could out color Color around comments.

    8. hypermark

      So totally agree, and there are two logical angles on this. One is using that graph data to feed similar content to you (content curation). The other is using that data to provide transparency to like minds (kindred spirit curation).Given the breadth, depth and quality of comment threads and link data that DISQUS has facilitated, you could create the un-walled version of Quora around this model. This would seem to explain, in part, Facebook’s move into the comments space.I think that my head just exploded. πŸ˜‰

  15. thewalrus

    i think implicit/explicit social graph is another way to say “other people as filters”. i totally agree this will be fluid, and important signal. and there are many other very important signals…. like geo, time, personal interests, etc.the magic is when these signals can be combined in new ways… mobile provides great new opportunities. seems color is looking for the secret sauce in social/geo (different user case, but similar signals as 4square, etc).

  16. reece

    I agree that managing social graphs can be a pain, so I like the idea of an “elastic” network (in the words of the founder).I’m curious where else this may be applied effectively.LinkedIn certainly gets stale over time – what if Hashable used your location in relation to others to really determine connection strength?Really exciting, slightly scary, can’t wait to see how it all shakes out.

    1. lawrence coburn

      A company called Lokast coined the term “disposable social graph” a while back. I like Elastic better.

      1. fredwilson

        i like elastic a lot too

  17. im2b_dl

    You gave me a great idea this morning for UI/UX and the storycube/transmed-story stuff with this post. One of those times when you are making pizza…you read a story about kitchens and you get a better idea for a better tasting pizza.

  18. Elia Freedman

    If I had $100 for every mobile, social app I wouldnt need my next venture funded. I could do it myself! $100M needed? No problem…

  19. Danny Espinoza

    I’m a firm believer that the an implicit social graph that is constructed from multiple transient social graphs (and Color sounds close) is best positioned to expose Facebook ‘s fundamental weakness: a social graph constructed from arbitrary binary (friend or not) decisions results in an internally inconsistent reflection of reality.As you point out, Fred, nobody likes to manage the graph. It *should* be implicit.In my opinion Color is a clever trojan horse. If successful, it has the potential to stealthily starve Facebook of it’s high growth areas (photo and video sharing) while building a competing network that accurately understands the connections between people, events in time and location. As the network grows, adding status updates and messaging (which are likely in place but not exposed) will reveal Color as an architecturally superior competing network to Facebook.Good stuff. And it’ll be interesting to see if it succeeds.

    1. fredwilson

      and remember that facebook is basically a photo sharing service

      1. Matt Kritzer

        Facebook has indeed gone too far with trying to fit multiple implicit graphs into a one size fits all graph. They should stick with the true “friends” approach. These are the people that truly care about your day-to-day life experiences, not necessarily your interests.Interesting thought though: If we become more focused on digital connections going forward and less so on real person relationships, does Facebook then lose context and it’s grasp on the “true friend” graph?

        1. Robert Holtz

          Here I find myself REALLY agreeing with Fred in that if you try to lop everything into a “one true graph,” you will get the wrong picture of Fred or anyone else. By definition, its need to be singular and conclusive makes it lack the necessary granularity to give you any real insight.Rather, many graphs that represent many aspects that a give service is designed to learn, observe, and in some cases be trained to understand very well. And I particularly agree with the NOWNESS of the graph in that we are all constantly evolving, changing, and then sometimes reminiscing. The snapshot of any one graph morphs over time and not necessarily in one direction of progression. If you’re talking about a musical graph for example, tastes ebb and flow so old tastes that have become boring in one snapshot re-emerge again on a future date.The future of Facebook if it wants to hold its place is not to try and be the ONE TRUE GRAPH but to interoperate with a variety of graph-oriented services to provide, if anything, a Meta-Graph or graph of all graphs that would have summarizing information as well as granularity and depth that could only be truly exposed and altered through the combination of many services rather than just one.But, to Fred’s point, even that top level “Meta-Graph” would need trending to be truly useful over time if one is to extract any real context. Alternately, any graph can only be relied upon to be good for the immediate period surrounding right now.Facebook or anyone with similar ambitions would need to take this approach. Otherwise, Facebook’s social graph will become one of many graphs that come to matter and it will fall on someone else to aggregate the Meta-Graph. Even if that comes to pass, the implicit input streams will come from many services and many apps that are distinctly interested/focused in their own affinities… their own graphs.

          1. Matt Kritzer

            Really well said. I would think that Facebook understands this to a degree and it is *super* early. Perhaps Facebook wins in the long run because they have the biggest war chest and can make these acquisitions to form a “Best Of” graph platform.

          2. Robert Holtz

            Facebook has proven highly adaptive so I wouldn’t count them out. But look at the backlash against Facebook Comments just in this thread. As people start to realize their actions are building a graph, even implicit graph-building becomes explicit to a certain hard-to-quantify extent.I’m reminded of the Seinfeld episode where George Costanza doesn’t want one circle of friends to meet his other circle of friends.For a successful “Best Of” graph to ever exist, such an operator would need to be very okay with not necessarily owning or branding all of the input streams… to allow individuals at least the appearance that their social circles are distinct and compartmentalized rather than thrown all into one pile.Yes, Facebook has a big war chest. But in a world where Color can come out of left field and bag $41 million in a first round, that buyer could be Facebook, it could be Google (who has yet to hit their stride in social and could do it in the name of search), or it could come from someone none of us has ever heard of yet.No matter what, you rightly refer to this as being *super* early. I believe in the many graphs model like Fred does and we are just at the start of the whole phenomenon before we have graphs to cover every subject, before we see consolidation of graphs, and before we see one graph success translate into the demise of another.No matter what, that’s what makes our field so interesting and dynamic. Yes? Get the popcorn popping and enjoy the show. πŸ™‚ That’s what I say.

          3. ErikSchwartz

            The thing that is so frustrating about facebook is that they have the data to do really nuanced and interesting dynamic social graphs and they don’t do it.Don’t just stick my TC comments up on my wall. Only show them to fellow Y! alums, or start up friends. Do not show them to theater friends and sailing friends and cooking friends. They don’t care.The power of facebook is not they they know who my friends are, the power of facebook is that in many cases they know why we’re friends. They underutilize that knowledge.

          4. Matt Kritzer

            Hmm. Trying to reply to your last comment Robert, but am unable to.I also think the continued adoption of Facebook Platform for 3rd party authentication is the trojan horse that might win the day. So long as Facebook is able to remain the dominant people directory, they can use that war chest most effectively (esp. compared to Google).@ErikSchwartz: I think Facebook will have to fall inline eventually. Us tech folk are ahead of the curve and are not really representative of the average FB user.It’s gonna be a great show. Just watchin’ the previews right now. I’ll be first in line to see how Zuck and his service responds to the pressures of the street.

          5. Dennis

            We’re actually creating a service for this right now – check it out and we’ll let you know when the private alpha becomes available – it’s time to not create a different network, but to create a better

        2. Universal_Mind

          Keep in mind they’re preparing for the BIG day down the road.They gotta do everything . And inturn,they will destroy themselves..HE HE HE !

      2. AVillageOfHockey

        i disagree, it’s not a photo sharing service. Among my friends, fewer people tag photos these days because you don’t know who they will be tag/shared to. The primary use now (judging by this small sample) is twitter for friends… just that you can use more than 140 characters.

        1. ShanaC

          I’m getting the feeling that it is more in between. There are tons and tons of photos that I see (especially baby photos as of recent…) otoh, I’m seeing as a percentage, less people posting those photos…

          1. AVillageOfHockey

            I guess it can be described in several different ways, however, the primary use that has held out and is of real value is posting either an opinion, link, photo, video or activity and people you know interact with it via comments or likes.Twitter: NewsfeedFacebook: NewsfeedThe core of these apps and the core value stays the same to the user. The rest feels peripheral. I wonder what else they can make mainstream e.g. they push one thing for 6 months on every facebook page and that becomes a part of life, just like the newsfeed.. baaaaa i don’t know. I just think these companies, like yahoo, myspace etc… could have focused on one thing for 1 year intensely and pushed it into peoples minds. Preferably something that provides real value

        2. markslater

          its not a photo sharing service – its a bigger version of twitter.

      3. Neil Braithwaite

        Where have I heard that before?

      4. Tereza

        I see it as a not-so-good social CRM, not so much about photo sharing. and that you hit the nail on the head where it doesn’t decay as needed.

        1. Dennis

          completely agree re not-so-good social CRM – this concept is at the heart though of what we are currently building, sign up at the website below to learn

      5. $1331601

        I have to disagree. Facebook is where I go to keep up with friends that I don’t have a chance to keep up with in real life.I do upload pics sometimes, but more for personal storage and friend access. All albums are set to “friends only.” I let friends tag themselves.I use Facebook to interact actively or passively with people I know I like. I enjoy the stories they post, because they introduce me to new things or validate my own opinions. I enjoy their kid pictures, because I get a grain of insight into their daily life. True, if they were close friends, I’d already have that. But there are only so many people you can be close with, but an infinite number of people you can care about.I also go to FB to play games when I’m bored.I have no idea what I could use Color for. There has to be another product baking because if this is it, I’m sad that our industry apparently didn’t learn the lessons of 2002.

      6. Steven Kane

        would zynga agree with that?

        1. fredwilson

          i doubt itthey don’t agree with much that i believe in

  20. lawrence coburn

    I would go so far as to say that 90% of the social graph is, as of yet, unlit.I am so excited for the companies working on this problem.

    1. fredwilson

      unliti love that termi’m gonna use it

    2. Rupy Yuan

      “Unlit” has broader connotations than the companies working on this problem, there is I guess the opposite of the social graph which is an unconscious transformation of how connectivity rewires our consciousness.An opposite of a social graph is an “Aristotle” that serves as continuity through the ages that remains as essential, well past the life cycle of a social graph.The graph of future generations will be radically transformed more because the past remains an unlit graph that adds dimensionality to what we are now are able to see through the networked relationship.[$M.]

  21. Alan Wolk

    Given the general inaccuracy of the GPS feature on most smart phones (I’m thinking of how infrequently FourSquare puts me in the right location, particularly in Manhattan) I keep wondering how many initial users are going to wind up with pictures of people at the restaurant across the street or at a party a few blocks away.That may kill off Color right away (sort of how many people gave up on Gowalla when it wouldn’t let them check in from their homes or offices due to faulty GPS)A flexible social graph built on things like music choice or reading material makes more sense – it sort of exists at some level on Amazon when they show you what other people who bought the same book also bought — I have found much new reading material that way. And what’s great is those people remained anonymous – I never had to worry about who they actually were, just that we shared similar tastes. Photos are by nature a whole lot less anonymous. which seems like another strike against Color.

  22. Antonio Tedesco

    Colors sounds a little too CSI for me.I guess you need $41MM to pry 12 people out of Google/FB/LinkedIn.

  23. kidmercury

    ” Seems like most anyone can do that these days.” #ohsnapgoing to beef with the one graph vs many graphs idea. as with all these false dichotomies — i.e. right vs left, yin/yang, open vs closed, etc — it is not a question of OR, but of BOTH. once we embrace data portability and federation, we will get a better idea of how many and one social graphs can co-exist in a complementary fashion. this is governance layer stuff, though, and we don’t get there without doing the necessary political work.

    1. fredwilson

      i want to pair you with umairyou guys would kill together

      1. markslater

        agreed fred.Haque has sullied recently but i’ll tell you he was blowing me away in 2006.i”ll always remember his deck on the music business, and how he used the DJ mix analogy.brilliant stuff

      2. Mark Essel

        Umair Haque (@umairh) is such a rebel, is that who you’re suggesting? His ideas are even more extreme than my own on social evolution. Fun to read. I bet he and Kid share some world views on government.

  24. giffc

    I could not help but think of Eric Ries and the story / trajectory of when I saw the funding announcement.

    1. fredwilson

      yup. this money will be a huge burden for them.

  25. Renee

    I agree with the many social graphs theory. My boyfriend and I added each other on Twitter and on the night we met, and our overlap/interaction on those two sites got me thinking that maybe we were compatible. πŸ™‚ Regarding Color: I’m a little bit confused about why I want to see the photos of other users who are in close proximity to me based on location (esp after our connection is established). If I’m taking a picture of something on my iPhone, it’s intended more for the consumption of friends who aren’t there. I don’t really want to see a bunch of other photos taken by other users at the same event since we’re seeing the same thing.

    1. CJ

      Marketing is the only thing I can see. Imagine digital flyers or virtual billboards, but why would you want to use the service in the first place to access these things? The world may never know…

  26. PC

    Interesting but slightly disagree. I think this is where the news feed model is advantageous over the twitter model. It’s much easier to have an automated weighted graph deciding what content I get as opposed to all content being equally weighted. Perhaps the implicitness you speak of would be better manifested as graph weighting and not graph pruning.Lately I’ve found twitter to still be very valuable because unfollowing is so easy. I follow and unfollow people frequently enough such that I can manage my stream but it hasn’t become a pain yet. Perhaps what’s more annoying is te plethora of “irrelevant” tweets I have to scroll through. Twitter needs a filtering option where I can either view all or view highlights, which would basically be a weighted system similar to news feed.

    1. fredwilson

      a relevance filtered feed is certainly necessary for twitter. they know that.

  27. shotbeak

    Awesome idea! I’ve thought of something similar (not photo; but updates) that can be used at places where people congregate. A great place for something like this would be a music festival! People sharing photos and updates and videos will make it such a more interesting experience!

  28. karen_e

    Could someone link me to a tutorial? I need a boost. I’m kind of irritated that I am not yet making good use of any of the music-sharing sites.

    1. fredwilson

      what services do you use for listening?

      1. karen_e

        1. Webwise and on the laptop, I listen to music *happily* on streaming radio. I listen to music *frustratedly* on Pandora and There are 25 wifi signals or more in my apartment, is this relevant?2. iPhonewise, nothing is satisfactory except iTunes. Am i doing something wrong?3. In the living room, we now have an HDMI TV and DirectTV and an AppleTV so I should be able to gizmo myself into some happy music listening & sharing, possibly with video to boot. All links to tutorials are happily accepted.

        1. fredwilson

          you want to see which of those services allow you to publish your listens tolast.fmconfigure the ones that do to do that

        2. RichardF

          do you have the lastfm iphone app Karen? That works well for me for listening (I think the website is better for discovery)

          1. karen_e

            thanks RichardF, I think you nailed part of the problem – trying to use the app for discovery. to be continued ..

        3. Dan Epstein

          Hey Karen,Not a user (just never tried it), but I’d rec. Rdio.Grooveshark meets Pandora.They also have a new Mac desktop app.

          1. fredwilson

            I like a lot

    2. ShanaC

      me too! I am so you on that

    3. Andrew

      Karen– I’m writing one right now (hopefully out this weekend). I had the same experience– tried and didn’t see the point until much later. Short version is, hook it up to your iTunes and other sites and wait a few months. The recommendations get better.If you’re interested I wrote four other parts to the series, covering my favourite music discovery sites. I hate directing people to my own site in comments (it feels spammy),but maybe you’ll find it helpful. Services covered so far are Hype Machine, CBC Radio 3, Soundcloud, and Hope it helps.

  29. Morgan Linton

    Holy moly – 40 million! Doesn’t that seem like an insanely high amount of money to raise for something like this? I’m not saying this isn’t a great idea, it definitely is, I’d just be interested to know what is the typical amount of money raised by companies in this niche? Is it a ton of employees that make the number so high?Thanks for sharing as always Fred, I’m going to be playing around with the app today.

  30. Matt Kritzer

    What I don’t get is why we can’t have multiple graphs within the same social service? By this, I don’t mean inside an aggregator, but within a specific social network. I really don’t want to discover or switch between multiple services constantly to engage the right audience. And as this fragmentation of social services continues to proliferate, is the general public really going to come along for the ride? It seems like switching between services also amplifies the need for curation. Perhaps it’s a future mater of acquisition and merging of services.I love twitter, but feel like it falls down a bit on the job in this regard as well. I don’t see why I can’t have multiple graphs inside twitter. This is what I feel twitter is missing:- Allow users to “self-curate” themselves by classifying content into topics before posting and/or selecting appropriate audiences.- Allow users to create different lists around topics, not just followers.- A tweet feedback loop that gets anonymously fed back to the content creator. This would be in the form of a quality score that would allow followers to rate a tweet’s usefulness. I can’t tell you how many famous tech personalities I follow on twitter that constantly push promotional tweets on their followers. I don’t want to unfollow these people because they do mix in an occasional great nugget, but I don’t want to troll through shameless self-promotion either. They need to get the message in a micro-sense, not the macro number of followers-sense.Seems to me that there’s much more than can be done on the machine front for curation before injecting the human element.

    1. kidmercury

      facebook has multiple graphs within the One graph. not really a fan of fb’s approach but i do think the concept of multiple within One has merits and will be something we see much more of.

      1. Matt Kritzer

        Yep, that’s true, they do, but it seems very few use it in that way. The graphs are walled off and standardized in format and function, which I don’t think will allow it to scale in a specialized sense (as currently constructed).

  31. ErikSchwartz

    Have you folks seen the satirical Color pitch deck?

    1. fredwilson

      yes, it is great

      1. Robert Verwaayen

        i especially enjoyed slide 45 – we’re gonna flip this b#(@ like flipper

  32. ShanaC

    Personal Opinion – I doubt I will use color. I’m already sensitive to my 4square usage. The idea of having an implicit social graph around picture locations where I am now just seems like I am revealing too much….

    1. falicon

      Did you see Brad Feld’s recent post related to foursquare sensitivity?… … horrible news for foursquare but def. something people should be aware of…

      1. fredwilson

        he either was tweeting out his checkins or he allowed people into hisfoursquare network he should not have

        1. falicon

          Yep – inevitable that someone would run into this…just stinks because it’s the type of negative people latch onto in their heads (even if it’s tied to user error or very tiny use cases).Even the foursquare service has to deal with spammers and low-quality marketers trying to game the system…the dark side of the internet revolution!

      2. ShanaC

        I’ve been thinking about this for a while now, though I haven’t seen Bradspost until now. I tend to check in as I walk out, and I don’t check intoevery place I go, to make patterns less discernible.Also, personal opinion as well – the easiest way to get over stalking is tocall it out, as a public.

  33. Denim Smith

    hmm.. .I have a really hard time with all the social graph talk. People are people and things change. Dust off the crystal ball for a moment and ask yourself how and why humans will interact with one another in 200-1000 years – unknown technology notwithstanding. The evolution of instant and global communication won’t necessarily lead to a *relationship* (even passive) with every touch point/ person and every interest out there. And I am not saying that a digital social or interest graph sliced and diced will not be valuable to someone (not necessarily the user). But such graphs are not static because we’re not static and discovery is saturated into our DNA – like finding that crumpled $10 bill in your pocket – and we were born to wonder and explore because of its intrinsic rewards –> development and personal growth. This is precisely why Google is so successful. So what benefit do I get from posting to Color (look at me everybody!)? What benefit will I get from being a voyeur on Color? The law of diminishing returns and 24 hours in a day suggests I cannot follow everyone on every platform and vice-versa (and now venues?). I don’t know about anyone else, but I do more in 24 hours than ever before and clicking through stranger’s photos from a venue just doesn’t cut it for me (why? Because I was there in person with the people I wanted to be there with). And, I have a daughter and I want to maximize my time with her. Sorry folks – life’s too important and way too short to waste on such frivolous things. We’re talking about photos after all and there is no shortage of sharing apps out there – private and public – but Color is how you share with strangers and/or a venues management? No thanks. Now go check out every photo from SXSW and call me in 3 hours.

    1. fredwilson

      hardly frivolous. this technology has already had global political implications

      1. Denim Smith

        Apologies Fred but my rant took a life of its own. My point was to say that this is a feature and not a company and that’s a lot of pre-launch capital. For them to not have had a presence at SXSW is odd considering the $41MM and reported 27 employees. After all, isn’t this going to be the Placebook?And just so I’m not misunderstood, the Internet was the technology in the political issues and not brands – and the issues were not global just because we could watch them or read tweets from across the globe. Revolutions aren’t correlated with photo sharing start-ups and won’t be increasing because Color launched. So we’re in agreement that the technologies were not frivolous – no question about it. My reference to frivolous was another frivolous attempt to drill deep into each location and every second of our natural world and pixelate and populate it for the benefit of an entity and that doesn’t make sense to me at all – I have better things to do – who are we recording for and what are our rewards for building an image-based place wiki? We need to know the rules of the game up front and the company needs to be transparent. Its just interesting to flush it all out in a community like this. Color will do well but won’t be for everyone.

        1. fredwilson

          i’m not even sure it will do well based on my use of it yesterday

          1. Denim Smith


  34. Michael Weiksner

    I used to LOVE, but I abandoned it a few years ago. Like many collaborate filters, it got to know my tastes TOO well, and then it wasn’t suggesting enough new songs for me.For optimal serendipity, you need a mechanism to occasionally leap well outside your comfort zone. You want to occasionally jump to a user with almost no overlap with you but is an opinion leader on a totally different continent. Potential guides to new genres.Facebook is particularly ill-suited for this role, because real social networks are very dense. Dense networks act like cul-de-sacs, dead-ends. This topology is terrible for discovery (see… ).Herein lies huge opportunities!

    1. markslater

      i completely agree michael – one of the reasons i started reading AVC was freds unabashed love affair with – me too. But it jumped the shark for me in 2008 or so when it just fell back in to the noise……

      1. Matthew Hawn

        You should give another shot. (Disclosure: I work there). The number of ways that we’re letting you explore recommendations are better now than they were in 2008.Here’s how they break down:ARTIST radio and TAG radio: These two tap the collective community to recommend music that’s similar to the music “seed: you give it. Tags equal genres here. It’s powered by more than 50bn scrobbles worth of data now and growing by a thousand scrobbles a second.Then there are the Personalized stationsLIBRARY radio plays music you already know from your listening history.RECOMMENDED radio plays music similar to what you already know but haven’t listened to yet.MIX radio (launched late last year) is a mix of the two. Stuff you known but also peppered with new recommendations.Social Radio:FRIENDS radio plays the music your friends are listening to. NEIGHBORHOOD radio plays music for people whose recent listening history looks like yours.Give us another shot. We’re planning a lot of renovation on this coming year – most of it is focused around the ideas that this blog post is exploring – and I hope we’ll bring back some people like you who haven’t been back in a while.

        1. markslater

          hey thanks for the response and the effort matthew – i’ll give it another try.

        2. fredwilson

          i use as much as i ever my library and neighbor radio on sonos is the very best “i don’t know what to listen to right now” radio ever

    2. falicon

      I agree so much with this…been thinking about it a ton as well. It’s a very difficult thing to do right, but that just means it’s going to be massive when someone does figure it out.

  35. Jim Adler

    I’m a “many graph” believer. We humans are just too nuanced for one size fits all. As a step in this direction, I defriended my business colleagues on Facebook (& reconnected with them on LinkedIn and Twitter) to create more of a “site barrier” between my personal and business lives — two graphs, three sites and, no doubt, more to come. Earlier this week I posted on my rationale and mechanics

  36. Cost2Drive

    agree that it needs to be implicit as it seems I’ll never get around to explicitly cleansing my twitter and facebook accounts. too much work.btw – did you drive or fly to spring break?

  37. Ido Salama

    1. looking at pictures of random people gets boring really fast 2. looking at pictures of friends is addictive3. $41M – I want to see this on the SNL “REALLY???” segment

  38. Philip E. Reichen

    First time post, long time reader of Very good post Fred! A couple of friends and I started building a location-based Q&A platform a while ago which we only recently turned into a startup. The amazing part is that if I read your posts, and especially this one, it enables me to call the “things” that we build by its name! This post is like a blueprint for the house that we already built. Thanks for sharing these awesome thoughts!And yes, I totally agree with the fact that soon enough we won’t “know” on which social graph we’re on right now because it will drift away from being people-centered to being interest-centered and ad-hoc.

  39. Mark

    After reading John Battelle’s post, I can see where people could really like color. I’ll check it out, but really have no personal interest. Yet, I’m willing to bet it, or a competitor, will be huge. Geez Foursquare should just lock the doors and crank this out ASAP. They would have such a headstart.That said, I can’t help but think that there is a boyband effect at work when you have $41M to start.

    1. sigmaalgebra

      So, you are saying that for the development Foursquare needs to hire about two dozen hackers, for that get another equity round, and there have a post-money evaluation high in nine digits? Or ten digits? Eleven?Break out the 12 year old barrels! Right, the 12 year old barrels of soapy WonderBubble! And find the switch to turn on the bubble machine!Then do an IPO! Don’t be in the doorway when the lockup period is over! Drain those fat boomer IRA accounts!

  40. markslater

    i am so happy you wrote about this.firstly – i am totally with you on social graphs.6 years ago a big wave washed over the planet forcing people to re-connect and stick their heads out in to the digital public. 95% of those connections made were “hey…..long time…..nice to hear from you………………….now lets get back to having nice lives apart…”you cant capture a social graph, and then assume it applies to perpetuity. It doesn’t. because people move on, they have different interests, and they go through changes. trying to drag your social graph along or in to your interest graph is a mismatch.why would an old boarding school buddy living in Hampshire be interested that i am looking to eat Italian tonight for instance?where your interest graph involves a purchase (which – shameless plug – is what Getabl is) then its likely going to have local implications in terms of who you want to share with. People who i share with on getabl are friends and local – it allows them to participate in my actions.So interest graphs are absolutely the new frontier – where they involve commerce they’ll be most interesting locally.

    1. fredwilson

      i’m glad we are on the same page. it bugs me when we are not.

      1. markslater

        now now fred – leave the sarcasm to us brits πŸ˜‰

  41. Mark Essel

    Agree with implicit graphs, I’m tuned into the same wavelengths and just wrote about implicit interfaces.Default inbounds are last decade, our attention changes and is our most precious asset for learning and contributing to projects we care about.I came up with an app concept just to do dynamic fast swapping of my twitter default follows. It’s in the half coded broken stage, but should only be a few hours away from going live. It was born out of frustration from manually moving everyone from my follower list and added them to a list (called inbound). The process was slow and somewhat painful and error prone, and won’t work with people who follow more than 500 due to the list limit (weird limit). The good part about doing it manually was being able to curate/modify the list as I went. What I really want is a hot swappable input channel. Now I use lists for that but when it comes to third parties that expect a “default graph” which is just about all of them, it’d be nice to hot swap or select a certain list.

  42. Aaron Klein

    @robgo points out this morning that it COULD be possible that a $41MM-funded 27-employee pre-launch photo sharing startup could actually be an early April Fool’s joke.Let’s pray that it is.

  43. markslater

    so i just got a picture of some dude holding his uknowwot out for the camera and he is less than 150ft away?stella – inside now!

    1. fredwilson

      chatroulette all over again?

  44. nickhuhn

    I don’t want to clutter these great comments with a product plug, but amplifying the implicit social graph is the underlying assumption behind a social commerce startup I’m building with some other smart people: only is this the next wave in social networking, this is the next wave in business. We’re pretty excited to help reveal implied or tacit intentions as they relate to peer-to-peer commerce.

  45. Steven Kane

    seems there is a type of person — or is it a stage of life? — that thinks the world is better, or more interesting, if experienced through a technology filter, that is always searching for something that (in the words of U2) is “even better than the real thing”

  46. Bart Denny

    Late to the discussion here unfortunately, and its so crucial to our company. Just days before, we launched our app Pixamid, which is also based on the idea of sensing who you are with to share photos with.We go about it by tying into Foursquare, Facebook and soon calendar. But our core model is closed sharing (share just with your friends here), not the open “share with everyone here model” (although users can do that). So a user may have to make a few clicks to add non-Pixamid users to the photo set, but that gives comfort that photos are not public.Now, as a bootstrapped startup, we want to do a lot more than what we have now (yes, Fred, Android is one of those, sorry). So we’re pretty happy with v1, and will make it much easier to share easily with the right people in the future.But most importantly from our point of view, we try to solve an actual problem we ALL have – we hate to organize our photos. We aim to solve this by federating in photos from whatever services your friends us, aggregating them around the event. Right now, works like a charm with Instagram – you get anyone’s Instagram taken at the venue, while you are there.In beta with us: add your flickr + foursquare account to let us retroactively organize photos around event. (this available only on the website now, which is available only to app users. hope to open up more widely soon)And from there, you have this great timeline / event-based view. Then you can start telling rich stories by adding a few events together.Sorry for the long pitch, but clearly we’ve got some thought in this area, and the hype that Color got today was frustrating, when you know you have a much better product! Blogged about that here:, I think you will find Color fairly baffling from a UI perspective, intriguing if there are other users around, but ultimately not solving any concrete problem. More private sharing really is key in the photo sphere. BBTW, thanks for the chance to pitch Albert & Christina in an elevator at SXSW. You probably heard me do a decent summary of our ideas in 40 seconds. Try Pixamid on the office iPhone, I think you’ll find it much better than what $41M can buy you.

    1. fredwilson

      any idea when you’ll have an android version. i love the sound of what you are building

      1. Bart Denny

        traction, funding, android… $41M gets you on multiple platforms quicker than our bootstrapping operation.As the first 2 are now rolling, earliest we will have an Android is in 3 o4 4 months. I promise you will be one of the first to get it!

        1. fredwilson

          but apparently $41mm doesn’t get you a good android appi expect yours will be better than theirs

  47. Gregory Magarshak

    Wow, this is pretty cool. The way we build social graphs is definitely in need of a revolution. Seems like people are moving in that direction now.Facebook and twitter are cool but there has to be a better way to connect People, Places and Events (and other things) that is more intelligent than just “I want to be friends with this person”. It should contact YOU when there’s something of interest. And it should be on your phone.We’re working on it πŸ˜‰

  48. Dave W Baldwin

    Doing quick read. Color follows the diary hook offering ability to shoot photos and they are placed on a grid recording location/time/known characters moving to unknowns. The unkowns would be teasers sent saying you’ve been tagged.This design is more real time than the one unveiled last year placing websites on a grid that could be called up per subject, the hook showing attending Angels and VC’s their websites for the ‘ooh and ahh’ effect.Color is good in one way offering the tags to be brought up, but it still is present/past. It involves however many steps.Still think $41mm is too far out there in investment, you give too much and you end up with a lot that isn’t rev producing. OTOH, bare level price points, you can offer this for $1 and with however many 100mm smartphones get the investment back…move down to $.10 p/month where the user okays the deal and so on.Good luck, but this is not the ultimate social graph that moves in real time working past/present/forward.

  49. Kashif Shaikh

    A generated/dynamic social graph based on other factors than relationship, tries to bring strangers together, and I’m not sure that is a good thing.Can you imagine that me living in Canada has the same interest in some music as a guy in India? What value would that ‘social graph’ bring, other than ‘we share the same interests’ and it makes the application easier to find music that ‘my dynamic social graph’ would like?

  50. Kashif Shaikh

    How would Color work in a public area? Imagine taking a picture of yourself in the mens bathroom, only to see other pics of men in same situation.

    1. CJ

      Damn – it’s Chatroulette!

  51. choicr

    Yeah we are creating the social graph in a different way, but I am digging color so far. Sitting in the office and a few random people showed up in my ‘graph’. Pretty neat! Check us out too. We are looking for that ‘dream team’ and are moving fast!

  52. Cameron Brain

    No doubt someone already said this, but the thing I’m most excited is getting connected (albiet loosely) with people in proximity to where I am. Fred, you were talking about, and someone else here was talking about Disqus. The difference (and I know you already know this, Fred) is that Color is mobile. I’m not in front of my computer. Like a little gopher, I can pop my head up at a restaurant, party, concert, airport, etc., and actually look for the person that just took that pic. Point being, there’s an opportunity to actually connect with real people, in real life.That is really exciting.

    1. fredwilson

      that’s one of the many reasons i love foursquare

      1. daveevans

        At first reminded me of how google maps shows nearby images without the map and adding sharingTemporary social graph is fascinating. Let Color create the graph for us and blow it up based on proximity (archiving, timeline?) Better than the awkward “suggested friends” on FB and Twitter.Wonder what the emerging use cases will be for Color? share photos at festivals, gatherings, etc. I guess it could be fun to try and identify people from the stream nearby. But how fun is that after you do it 10 times?LOL @ Chatroulette all over again.I like Assisted Serendipity, shows then the gender balance falls in your favor at your local bar based on Foursquare. That gets me off the couch every time. Very strong use case. But location-based Instagram, not sure.I do like the app, very smooth, although groups are a big ??? right now as to how they work or why useful.Let the kids and the pervs use color for a few months, then we’ll see what’s up.

        1. eyesparky

          Some of the streams in and around hotels could get interesting. Joking aside, it will be fascinating to see how things shake out over the next few months, hopefully with the apps having iterated to something a little more polished. They are going to have to address some of the issues around the underlying use of contextual data with regards to privacy (the TechCrunch comment about Color using all the sensors on your device, including microphone to pattern match sound to other devices to discern co location, certainly made the hairs on the back of my neck stand up).

        2. markslater

          right dave – lets the kids and the pervs figure out the product market fit.

  53. Mark Mayhew

    I just joined Color and, from what I’ve seen so far, it’s a worthless site/network

    1. fredwilson

      i haven’t had much success with it today

  54. awaldstein

    Disqus is on the top of this thread with 21 likes. It’s been here before.Be great to have a post where they field questions from the community. There are a lot of fans here with a lot of ideas.

    1. fredwilson

      i know that Ro and Daniel read these comments

      1. awaldstein

        Sure they do…they have a real base of believers here.BTW…just saw your Tumblr pics of Aspen…serious snow and ski envy. Enjoy.

    2. Ro Gupta

      we haven’t done that in a while on our own blog so maybe can do another one soon. but you’re right, this particular community tends to grok disqus better than any. sometimes even better than those of us that work here.

      1. awaldstein

        Ro…on a side topic, articulating ‘why disqus’ to new users is still tough.If you comment a lot or have a lot of comments on your blog, Disqus’s value is self evident.I’ve suggested to a number of wine blogger friends to give this a shot. And it’s a self fulfilling prophecy of sorts…If you have Disqus installed the subsection of the community that is interested in your topics will appear over time.However being able to articulate above is a challenge. If you can, you can speed up the adoption curve.You guys and your product rock.

  55. Eric Friedman

    I think this is a powerful idea and one that is basically an elastic social graph. I blogged that Google as the social network that is there when you need it, and not when you dont.

  56. Peter

    Great Post Fred. We’d love to have you visit the Color office next time you are in Palo Alto. Clearly you see what we are trying to do.

  57. John Rorick

    The other downside of all this press and cash is risking becoming the Ishtar of consumer apps. But business school classrooms will benefit from years of case study, and they could retire the annual AOL/Time Warner merger critique. Dustin Hoffman FTW!

  58. Derick Rhodes

    As the founder of a very new start up playing in this sandbox, Listgeeks (, I’m 100% preoccupied with all things social recommendation/social graph-oriented. I think it makes sense to describe where things are heading, when we’re talking about companies like foursquare, Color, or any of the numerous recommendation-oriented platforms (Hunch, Quora, etc.) in these terms: We’re building towards intuition.It doesn’t really matter to me what the actual goal of the app in question might be at any given moment, the thrust is the same: We’re building tools that make it increasingly easy to provide people with things they want (information being perhaps the most important) with minimal effort on their part. The level of required effort on the part of the consumer is decreasing as the quality/accuracy of the experience/thing provided is increasing.Eventually our apps will function essentially as intuition, but better. There will be apps that intuitively install and then merge data/information from an array of tools specifically tailored to you: where you live, how old you are, what you like to buy, what kind of of music you like, etc.The experience they provide will be so invaluable, you’ll essentially have no choice but to want them: they’ll know you want them before you know you want them.We’ll be able to rely on these systems to give us exactly what we’re looking for (or to tell us if what we’re looking for will be available anytime soon, and where), without our doing much of anything.I think that’s what we’re all building.

  59. apartmentsdirect

    While this is not always the case with the blogs listed, I love where they are going with in lanzarote

  60. Pstoppani

    Color is wasting $41M… OMG, it is beyond terrible in terms grabbing user attention!

  61. Dan

    Were you successful in getting your family on it and try? What are your thoughts after trying? P.S. Disqus could use a little work on the mobile experience (typing this on iPhone) although bet they have been working on this for months already…

    1. fredwilson

      the experience hasn’t bee very good so far

  62. paramendra

    Google Images, Facebook Photos, Twitpic, Instagram, FoodSpotting

  63. Nadim Hossain

    This raises an interesting question — what does the Implicit Graph mean for online shopping? Just like your music example, I care about what people like me are buying on Amazon, potentially. I may not know them, but if I’m a professional photographer, I really care about what equipment other photographers are buying. Ultimately, the mashing up of one’s social graph, implicit graphs and commerce will make for a better user experience.

  64. 24hoursplumber

    I have long desired to be able to select definite people and have their comments automatically highlighted whether i am here or at cdixon or msuster.

  65. Luke Iseman

    I think the key to delivering the right implicit social graphs to the right services is weighting friendships: my 500+ Facebook ‘friends’ aren’t the same degree of closeness offline, so why should they be treated that way in the digital world? This is a big part of what I’m trying to do with WeTique, my (garage stage) startup.One of the uses of the weighted social graph we’re creating will be to recommend the relevant subset of ‘friends’ one should add on a given social service by analyzing one’s relationships as well as one’s interests. With our friend weightings, this will be straightforward to achieve more effectively than I’ve seen any other social service do this; analytics on user behavior in response to our recommendations (do you actually add the Facebook contacts we recommend for location-based services on Foursquare?) will enable us make this world-changing:)It’s very much proof-of-concept right now, but feel free to give it a try at . This version delivers recommendations based on 1. what your friends have liked on facebook and 2. the weighting you place on each of those friendships. Over the next 6 months, we’ll utilize Mahout, create ‘lenses’ through which to view other social sites, and some secret sauces that will make things really get interesting.

  66. Sunil Madhu

    Back in ’07 when my buddies and I started tinkering around with tools like Jena and languages like OWL-DL, we recognized the power of the semantic web and machine based reasoning to be able to expose the connections between things that would otherwise remain hidden to the naked eye. Hidden in massive social graphs like Facebook and Twitter, are billions of attributes about people, things and places that connect us. When we meet someone new, its often the things we have in common that attract us but the differences that ignite the flame of curiosity. Machines are now able to help us discern these subtleties in connections. Knowing what connects us and how is cool. But all connections aren’t created equal. Knowing that for example, my best friend and I are connected through hip-hop music is one thing. But learning that my best friend influences my hip-hop music purchasing decisions because I trust his taste in classic hip-hop music more than say 10,000 reviews from people I don’t really know or trust, well, that’s more interesting.Color, in essence, distills down Facebook and its massive social graph with a focus on photos. A lot of market research has shown that 2 of the top 5 things people do on most social networks is share photos and eyeball shared photos. So dispense with the textual nature of life-logging and use photos instead because pictures say a 1000 words. Not ground breaking. Color aggregates photos around physical locations — geo-caching photos. Not new. Color attempts to figure out where you are. GPS is inaccurate at best so in densely populated locations like New York, you can’t really tell if all the photos you’re seeing are pinned to the place you’re actually at or nearby. Does that matter if you are constructing a social graph of photo experiences centered around location? What happens if the happy smiley faces Color shows me at the restaurant that I’m near which presumably is supposed to influence me to gravitate to that restaurant, actually turns out to be photos of another restaurant nearby?What about noise? Increasingly we are inundated by services that allow us to express ourselves in volume but not necessarily in quality. In fact, the way we communicate is becoming increasingly distorted and lossy making us more malleable to bias. $41M for Color is certainly brow-raising. Perhaps they have weaved some cool story around photos + social influence + object recognition (like Google Goggles) to help investors part with their monies. Either way, Color is only pushing an envelope that was already set in motion by others before them. Wait till the hype dies down. Then lets see who’s pushing what social graph envelope.

  67. paramendra

    The Color Social Graph Might Work Better For Books, Movies, Music

  68. Alex L

    I know I’m a little late to the game, but the implicit social graph has another big advantage you haven’t mentioned: the ability to meet new people. Facebook and Twitter are great for connecting with people you’ve met or heard of. But they’re not good at meeting new people.I work on a wordplay game,, which more or less has an implicit social graph, and about 100 core/power users have developer insanely strong friendships. They’ve even started meeting each other in person.It’s been a wild experience, creating a creative wordplay game, and watching the power users make friends with one and other. And the payoff of seeing these awesome people make new friends is worth more than anything else, at least for me.Anyway, great post, and I totally agree. We need a way to make new friends that have similar interests, not just stay in touch with the real-life people we already know.

  69. fredwilson

    that’s a good point Kirill

  70. cyanbane

    “did a check-in on FourSquare at Bergdrof-Goodman (men’s side) and has a Disqus profile with strong conservative sentiment,”Are you basing the decision of ‘strong conservative sentiment’ on the content of the sites the user is commenting on or the other “neighbors” that a given site determines (ie Are you relying on the niche site for the algo (that you can assume they know well) or your own. I have to think this is similar to what Google has to do with personalized search results.

  71. Lee

    It only needs to be probabilistic the first few times in a given scenario.Once the individual acts on a given message (or doesn’t) then you start to build the deterministic data you need next time.

  72. sigmaalgebra

    What do you want to estimate, a ‘social graph’ or an ‘interest graph’? [Note: Here we are using ‘graph’ in some ways not in the applied math subject of graph theory.]It sounds like you want to use an estimate of a social graph to estimate an interest graph to estimate ad click rates. So, have about three steps of relatively noisy data, and generally we have to suspect that the result will be too ‘noisy’ to be very effective for the intended ad targeting.If what we want is an interest graph, then we should go for an interest graph more directly: Social graphs, restaurants, other people in the same restaurant, clothing stores, other people in the stores, etc. are very crude data for ‘interests’, and here’s why: ‘Interests’ involve ‘meaning’ of content, human ‘intelligence’, etc. and are tough to handle effectively with software. That, say, ‘circumstantial’ data will do very well getting at ‘interests’ is a bit far fetched.Or, is the ‘social graph’ really more about just the ‘social’ itself? If so, then for Color I’ve already heard the word voyeurism. “Why is that person at the table in the corner wearing a paper bag over his head?” For people, in the past ‘social’ mostly meant ‘make business connections’ or ‘romantic connections’. Maybe Twitter and Facebook are examples of more variety in social connections. Whatever, the ‘monitization’ of the purely social, that is, without ‘interests’, is vague.Since “people are social animals”, I can’t say that the purely social can’t lead to monitization. So, to me, Color looks like ‘participant observation’ (from sociology) research. Or here’s $41 million: Use the TIFO method, try it and find out. So, use “ready, fire, aim”. For something closer to “ready, aim, fire”, I thought that…had some merit.I would also have a broad objection: “Without a destination in mind, any wind will take you there.” or some such from Mark Twain. Or, “Don’t start vast projects with half-vast planning.”. Or, for a project, let’s start with a clear goal, i.e., a specification, of what this project is to do. I know, a current view is that a specification is hopeless because “A startup is a temporary organization to find product/ market fit” so that the $41 million is to find a specification, not to achieve one. So, Sequoia is funding participant observation (see ‘Tallys Corner’) sociology research; who would have thought? Such things have been approached with applied math — J. Coleman, P. Rossi, D. Entwhistle, and others. Even in ‘sociology’, the mathematical approaches have much more respect. The usual steps are (1) participant observations, (2) guesses at what is going on, ‘social theory’, (3) statistical tests of the social theories, and (4) mathematical models. Sequoia is starting with (1) and has a way to go to get to (4) and estimates of click through rates.In more detail, for the corresponding data manipulations, we’re working merely heuristically here. Instead it is more solid to say what we want, what manipulations will approximate that, and why the approximations should be close. E.g., we’re trying to estimate ‘click through rate’. Okay: Let’s see the math for that estimation and the derivation of the good properties of that math.From 40,000 feet up, I believe that real value for users and monitization are both more promising pursuing ‘interests’ more directly.For ‘social’, that sounds like trying to get something substantive out of what happens in high school, passing notes in class, texting, glances in the halls between classes, seating arrangements in the cafeteria, gossip, etc.

  73. PhilipSugar

    Is it really not violating my privacy??? Because it sure feels like it to me.My test would be could you put what you’re doing in bold print with easy to understand terms on the home page and not have people care.For instance would you put?We are using your browsers unique settings to identify you to gather information about you, to market to you. We’re going to connect it to other social graphs which can identify you but we won’t”The consumer is in control. I think we’re in a period of transition where technology can be used to try and circumvent that, but it is coming.I think there is going to be a good opportunity to do just the opposite. Separate all of your social graphs, and insure people can’t connect them.I believe this because it mimics real life. For instance: I don’t want my work and my religious social graphs inter-connected. One of my partners is serious atheist. I am a devout Catholic. I belong to a Catholic Charitable Fraternal Order. Do I want those social graphs connected??? Heck, I think there are people that read my comments and my blog that really didn’t want to know that information and might think less of me because of it.Do I want people to be able to connect my social graph on my boating activities and understand what I do with my free time?

  74. markslater

    its a violation of my privacy that you think you can push anything to me without me gesturing that i am “in the market to be marketed to”.just because location, time, and profile give you more “context” does not mean that i want a better targeted ad. who said i want an ad?

  75. ErikSchwartz

    Lala started as a CD trading service. This team knows how to change plans and they have a long runway.The problem is they can’t settle on a really good, but merely medium sized business. Assuming they raised this at 120MM post, and also assuming there’s a liquidation preference with some multiples (at this valuation it seems likely) they’re needing a giant exit before the founders/employees see anything.

  76. fredwilson

    why not make it free so everyone in the world can read it

  77. ErikSchwartz

    At some point a founding team and company is going to get its start in these comments.

  78. ShanaC

    !!!! Really !!! That could be so cool…(ok dorky that I like having my comments on tech and my fashion searches integrated together)

  79. Dave Pinsen

    It seems like there are more direct and immediate ways to reinforce pleasant behavior on the part of a Starbucks barista: reciprocating with your own pleasantness, and putting a dollar in their tip jar.

  80. fredwilson

    cool idea

  81. Tereza

    You mean the “we’re not sure how it’ll make money yet” part? Or the “maybe banner ads” part? πŸ™‚ They can — and should — do better than that.That said i’m trying it out right now. Cool…for 5 minutes…a fun party feel. I look forward to trying it outside of work.2 star rating in the App Store kinda stings though.The App descrip though — use on playdates. Um, throw pics of my little kids on a live geostream? “Yoo-hoo! Hi pedophiles! Anyone nearby? Here I am!”My advice would be stick some parents on the team to work that out a little more carefully.Also curious about naughty photos. Are they screening those out? Any hotel stay is gonna be like a proximal ChatRoulette. Will be interesting!

  82. Tereza

    I get it. I know that. Just thought it important to express it.

  83. ShanaC

    That would be interesting. Truthfully, I’m lucky to have met some of the smartest people I know here…And it would be so insightful to read all these posts and comments 20 yrs down the line in order to see what happened..

  84. ShanaC

    I’m with you on that. I think we should just include gratuary or give waiters an actual salary…

  85. Robert Holtz

    LOL! You will be assimilated. Resistance is futile. πŸ™‚

  86. andyswan

    It already has.

  87. Donna Brewington White

    A “dream team” could be assembled here.

  88. kidmercury

    YES! thank you for reminding me charlie, i almost forgot to bring thisup in fredland. for those unfamiliar with this story, do yourself afavor and search the web for “defendant convicted of minting his owncurrency.” would drop link here but disqus eats them too often.(serious problem, disqus)people, this is a HUGE story! but anyway, it doesn’t matter what thefeds try, with the right approach, we are going to win. i encourageeveryone to be a leader and join the side of the winners. roll yourown currnecy and/or use alternative currencies. more and more peopleare waking up. there are bills of legislation in MULTIPLE states inthe united states that call for the issuance of state currencies,local currencies. time to get educated and take the power back! thereward for doing so will be a genuine bull market and a foundation forlong-lasting economic prosperity for all.

  89. Rocky Agrawal

    this is a logical extension of what foursquare did with AMEX at SXSW. credit card companies have been sitting on tops of giant piles of incredibly valuable data for a long time and have done nothing useful with it.even the special offers i get printed on my amex statement are usually pretty far off target, despite having 18 years of my purchasing data.i did a blog post in which i talked about what a foursquare/amex co-brand card could do.

  90. Adrian Meli

    Interestingly, though, the high vauation will make this a self fulfilling prophecy given how much exposure it has gotten them. I wonder if some of the high dollar raise was to get extra exposure so that they create the killer app and Color competitors are left in the dust. In other words, some of the extra $ could be seen as a marketing expense and justified though $40mm is clearly veyr high…

  91. Universal_Mind

    I went to your blog. As they always ask start ups.”What problem are you solving” ?

  92. CJ

    Dating? Is that where this is going? That’s what your comment on relationships (undiscovered) makes me think. Think about this – you go out to a concert, take a pic of yourself and ‘post’ it on color, someone else sees it, gets in contact with you, meet up, etc. You assume you have at least one thing in common if you met at the Bon Jovi concert.

  93. Rocky Agrawal

    No, not really… only to execute. :)Companies like AMEX have all the data they need. But generally they didn’t think about these sorts of things and have limited ability to execute on it. I had come up with these concepts while working for a large Internet company and couldn’t push it through due to internal resistance.The really interesting parts of what Color is doing (ignore the photo sharing — that’s a sideshow) could theoretically have been done by the mobile carriers. They have had tons of data in their systems to make such connections.

  94. Rocky Agrawal

    And, yes, the 4sq demo would be a nice win for AMEX. Their customers tend to skew older.

  95. daveevans

    Glad to hear dating brought up, that is my wheelhouse and I have been talking about leveraging social graphs, parsing Facebook feeds, mining for taste, preferences and personality for years now. $1.5 billion dollar industry in the US and only 40-ish percent adoption rates.Show me what you Like, who you hang out with, last few concerts, photos, conversations, etc. Dynamic data instead of set-it-and-forget-it dating profiles.Instead, everyone is trying to have “friends match friends” on Facebook, sigh.

  96. ErikSchwartz

    Perhaps it’s 40 mm in 3 tranches. 5MM now, 10MM upon some milestones, 25MM upon some more milestones.That way you get press coverage on CNN and TechCrunch and the huge free PR bump for raising this monster round but you’ve really only taken 5MM (with a commitment for 35MM more).

  97. falicon

    having the contact info shared in the comment would be *so* funny if you had started this comment with “I’m single”…btw – my guess is that it was really just a glitch in the email-to-disqus send…hopefully my response will bring it to your attention (in case you want to use the ‘edit’ feature). =D

  98. markslater

    100% charlie. i wont be. i already dont use any of the check in guys because of this very approach. And i am going to hazard a guess that the traction from this approach is questionable at best.

  99. PhilipSugar

    I think you are right you have to decide where to draw the line. I think the example is a little flawed as you access data that resides on my system. Kind of like downloading a unique key from the Coutach’s computer.Also just because it is legal it isn’t necessarily right.For instance I know how I feel if somebody took pictures of my kids outside my house and posted them on color with the geo-code.I bring this up because I live in a very quaint town that is over-run by photographers on weekends. People take pictures of my kids which is legal, but I can tell you what I feel.