What If Web And Mobile Apps Are Like TV Shows?

I was having lunch with a veteran of the entertainment and the video game business this past week. It was an interesting and wide ranging chat. One of the things we discussed that stuck in my mind was the thought that web and mobile apps might behave more like TV shows than traditional software applications.

I've watched my kids go from myspace to facebook to instagram over the past seven years. And who knows what social app will be their "go to app" in five years. This has always been the case in videogames. Farmville to Cityville to something else. Words With Friends to Draw Something to something else.

This round trip from nothing to everything to nothing again is also true at some level with many tech companies. Digtal Equipment Corporation was founded in 1957 and shuttered in 1998. RIM was founded in 1984 and in all liklihood will be gone before the end of this decade. Same with Sun Microsystems, Silicon Graphics, and many more iconic tech companies.

This concern or observation depending on where you sit has wide ranging implications for valuations, returns, and many other aspects of the startup economy. Companies are worth the net present value of their future cash flows. Said another way, if you knew that a company was going to earn $1mm a year for the next ten years and then be shut down, there is no way you'd pay more than $10mm for that company and certainly you'd pay something a bit less than that.

There are web and mobile applications that seem more immune to the "here today gone tomorrow" concern. Utilities like search, email, calendar, document store, etc feel less likely to be subject to this issue. YouTube also feels fairly secure. But purely social apps, the ones that depend on having your friends on them, seem quite vulnerable to a mass exodus. RIM's demise among my kids' generation had more to do wtih everyone leaving BBM than anything else. For as long as all of their friends were on BBM, they all wanted to be on it too.

Network effects are powerful in both directions. They can help you grow exponentially. But when they are going against you, they work just as fast. Myspace's decline was mind blowingly quick. RIM's has been as well. Who is next?

I am not writing this post to pour cold water on the internet sector. There are so many amazing things happenning right now. We are investing actively and agressively and are not the least bit bearish.

But it is important to understand the entire life cycle of what you are investing in. If you are playing a game of musical chairs, you have to know that's what you are playing. Or you will be the one left standing with nowhere to sit. And that sucks.

#VC & Technology#Web/Tech

Comments (Archived):

  1. Mordy Kaplinsky

    This is a very good point I discuss often. Utilities like telco’s, CableCo’s and similar basic infrastructure, while not sexy are basic necessities that have great long-term revenue possibilities although they cost a lot more to create while many cool software companies have little to fall back on when the market changes and stops favoring them.

  2. JimHirshfield

    Could it be most startups lack long term vision – is it a form of ADD? Build em and flip em attitude by the founders and early execs, as well as the userbase always looking for the next shiney thing?Do most startups lack a sustainability gene?

    1. fredwilson

      well facebook’s purchase of instagram shows that they have one

      1. awaldstein

        Not sure.Big Facebook booster for a lot of years. They feel stale now. Their gestures trivialized. And  at a broader market level, many are moving to Twitter.I think Facebook knows clearly what they are about today. Not the same to me as what they will be about tomorrow. 

        1. Matt A. Myers

          Moving to Twitter and G+, and I see the hooks into Twitter and (future hooks) into Google+ will just reenforce all of that.It’s a slower path, though slow doesn’t matter if you still get to the finish line.See The Tortoise and the Hare :Phttp://en.wikipedia.org/wik…

          1. awaldstein

            Don’t know about G+. I’m not a believer, nor user.Re: Facebook, i’ve become a contrarian. And see what it doesn’t do, specifically engagement and transactions, as pushing innovation elsewhere. Check out my post of last night if interested in this.

          2. Matt A. Myers

            G+ isn’t there yet — though there’s still time for something else to disrupt what they’re trying to do..Re: Facebook – I’m definitely a contrarian as well. Some people think I’m crazy even when I tell them Facebook will be “much less relevant in 3-5 years from now”, though it’s not something they study, so I just smile in response – and then come to AVC for thoughtful dialogue! :)Re: Blog post – Beautiful, I will check it out. Thanks for pointing it out. I’ve been on a writing binge the past 2-3 days, so hopefully I have some thoughts to add – though you usually cover all of the bases pretty awesomely, and more elegantly than I wonder if I’ll ever be able to. 😛

        2. JLM

          Your notion of feeling “stale” is a very critical observation. Puts a spotlight on our expectations.We have become addicted to the adrenaline of speed and changing offerings of the Internet. We want that new startup excitement.Twitter is more exciting — like a provocative shout across a crowded bar and the discipline that it must be done in 140 words awakens our sense of creativity.

          1. awaldstein

            True…i like the new.But here, the new is the engagement and the banter. And the platform, Disqus, is the buoyancy that lets us move around and interact without getting lost in the current.On Facebook, it’s just a fast moving stream. Engagement is difficult. Parsing by context near impossible. And more so, the gestures feel like primitive tools, limiting expression, holding us back from connecting.Likes changed the world. For the better. As a platform. They feel like a feature now.

          2. JLM

            There is no question that the one to one banter and engagement is what the world is looking at.I cannot express what a joy it has been to meet so many accomplished and interesting folks on this salon.And, ultimately, in person.During SXSW, I had brunch w/ Grim and Wm M in Austin. It was wonderful to put people to their web presence. Funny thing was I liked them more in person than I might have otherwise on the web.Well, until Grim ate the dinner party at the next table.

          3. awaldstein

            Couldn’t agree more.I’ve made some great friends and clients from hanging around here for what, 3 years now.Nothing is more valuable than friendships. Nothing is more important than the connectors that let those friendships happen.You and I sometimes will I”m connect. When in NYC, ping me.

          4. William Mougayar

            Next time, I want a ride in your other car 🙂

          5. John Revay

            Having brunch w/ Grim – Did he have the head piece on?

          6. Emily Merkle

             Well played. Onward we flow.

          7. William Mougayar

            In the Disqus context, the UpVote is similar to a Like I think, except that they dynamically bubble-up the threads that have the most upvotes.

          8. awaldstein

            The difference is more profound. A like is a solitary act. A thumbs up. An UpVote is a group social act, a dynamic of community.All gestures are not equal. The best will become language in their own right . That is what I was trying to get to in my post this morning @ http://awe.sm/5kwW4

          9. panterosa,

            You know I find it fascination how no one is mentioning online dating and winking and liking int here conversations about engagement and FB and so on.Is AVC such a partnered crowd? or is dating not connected to this engagement activity as a lower form of connection?

          10. awaldstein

            I can only speak for myself on this. Not a segment I know. I’d be interested to read something if there is someone blogging on social design and gestures in dating services.

          11. panterosa,

            Arnold, I don’t know who would write about it. I simply have experience with friends and how they manage their dating. And it is most interesting

          12. panterosa,

            Interesting blend of concepts – stale and speed. I avoided FB until a biz partner needed me to “exist” to post my comments a month ago. As you know I’m happier pseudonymous. But I have almost no interest in having friends in public. For someone who stands out so much in a crowd on the street, I have a very back room mentality. Somehow these two opposing traits haven’t melded into some way to use FB.All FB traffic in my mind was stale – and kind of static – pulses of broadcast info. FB seems to be all people who are happy to live in public, almost need to live in public. Most artists and designers I know post their own work or what they’re looking at – yet most of them say that new work for hire doesn’t come from their postings. Yes they have increased traffic and ‘friends’. So I wonder.Speed and adrenaline is twitter, yet engagement is still in my mind simply a broadcast mechanism – I tweet my convo so you read what I say to Joe. I enjoy to read some people’s ideas, or see their links. But often the engagement – the conversation – is yet another broadcast.I haven’t reconciled myself to the vanity and self-branding of it all. And I am an instigator, a disruptor. I agree Disqus is my favorite place. I also agree I have totally enjoyed meeting people from AVC and Disqus in real life and so agree with you and Arnold and William. Perhaps my frustration is the ever imperfect technological methods we seek to mirror and reflect ourselves. In particular, how many use this as ever increasing way to see themselves rather than connect with the world. 

        3. ShanaC

          They do, but the older I get, the more locked into it I feel.  Twitter still is much harder to use.

          1. awaldstein

            Funny…I just want more. Better means of communicating and connecting my off and online life.There will be many other options coming.What I really want is mobile based communities. That’s the future I want embrace.

          2. ShanaC

            I agree about the mobile. I want better options. But I am less sure that they will exist

          3. awaldstein

            I just expect better options all the time. The social web is at its infancy. Everything will get better if we work to make it so.

      2. Matt A. Myers

        The problem with Facebook’s purchase of Instagram is I think Facebook is a bit delusional, and/or just assume they can maintain a controlled ecosystem to kill off competition (which exists), and hope their massive size will stay there (though their users are mobile) – instead of settling into a place that their current metrics will allow them to sustain and survive long-term. And if they’re basing all of their decisions off of a faulty thesis, then they have potentially to completely die off.I have a blog post draft waiting to be prettified, and one thing I say in it is something along the lines of: It’s the uglier version that dies out. MySpace died out to Facebook, and there will be prettier versions of Facebook – more elegant for usability reasons and other metrics will be allowed that Facebook’s model doesn’t allow for, but that is of benefit to the user.It’s possible they have plans to eventually remove advertising (unlikely), though their setup of newsfeeds / timeline isn’t optimal.

        1. JLM

          I agree w/ much of what you say.I would point out that you use your strengths strategically — you have nuclear weapons when the rest of the world does not, you focus on nuclear weapons.  You prevent shit heads from getting them when you can — Iran.FB is photos, engagement and cash.They moved with deliberate speed to use their cash to protect their photo franchise because I’gram was putting heat in the seats at an alarming rate.That is real strategic thinking and action. Plus $1B is chump change to protect a $100B value.

          1. Matt A. Myers

            It’s all intent-based, where you go / end up. If you have nuclear weapons and want peace, you’ll support peaceful actions. If you’re part of a military industrial complex and intent is on making profits, then you’ll end up with perpetuating war as your end goal / where you end up.I agree fully with you it was a protective / defensive move – I just don’t see Facebook as the end solution, and so anything that they do within their current strategy is futile.

          2. JLM

            The military – industrial complex does not really want war, they want the fear of war and the unknown threat.Today the Air Force is developing a fighter that is 3 generations ahead of the current inventory. This is driven by threats that do not even exist today. That is what the military industrial complex wants.Our current inventory of weapons is superior to all known and unknown competitors and we really do not need a better 8″ howitzer as an example.But the military industrial complex can whip us into developing one w/ counter battery and automatic loading capabilities which are simply not necessary.The military industrial complex does not want to fight a war because it proves up these myths.Look at the game changing implications of drones replacing $200MM aircraft.

          3. Matt A. Myers

            @JLM:disqus  Our biggest potential threat legitimately is an aggressive alien species, though if we want any chance of defending against them we’re better to put money into space exploration and technology..Maybe I watched too much Startrek, Stargate, etc..Fear is the easiest way to get people to do things, to be ‘productive’ – mostly because it’s not using analysis, eg: fear-based decisions, so they’re making impulse decisions.I find it interesting your insight is to say it’s more just perpetuating the status quo of developing technology further for use “in case” – and not wanting public opinion to sway dollars away from supporting further bloodshed. I guess the same can be said relating to education, which I see the resources of needing to shift into more of a synergistic setup with more micro-sized educational pieces, eg: using the physical/research resources universities provide, as well as human.

      3. JLM

        I think the acquisition of Instagram is a merger of photos & photos with the larger imbedded base fearing the growth of the smaller imbedded base.I think they have eliminated their competition with funny money thereby protecting their franchise.This is like bribing the Shiites or the Sunnis in the Al Anbar province.  It is smart but just a bit unconventional.Who cares about spending $1B to protect a $100B value?  1% probably gets lost on the postage meter annually.

        1. Matt A. Myers

          It’s only temporarily protected though, slows down what will be the end result.. There’s no magic technology behind Instagram, though have a good brand, executed well – filled a need. There will be others.So yes, short-term to protect the perception of a $100 billion valuation (based on their userbase size) from looking like it’s easily toppled by an app like Instagram – makes sense … but only if you’re trying to pocket money yourself, and don’t care about the investors – but isn’t that part of fiduciary responsibility too? Or is that responsibility somehow confused and only created after you’ve IPO’d, and therefore doesn’t transfer?

          1. JLM

            I am with you up to your sense of dividing management and shareholders. Management (Z-er) has so much ownership they cannot fail to act in the interests of the shareholders. That is why management owning a big stake is so critical.

          2. Matt A. Myers

            Right, but it’s a catch-22 isn’t it, if the management/pre-IPO shareholders only care about a short-term high gain? Eg: IPO is their pay window, and they don’t have any real vested interest in how the new shareholders do, or rather the risk-reward balance far shoots to them seeking higher reward as they won’t lose their current reward (assuming they cashed out some).

          3. JLM

            @mattamyers:disqus Current management owns too much to cash out through IPO.

          4. Matt A. Myers

            @JLM Right, but what about just cashing out 10% of their ownership, which in some cases is still a lot of money – enough that say qwells the risk of fear enough for them to take bigger risks than they should..

        2. fredwilson

          Insurance/hedge

          1. JLM

            EXACTLY what people with real money do. Well played, in my book.

          2. LE

            Taken as one example makes sense and seems appropriate. But unfortunately there isn’t enough money around to protect against everything that is a potential threat. Add to that the fact that many of the people who might be perceived as a threat won’t act rationally. Instagram started their negotiating position at 2 billion and ended up at 1 billion.  What if they had said 5 billion and Zuckerberg walked? Then the threat actually happened. It would have seemed the right thing to do to pay 5 billion for the company (ah with 11 employees or whatever they had). (On a separate note remember how Msft got rid of Netscape? “We’re going to cut off their air supply”. )

        3. Rowan

          Instagram is a very different network and product to Facebook.Aside from ‘eyeballs on screens’ time it wasn’t ever going to be a direct competitor for Facebook.Its more comparable to Twitter (but visual), and given its mobile focus and location capabilities its also a threat to Foursquare.If Facebook are smart they really will continue to develop Instagram as its own product and wield it against other companies in the space.

          1. William Mougayar

            Agreed. The minute they attempt to “integrate” Instagram inside FB, it becomes the end of it.

          2. Emily Merkle

             I see things a tad differently. Twitter, Facebook, iTunes, Instagram, Foursquare, Google. All scratch distinctly different itches. Twitter: newsfeed (push) / distribution networkFacebook: happy hour meets Christmas morning meets laughter, family, friends, love. No more ads. Zynga has the necessary monetization schedule to keep the trains running on time. iTunes: the soundtrack of life. Foursquare: when you don’t know your avenues ;)All together via API bundles.

        4. LE

          “fearing the growth of the smaller imbedded base”I think it’s much simpler than that. I think it’s the work of a guy (Zuckerberg) who simply decided that something that he has seen happen a few times in his relatively short business career wasn’t going to happen on his watch.  In a sense this is somewhat similar to a doctor ordering up an extra medical test or two for fear of a malpractice lawsuit. He’s like “look what happened to Microsoft, look what happened to Google, that isn’t going to happen to me, I’ve learned from history”. Maybe he doesn’t remember how Yahoo bought broadcast.com and what happened with that purchase.What he doesn’t know is that there are thousands of opportunities that every company passes up on that don’t amount to any threat at all or an opportunity. Besides if users (as Fred has pointed out) are so mercurial (which I fully agree with) what are you going to do buy and block every next thing that comes along that is a threat? Why not put effort into diversifying into something totally new (like the tobacco companies have attempted to do knowing that there is no way to defend the threat of selling cigarettes).  (Make sure to check with Steve Case though first on that strategy.)

          1. JLM

            I would say we are in the same church or synagogue though maybe not the same pew.I agree with all that you say — in particular, the volatility and mercurial nature of the modern platform customer — but in this particular instance, it IS pictures & pictures.I also think that Zuckerberg is plenty smart — whip smart.I think the whole business of AOL, Case and Ted Turner is almost Garden of Eden apocryphal.The young brash guy who promoted the old industry guys with the “new” platform and the old guy who should have followed his own instincts and cashed out.Ted Turner should have gone pay window on them and failed to do it. Big time.Big learning there.

          2. Carl Rahn Griffith

            At least he had Jane Fonda as a distraction 😉

    2. Matt A. Myers

      I’ve always liked Warren Buffett’s question for investment:Would you want to own it 10 years from now? If the answer is yes, then invest, if not, then don’t.It definitely helped guide my thesis evolving, definitely has made things take much longer and made me learn to be even more patient. Thank goodness for yoga..

      1. fredwilson

        but that is hard when you are investing a world that is changing as fast as ours is right now

        1. Matt A. Myers

          Very true, however someone like yourself and many others on this blog, who have been studying the forefront of this changing world will have much better insight into it.

          1. JimHirshfield

            Looking back 10 years, I could never have imagined that in 2012 I’d be over paying for shave tech. ;-)Too hard to see these things when traveling so fast on a track that’s only 17 years long.

          2. Matt A. Myers

            I’ve been web developing in some form since I taught myself to program when I was 11.. I’m turning 29 in 7 days, so, let me pull out my calculator here… 18 years I’ve been involved so far.Maybe I’m a very unique case though it doesn’t feel too hard to see for myself, if you just observe/pay attention to all of the little things, slowly you’ll start to determine what’s important to pay attention to.We’ve had lots of cycles occurring that have shown us large successes and then their collapses, allowing us to contrast what happened, and therefore highlighting the important metrics that lead to success / allow it to be maintained:Usability / Design / Speed- Understandings becoming more evolvedEngagement- Strong metric, if defined properly anywaySocial Aptitude- Helps guide reducing friction; Make things easier / more fluidTrust- Don’t fuck with people; People are fragile and will prefer the friend who’s always been nice to them than the one who’s been abusive: Be a great friend from the start, help people learn – be a teacher.Social Stance / Worldview- Do I want to get rich, or do I want to help the world become richer?- This includes being authentic/genuine; Many people aren’t good at seeing this though, though traction will be started by those who do see the better option, who were seeking for it, and that will help lead things to a tipping point100% Benefit- Build for 100% benefit to the consumer – give the user what they want/need; they are mobile afterall and will move to the better solution- People don’t know what they want or needed until someone shows them the better option; Think MySpace dying quickly when Facebook was shown to people. Similarly, viral is just a signal that a need was missing and (even with ‘small’ viral things, like pieces of information or entertainment, it was something needed by a community, that drew interest)All of these have their nuances, too, of course.

        2. Cam MacRae

          You still invest for 5 or so years though, right?

      2. JLM

        I am both a huge fan and a skeptic of Buffett.His advice has been very good for him but he has not been immune to some horrific judgments witness his involvement in derivatives and his timing and picks on the “recovery”.Having said that, it is my sense that Buffett himself proves the wisdom of Fred’s musings on the life cycle of things.In his dotage, Buffett is being passed by the years and yet he makes some very basic and fundamental bets that only someone commanding his capital is able to do.Trains — OKPlanes — not so goodOn a personal note, he is a very weird cat in many aspects of his life.

      3. DanielHorowitz

        I like Buffett too, but I’m not sure this idea of long term value investing is as applicable today as it was many years ago. Fortune 500 turnover is rapidly accelerating. The time horizon for which we can predict with reasonable confidence is decreasing. How many 2022 Fortune 500 companies don’t even exist today? Probably a lot. We have to be willing to quickly adapt, and I don’t see Buffett doing this. WRT to VC, I think we need to diversify more across time/geography/industry to combat this increasing uncertainty. Rather than 10-12 year fund, we may now be looking at 8-10 year funds and I can see arguments for how many years deploying capital go both ways. 

        1. Timothy Meade

          Maybe 500 is too many companies now, the true conglomerates with true long term value don’t number 500?

          1. DanielHorowitz

            No doubt, but it’s very hard to build long term value. You have to continually adapt and reinvent yourself. I think Google is one of the most important technologies/breakthroughs/companies of my generation. There was before google, and after google. They created a massive amount of value and wealth.But, that adwords cash machine might not be spitting out so much cash in five years, so Google is trying to find other things. Will they be the biggest AI company? Will they be the biggest cable/tv like company? Will they be the biggest VC? I don’t know, but they have vision, and they recognize the need to continually create new value. Google knows that self driving cars/transportation is the future. Most people don’t really see this. Founders Fund knows that nanotech and AI are the future, but this is too far out for most people to see, so they don’t invest or pay attention. I’m rambling, but it seems clear to build truly long term value you need a visionary leader, and even then, it’s a lot of hard work, and some luck never hurt either. 

    3. PeterOrban

      @JimHirshfield: not sure if it is genetic but certainly the market rewards this behavior. As long as the “flip it/buy it” practice is perceived to be less risk than building internally this will be the case. The “flipside”: of course we all know the disastrous “integrations” (MySpace, etc) that refute this point…less news about successful integrations.

      1. JimHirshfield

        Agreed

    4. Emily Merkle

      The whole vision thing….tricky. I think of vision in terms of hope/fire/drive/faith/motivation/possibility.Pivot, a’ 1-2-3-4.

    5. ShanaC

      we’re funding that.  So yes.  It is easier to build a product than a vision.  The question I have is can the process of building a product help build a vision?

  3. Brandon Burns

    If you let the blind lead the blind, everyone’s going to fall in a ditch.Reading a list of quotes from the top rank and file of RIM or New Corp (who was responsible for MySpace’s demise) should come withe a laugh track; just complete idiocy. I wouldn’t point to their products, but to their people.There are many long lasting TV shows, they just have good people behind them. The Simpsons has been on TV for over 20 years. American Idol is still on top, as was American Bandstand for almost 40 years (R.I.P. Dick Clark). Meet the Press, Today, Jeopardy, all the Soaps — they’ve been on TV for 50+ years. The list goes on. Even in TV, smart shows that do their special breed of programming uniquely well, keep up with the times, and have smart people behind them keep on ticking. 

    1. Matt A. Myers

      I’d really like a chance to turn RIM around, it would be a good challenge – but I see it as being doable. It might not be around, or much too late, in 4 years time when my own projects have enough umph behind them to have time to work with RIM.If anyone from RIM is out there, maybe reach out to me. :PYou’re correct though it’s the people involved, not the company you see. It will be interesting to see what happens to Apple’s releases in 2-3 year times, when decisions and future planning no longer has the filter of Steve Jobs overseeing things. I’m hoping he planned out a 10-20 year path for them – just because I’m curious. I know where I’d take Apple next.

      1. Brandon Burns

        where would you take apple?

        1. Matt A. Myers

          Oh, that’s my secret! :)If they don’t go there or someone else doesn’t, I will in 10 years – when hopefully I’ll have the resources and experience to put behind it. :)No real rush with most technology of where things will be.

          1. Brandon Burns

            hmpf. inquiring minds want to know!

          2. Emily Merkle

            Do tell.

      2. Timothy Meade

        7 inch tablets are huge Apple should get into that. Hell they should just make the iPad 7 inches, since 7 inches is such a big category in tablets, I mean look at the Kindle Fire, it’s such a success (if we ignore margins). I mean we’re Apple we can’t possibly fail as long as we do everything the market wants us to do, and listen to every focus group, and every consumer group, and every business…That’s how we got here right?(I know Tim Cook is smarter than this, I just hope the rest of AAPL is)

      3. JamesHRH

        Sorry to be such a grouch Matt, but you don’t want this challenge.Unsalvageable.

        1. Matt A. Myers

          I think it’s an unknown to me if it’s unsalvageable when I don’t know what they’ve been hoping to release, what their longer term plans are – what’s in the pipeline..I agree it is an undesirable challenge, and the likelihood the opportunity would arise is slim – especially because the circumstances would have to be very correct, and focusing on my own efforts can potentially yield better results anyway. 

    2. fredwilson

      Great points

    3. ShanaC

      Most of the Soaps have been canceled because of Feminism and technological changes involving time.

    4. JamesHRH

      Brandon – an interesting thread to your post…every TV show you listed had no specific generational appeal. They are timeless, generic vessels – they are the utility part of TV.Miami Vice & 90210 were huge – they defined the 80’s. Three’s Company & The Rockford Files & Carol Burnett in the ’70’s etc.. Seinfeld is the ’90’s. They are like flowers or flies – they have a season where they bloom & then they are done.That’s not FB or Twitter. It might be Pinterest.

  4. Cam MacRae

    Taking the excellent xkcd google trends comic from the other day and changing ‘blog’ to ‘blogspot’ you get this:http://www.google.com/trend…which I think shows exactly this process.

  5. Renato

    My guess is that next one without a chair will be Nokia phone unit.

  6. Jan Schultink

    So a tech company should see the end of its life cycle and divest/acquire to restructure itself to justify its valuation beyond a finite number of cash flow years. IBM did that.

  7. Tom Labus

    In Inside Apple, Lashinsky talks about how cities never (or rarely die) but companies do all the time. Fast growth, peak, bureaucracy and death.But it’s so interesting to see some companies avoid that trap.MSFT may be one of them.At Apple who will take the next leap of faith to a new product(s) when they hit a crisis and will they believed internally.

  8. JimHirshfield

    Sustainability by acquisition is like healthy living with lipitor and stints. You’ll live, but quality of life not the same. Better to take care of yourself (innovate) than have side effects (C-level palpitations, staff leakage).

    1. awaldstein

      Yes and no.Sure…integration is a bear. Culture non withstanding but certainly the only solution is not to built it yourself.On the flip side, you can buy product and people but can you buy vision?

  9. jason wright

    If you want to win at musical chairs be in control of the music.That’s how governments and vested interests think about the interweb.

    1. Matt A. Myers

      I dance to the beat of the universe. I just want to manage as best I can which seat I am occupying; Some seats have a better view than others. There is enough seating for everyone.

      1. Emily Merkle

         Come on in – the water’s fine.

        1. Matt A. Myers

          I like to swim nude – is that okay?

          1. ShanaC

            O_o

          2. Matt A. Myers

            I think you got something in your eye – it looks a bit swollen..

          3. ShanaC

            You don’t watch enough anime. More like slight shock.

          4. Mark Essel

            There’s not time for even the naked truth, this is all so brief.But that respone got a cackle, and Shana a huge eye :).Fred’s comments on apps as fads is spot on, but the same holds for businesses – with a few exceptions.Buyers aren’t buying an eternity of cash flow, they’re buying an existing revenue stream conditioned on the market as it stands. The state of the market is continually drifting and evolving. Some collapse entirely while others bubble up from emptiness, apparent or otherwise.

          5. Emily Merkle

            sure 😉

      2. William Mougayar

        Good answer Matt. But the best view doesn’t always give you the best returns. If you’re retired, get the best view. If you’re still working, get the seat with the most food.

        1. Matt A. Myers

          Retirement just means being able to do what you want, which is where I want to position myself as best I can – to enjoy life with what I am doing.A seat with the most food might mean you’ll get fat too, and more doesn’t mean more quality. And there are many different tables of food that can be accessed from any given seat, and some with lots of food. :)I don’t need a feast for every meal though. I’d rather make $100 million a year, forever – than $1 billion in a short period. Low risk vs. high risk.I do understand what you’re saying though, and I plan to be full. 🙂

          1. William Mougayar

            I know you did. Get full in $$$ not in the stomach. I should have said “best” food 🙂

      3. jason wright

        There are enough chairs for everyone, but some chairs are called thrones. Not everyone gets to sit of one of those. That’s the present order of things.

    2. William Mougayar

      True. Few get to control the music of life and business …except on Turntable.fm

      1. Cam MacRae

        Can you get on Turntable yet? We southern folk will be last in the world, as usual :/

        1. William Mougayar

          Not yet. Still waiting.

    3. ShanaC

      I dance to the sound of silence….

  10. matthewdbenson

    Fred,The audio version from @VoiceBunny says that “I’ve watched my kids go from myspace to facebook to instagram over the past five years” whereas the text version above says exactly the same, but over “the past seven years” …Does @VoiceBunny ‘disagree’ with you (!), or did you subsequently edit your post…?Matt

    1. fredwilson

      i edited it

      1. JLM

        Wow, these guys are really paying attention.  Comparing the audio to the written word.  

        1. fredwilson

          Keeping me honest

          1. Dale Allyn

            Everything you say, can and will be used against you… 😉

  11. Bill Clark

    The obvious one is Groupon. It has gone from a service with a few good deals a month to discounts on massages from places no one wants to go to. I can’t remember when the last time I purchased a groupon but I still will look at the headlines on my email. One of these days I will unsubcribe and so will the rest of their user base.

  12. Dave W Baldwin

    Great post!  As you know, Fred, the key is to think of all things forward 5 years.  Don’t get too wrapped up in what is today, like Groupon, Google+ and so forth.Using Television Shows as analogy is right on.

    1. LE

      “Using Television Shows as analogy is right on.”Television shows that are successful many times are just a re-jig of a previous concept that has worked before with a new skin or a minor change. Because people and their likes and psychology  stay the same. For example  “What would you do” on ABC is just a variation of “Candid Camera” which had considerable success.  (Probably the one of the most ubiquitous  concept of all in entertainment is “fish out of water”.  Examples are (literally) Splash, Crocodile Dundee, Legally blond,  Private Benjamin etc.)So there is no need to beat your brains out trying to be original! Just go back in history and find something that worked at a previous point in time business wise and do a variation of it. (Remember the Pet Rock and what has followed that?) Dutch tulip craze? 

      1. Dave W Baldwin

        Agree with criticism re originality… The television analogy points more to hours usage and who is using it, not placing TV on the phone. When you move out 5-10 yrs, looking to rehash old television shows doesn’t cut it.

      2. Carl Rahn Griffith

        Our industry is going through a ‘jump the shark’ period…

  13. Dan Lewis

    This isn’t new. It’s a minor change, in fact.Video games — platform ones, I mean — are very much analogous to movies: big production and marketing budgets, relatively long production cycles, and large price tags for the consumer. We’re seeing some disruption there from things like Steam and, of course, iOS/Android apps (and to some degree, Facebook games). And the model there is akin to a TV model: lots of studios/networks, lots of games/shows, few winners, even fewer long-term icons. I think Zynga and Rovio get this. I think OMGPOP did too, which is why they felt comfortable selling after one hit.  So yeah, I think in that regard, you’re right.RIM, I disagree. RIM failed fast because their products got worse while competitor products got better. When the iPhone came out, RIM ran into the touch screen universe… and did a bad job of it. They went into the app universe… and did an even worse job of it. Meanwhile, they still haven’t fixed their browser. If you want to blame their failure on an environment where iteration comes rapidly, that’s fine, but it’s not really analogous to why TV shows go from super-popular to off-the-air.Myspace I’m on the fence about. While it was very popular, it wasn’t anywhere near as permeating as Facebook. Ask any 30-something how many of his or her friends were on it and you’ll get a very low number; ask the same question about Facebook and it’s almost everyone. If you need a TV analogy, I don’t see MySpace as the Internet version of ER; I see it as the Internet version of a 1980s show on USA Networks which lot of people who watched, but a lot more people looked at, said “this is creepy/weird/not for me” and skipped it. That said, the reason MySpace came down so quickly is exactly because of the network effects you’re talking about.Facebook/Instagram, I think the jury’s really still out on. I also don’t know how great your children are for measuring Facebook’s impact. They’re still in school, right? Not yet in college, I mean? That means that they see most of their friends on a day-to-day basis; they have really one or two social circles (camp friends perhaps as a possible second one); and they’re in the same station in life as all of their friends. Compare that to 30-something me. I have friends from high school, college, law school, and three different work places. Some of them are having kids, getting married, moving, changing jobs, etc., all at different points in time. Staying connected with them is virtually impossible without Facebook, and while I agree the network effects could make its collapse happen seemingly overnight, it is MUCH more likely that network effects will *prevent* that outcome. TV shows, not so much.

    1. JLM

      Well reasoned and convincing to me.  Thanks.Well played.

      1. Dan Lewis

        Thanks!

        1. Timothy Meade

          I do not see how your last statement, that Facebook will fail because of it’s closed infrastructure, follows in any way from the entire argument you seemed to be presenting.It’s the exact opposite, the closed nature and the costs to acquire that massive network of people who have invested in Facebook as a platform for organizing their life that will keep Facebook relevant for years, perhaps decades.It is true that the closed nature is a challenge to those who want to use it as a gaming platform or as a ‘social media’ platform to enrich their own business, where Facebook’s revenue is dependent on that usage, but relevance is dependent on the opposite drive to remain more signal than noise. But for you, and me, who use Facebook as the “universal social utility” it’s not really relevant, as long as the friends we’ve added since college, including everybody we know from high school, college, etc. remain part of that network updating us with their life changes, it will remain dominant.

          1. Dan Lewis

            I didn’t say any of that. I agree with you, entirely.

    2. fredwilson

      Two of my kids are in college. One is in high school

      1. Dan Lewis

        Well, then, I guess I screwed that one up 🙂

        1. fredwilson

          Not at all. Great comment. Just wanted to answer your question

    3. Matt A. Myers

      I don’t think Fred was questioning the value that exists with a structure like Facebook.  The need is there for such a connection point exists, using your example of a 30-something year old as social proof.However that same structure can exist elsewhere, without Facebook, and it can have an even better structure.Edit: To further this…People can change a channel to watch a different TV show. People can change what application they are using by downloading it / visiting a different site.TV shows can be created relatively easy, the infrastructure, processes and human resources and other have been evolving for many decades longer than online tech. However applications are relatively cheap to create now.The hard part is discovering what will work. Add a laugh-track to a show? More people will laugh while watching it. Have an easy-to-use/faster website? More people will use it.A movie like Avatar takes much more planning than a TV show (or perhaps a better analogy is a one-off documentary, or perhaps news excerpt). This planning time likely can fairly include the lifetime of James Cameron influencing every decision he made for the film. It’s a big ecosystem vs. a smaller one. The smaller it is, the less variables you have to account for – the less potential competitive factors to manage.Similarly, a large ecosystem, a large network of engaged users, will take longer to plan – the details and it’s the nuances to figure out, nuances being the smallest/most subtle of decisions that have the largest impact on overall success (with everything else as a strong foundation).I guess all I am saying is all it takes is time for the better new to come into play, the old less better can quickly fold – like MySpace happened with Facebook.Why Facebook will fail is the controlled ecosystem.You can’t know what will work because of competitive factors, so the next best thing you can do is create an ecosystem/environment that allows for things to naturally unfold – create an environment that facilitates this.The thing you need to eliminate is control.

    4. Dale Allyn

      Dan, I agree with your comments. I’ll add that to me the MySpace failure is largely a product of signal vs. noise. Of course there’s a list of contributing factors, but there’s a fundamental issue of noise there and a complete lack of continuity. There’s no cohesive UX aesthetic or framework, so the content-noise is severely amplified. The disjointed mess offers individualism at first, and then dies of loneliness in the crowd. 😉

    5. JamesHRH

      Myspace is a non-standard experience. You cannot scale a nonstandard consumer experience into a mainstream brand.People need the comfort of consistency to invest in the network. MySpace is still huge – but it is not mainstream because it is not a uniform experience for all members.

  14. kidmercury

    i think cities are a better comparison than TV shows, although as brandon noted elsewhere in this thread, there are long-lasting TV shows. a big part of the problem is the focus on apps; companies are building themselves to design apps. in reality, they should be thinking more about building social graphs — building communities. related to this is the extreme focus on software engineering and user count as opposed to the myriad of other skills needed to build communities. basically, social networking companies are trying to use the same inputs a company like google does. this will not work and will result in parabolic rises and declines. the other issue is that many entirely digital companies won’t have the business model capabilities that we associate with traditional offline businesses. pay what you want models i think will be increasingly common, as will social networks built around individuals. avc.com, for instance, will last as long as fred wants it to and faces no real threat of going out of style like the big social networks do. avc has no resale value, though; take away fred and the whole thing falls apart. we have yet to see a social networking company do meaningful things outside of software. they don’t build computer hardware like apple, warehousing centers like amazon, or self-driving vehicles like google. the real purpose of social networking is to re-wire the world and this does mean finding your way into real matter. failure to innovate along the right trajectory can be fatal to specific companies, although it does not condemn the entire industry. 

    1. awaldstein

      “a big part of the problem is the focus on apps; companies are building themselves to design apps. in reality, they should be thinking more about building social graphs”Right on Kid!Apps solve problems ostensibly. Train schedules. Where am I?Big believer that the web is about me, wherever I am. And for most instances, on mobile, there is no community that really connects us.Verticalization to create a more contextual sense of intimacy and gestures as language to overcome the input restriction of mobile are key here. You might like my last post, a meandering rant about gestures and language @ http://awe.sm/5kwW4

    2. Brandon Burns

      community point = spot on. from the simpsons to american idol to all the soap operas… the most successful “shows” are more than that. they’re brands that have woven themselves into the fabric of people’s lives. they have communities of loyal followers who are obsessed with their every move in between episodes and seasons. reminds me of my advertising roots. i always told clients: sell the brand, not the product. sell the emotional connection, not the function. that’s how you’ll grow your community. i think i have some copywriting changes to make on my own product now… :o)

      1. ShanaC

        It is problematic to build communities on the web – remember DIGG – if you move ever so slowly in the wrong direction, you get that nightmare.If we continue with the City metaphor – there are some cities that are classic, and some that are not.  No one you know has deep plans to go live in Timbuktu anymore, but probably most of your friends if they could would be happy in Paris.

  15. Sanford Rich

    Great topic. I call this my “Unstable Power Order Rule” of investing in technology. The winners and losers in an App or Web site eyes attraction contest suffer from an organic Power Order function where the successful get the vast majority of attention and others get the scraps and die. The Unstable part is the tendency for eyes to be fickle and flow to the next attraction in near lock step, another source of death. These tendencies make finding an attractive investment in this space more a function of luck, in my opinion. Again, great topic.

    1. Emily Merkle

      And luck is a lady, Thank God.

  16. Steve Pell

    Fred, such a great post that needed to be written on the reverse networkeffect. Thepoint on social that you note with your kids is so valid. Establishing a profileand network was harder on Myspace than Facebook than Twitter than Linkedin than Instagram… Each time it’s been a substantially easier process to get to thepoint of active engagement and benefit from the network (and there’s no waythis has been a linear progression!). In addition to improving techfacilitation, I feel like we’re each getting better at building and replicatingonline social networks (it’s almost core skill set for this generation). 

  17. David Miller

    It’s easy to confuse rapid customer growth with a durable, long-term business, especially as we’ve come to believe that networks of users lead to higher degrees of customer “lock-in”.The real way – the only way – to maintain a business over time is to continue to add value for customers, and that changes over time.  Maybe what you’re saying is that social networks may have a harder time adapting over time.  Maybe their customer lock-in is more perceived than real. Is Zynga really going to generate more than $180 million in profits from Draw Something?  I’m not so sure, especially if the half-life of a social game turns out to be six weeks.

    1. Luke Chamberlin

      In addition to Draw Something (the game) Zynga bought the right to market new games directly its 50 million users. A certain percentage will play Zynga’s next game when they get bored, and this process continues. That’s a big part of the valuation.

      1. Emily Merkle

         I see this as a pivot.

        1. Luke Chamberlin

          If you mean a pivot toward games that are actually creative and fun I agree with you.

          1. Emily Merkle

            I’m saying let’s go yard.

      2. fredwilson

        Yup

      3. David Miller

        I get that … but how many of these users were new users to Zynga?  Not nearly all of them.  Was some part of it defensive?  That is, to prevent another company from becoming more dominant in social gaming?I’m not sure I understand the model well enough.  It just seems that a lot of things have to go right for the financials to follow the users.

  18. David Repas

    I actually think that you can see the destruction of life cycles in many other areas of society as well. Social movements more and more often seem to have exponential growth, but also have much less staying power than in the past. It’s become so easy (maybe too easy) to become part of something and then just move on to the latest greatest whatever – no investment is required.

  19. sigmaalgebra

    Fred, you are trying to make sense out of some complicated reality, and that can be tough!Your “large network of engaged users” has some strong points, e.g., network effects and switching costs. These strong points can provide a barrier to entry, a longer company expected lifetime, and a higher net present value.High switching costs can be cute: If some software is more difficult to use, maybe with more functionality, then people who have made the effort to get skills with the software can be reluctant to switch! E.g., I rarely use Microsoft Word so that each time I return to it takes me some minutes until I stop swearing at it, but people who use it a lot and are good at it apparently do NOT want to change! If want to make software more difficult to use, then maybe have the software easy to use for first time users and have the difficulty optional only for experienced users! But, surprisingly, apparently some such difficulty, even without extra functionality, can be an advantage!For NPV based on discounted future earnings, maybe many investors in liquid stocks don’t worry about that and, instead, just pay much more, get happy if earnings and stock prices rise, but sell within less than a second at the first news earnings are starting to fall! Hmm …: Driven just by Brownian motion, does such behavior by many investors yield some predictable price changes?For DEC, they had a lot going for them. At one time DEC was getting more revenue from DuPont than IBM! Why? Because the DEC VAX was big in the DuPont labs, and when a new lab project went into production the VAX software the lab had written went into production too. Then DEC VAX found that they also needed business middle-ware, e.g., database, but were doing that. So, DEC VAX was moving into more traditional IBM territory!As microprocessors started to dominate, DEC failed to ‘get it’ effectively and, then, slowly died. Sad: There was no good technical reason at all, none, for DEC not effectively moving their VAX business onto microprocessors (apparently the DEC Alpha chip was not part of an effective move) where they, then, would have retained their advantages of polished VAX software, expert VAX programmers at DEC and customers, existing customer VAX software (lock-in, high switching costs), a golden reputation for quality, an excellent marketing force, a terrific brand name, good R&D and manufacturing, etc. Instead they blew it.Similarly for Prime and Data General.Similarly for Honeywell and Multics.Actually, the Intel 286 chip was a close cousin to the Prime hardware which borrowed a lot from Multics. Prime had some very nice software that would right away have been the best in the world on the 286 chip. But Prime blew it.For Sun, they were too slow to move to Intel hardware and to lower their prices.Apparently what happened to each of DEC, IBM, Prime, Data General, Sun, etc., is that at one time they had a successful ‘business model’, built a management tree to ‘manage’ that model, and expected that situation just to continue. But that management tree was not hired, expected, or rewarded to move in new directions and, instead, fought new directions internally. So, as technology and the market changed, the company went out of business. On the way down, the more powerful people said to the others, “Ha, ha, yes, the boat is sinking but as your end sinks, my end rises!”. Too soon both ends were under the waves heading for the bottom.A smart, powerful CEO can keep up with change, e.g., the Gates memo on the importance of the Internet. Apparently Microsoft has done a lot to win in server farms and now might be on the way to winning in operating system software for mobile devices. Maybe.One way around such problems that might work is:(1) Problem. Pick an important problem, one that likely won’t go away. By “important” we mean in part that a solution that is by far the best will be quite valuable.(2) Solution. Find a solution that is by far the best. Here one approach is to do some original research that can result in powerful, proprietary ‘secret sauce’ that provides crucial help in making the solution by far the best.With a Web app, that secret sauce can be in software secure inside the server farm and not at all visible to the users: No one outside the company or even outside the CEO’s office need have even a weak little hollow hint of a tiny clue what is in the secret sauce while all the users can easily see that the solution is much better.Having the better solution secure inside the server farm can be a ‘barrier to entry’.There can be other advantages: A broad area of advantage is data. Some cases of data might be regarded as a valuable company ‘asset’. Maybe all the stock market data on all the stocks on all the exchanges, tick by tick, going way back would be an example of valuable data.For a Web app, another possible example of valuable data might be the history of usage by registered users. Such data might be used for ad targeting and ‘personalization’ more generally. As privacy concerns throttle centralized collections of data on individuals, a company that does its own ad targeting on its own users using the data only it has on those users can have an advantage.Of course, in part here we are assuming that the current situation in on-line ads will remain! It might not! If the nature of such ads changes, then the revenue model might have to change.Even if are successful in picking a problem that doesn’t go away, still will likely need to keep up with some cases of change. Here at least the CEO has to be effectively involved.

    1. ShanaC

      how long will it take to move beyond data as a competitive advantage – data is bought and sold these days.  New layers of targeting are put over it in terms of ads.*sigh*

      1. testtest

        difficult to move past ‘data’ when it’s an integral part of information systems. can think of it as:              data                  |input –> program –> output(see if this looks ok after i post it)with data being the most scalable aspect: you can double the data without doubling the lines of code. 

        1. Pete Griffiths

           Peter Norvig (Director of Research at Google) discusses exactly this point brilliantly in his talk “The Unreasonable Effectiveness of Data”http://www.youtube.com/watc…essential viewing!

          1. testtest

            yep, amazing. that’s where i got it from pete

          2. Pete Griffiths

            :)))

          3. sigmaalgebra

            How modest of Peter!As at     http://www.sciencedaily.com…is> ScienceDaily (July 25, 2003) — Santa Barbara, Calif. — Three years before he received the Nobel Prize in Physics, Eugene Wigner published an article entitled “The Unreasonable Effectiveness of Mathematics in the Natural Sciences” (1960).  He marveled at how often physicists develop concepts to describe the “real” world only to discover that mathematicians–heedless of that real world–have already thought up and explored the concepts.  His own experience of the uncanny applicability of mathematical insights to the physical reality of quantum mechanics led Wigner to observe “that the enormous usefulness of mathematics in the natural sciences is something bordering on the mysterious and that there is no rational explanation for it.” 

          4. Pete Griffiths

            Exactly – that is precisely the paper that Norvig referred to and that his talk was an homage to.

        2. ShanaC

          True – but if your data set is easily repeatable (ala advertising) then it is your program

      2. sigmaalgebra

        In principle and not always but commonly and pervasively in practice, data is and can long remain one heck of a valuable business asset and competitive advantage. E.g., for real valued random variables X and Y (in practice X and Y can be observations of anything at all where there is no intuitive role for ‘random’) f(X) = E[Y|X] is the best non-linear mean square error approximation of Y given X, and the discrete approximation here is just simple cross tabulation. With several variables, cross tabulation is a prodigious user of data, but we were talking lots of data, right?So, want Y because with it get a good trip to the pay window. Have X. Then get a lot of data and cross tabulate. Yes, that replaces linear regression and much more. Did I mention that we can need a lot of data? X can also be a vector — now can use MUCH more data!The senses in which data is powerful in principle are astounding. E.g., can argue that in principle with enough data and enough computing, achieving the ultimate possible in principle ‘singularity’ of artificial intelligence would be routine. The software is shockingly simple, just stochastic optimal control. While I did my dissertation in stochastic optimal control, I didn’t dream up this ‘singularity’ observation but had to agree with it roughly when I read it on one of the far out blogs.I say only “roughly”: The data volume might be so large it would fill the universe at any conceivable data density!Here’s stochastic optimal control in a nutshell: Consider monthly business financial planning for five years. Develop a spreadsheet for this with one column for each month and one row for each place to make or spend money. Fill in the spreadsheet in the usual way with random numbers for some cells and blanks for cells where have freedom to make decisions. Then stochastic optimal control is the best way to make the decisions to maximize the net expected value of the company at the end of the five years. Alas, the decisions are of a special kind: There are some huge tables constructed on the side that, at each month, in terms of all possible ‘states’ of the business, say what decisions to make that month. With enough data, get to say what the distributions of the random variables would be and also the ‘response’ (‘plant dynamics’) connecting the months (the algebraic expressions typed into the cells) would be. This is the discrete time version of stochastic optimal control. For most significant real problems, we’re talking ‘big data’ and more than current super computing processing. Still, occasionally some practical problems are quite doable now; nice when that happens!In practice, yes, data gets bought and sold, but at a price, and the price doesn’t have to be low!I mentioned Internet privacy: Off and on Congress gets concerned. Eventually the idea of having ad networks track users via third party cookies and even their IP addresses and, then, use/sell the data may get outlawed. But then just within one Web site there may be freedom to continue to collect data on that site’s users for, say, that site’s ad targeting.If in principle with enough data can have enough artificial intelligence to play God with the universe, then a lot more in ad targeting and much more in business should be reasonable to do and valuable!In a sense with enough data can knock off problems by brute force with simple methods. More complicated methods can be terrific with more reasonable volumes of data.In practice, to see if there is some value in data, usually have to evaluate carefully one box of data at a time!For solo violin with orchestra, Max Bruch wrote the ‘Scottish Fantasy’ based on traditional Scottish folk song themes, and the Heifetz CD performance is one of the best examples of his magic. With his playing, the ‘Andante sostenuto’ section is one of the best artistic representations of the soaring, indomitable human spirit I know. I recommend it! So, I copied my CD of the music to my computer and have my computer playing it as I type. It’s magic! Beyond belief!

        1. ShanaC

          Agreed that lots of data can be a potent thing – but at some point repeatability of your dataset (as massive as it is) decreases its usefulness.And I don’t think advertising data sets are going to be regulated away….

    2. Pete Griffiths

       The thing that killed DEC, Data General, Prime etc was UNIX.  The high end (mainframes) was dominated by IBM plug compatibles like Amdahl.  The bottom end was Wintel.  But the mid-range was an alphabet soup.  The problem for all the second tier mid range companies (not IBM or DEC) was that they couldn’t attract software developers because their installed base wasn’t large enough.  That’s why the disruption of the mid range started with 2nd tier companies organizing around the UNIX platform.  IBM and DEC had no incentive to support such a move – far from it.  IBM had places to hide but DEC, Data General etc didn’t, so they fell prey to the innovator’s dilemma.  They were slow to react and got their lunch eaten as the UNIX market share grew and grew and their customers saw their reluctance to commit to UNIX and their positioning suffered accordingly.  It was an interesting period with important lessons ot teach us about what is likely to happen in mobile.

      1. sigmaalgebra

        Good analysis.But I believe that DEC and Prime had a chance to keep UNIX at bay, keep their customers, grow, and take what is now Microsoft’s position and more:Around 1980 I served on one ‘midrange’ selection committee, soon was the technical lead on a second committee, and soon was the lead on a third one. The first two acquisitions were for military systems analysis, and the last was for graduate academics.For the first committee, our IBM TSO time sharing bill had reached $80 K a year for just two programmers, so we got a Prime for $120 K and cut the wire to TSO. Our computer usage exploded to 25 users, then 45, then much of the company of 600. Then I was the technical lead on the committee to select computing for the company.I ended up managing, at one level or another, the sites from all three committees, the third one as a chair of a faculty committee.Besides technical calculations, we were keeping printers (daisy wheel and then laser) going for word whacking. The last site became a world distribution site for Knuth’s TeX (due to J. Crawford) on Prime.Then near 1980, DEC and Prime had no shortage of users, sites, and software developers and were paving Route 128 around Boston with gold. In 1980, Prime gave the largest return of any stock on the NYSE.Generally Prime had the lowest prices and best ‘ease of use’ and system administration. DEC had the best collections of users and applications software. Soon Data General came along with the box in Kidder’s ‘Soul of a New Machine’. There were a few other players that really didn’t catch on. Then an IBM box with comparable power had a console that cost $250 K.Then UNIX, if only due to licensing considerations, was significant essentially only at AT&T and in some university settings, usually or always on DEC hardware. Later that university position was very influential for UNIX, Linux, C, C++, BSD, TCP/IP, etc. to the present with Android, etc.In the early 1980s, as UNIX was growing in popularity, Prime considered a simple option to let a Prime login look like UNIX. That option would have been easy for the Prime operating system (and likely also DEC’s), but I don’t recall what became of it.As athttp://en.wikipedia.org/wik…the Intel 286 was introduced on February 1, 1982.At that moment, Prime had an opportunity: Since the 286 was a close cousin of Prime’s hardware, Prime could have relatively easily ported their operating system software and users to a 286 box. They would have had for years the midrange box in computing that was the most powerful, cheapest, and easiest to use, with the best collection of users and software, and could have covered the worlds of technical and academic computing and smaller business offices with it.Why port to the 286? As soon as the 286 was introduced, it didn’t take Gorden Moore or Robert Dennard to see that microprocessors were the future of processors. So, Prime could port or die. They chose to die. By porting, starting in 1983, they would have had by far the best software (basically the Multics operating system that in important ways Microsoft is still straining to equal) for 10 or more years and maybe kept UNIX, etc. at bay. UNIX was built at Bell Labs to run on an 8 KB DEC and was a toy compared with Multics.Prime was started by a team who had worked on Multics at Honeywell. The war story went: “We believe that we can bring up Multics on a bit sliced super-mini computer, sell it, and make money.”, and the Honeywell response: “We don’t believe you can bring up Multics on a super-mini computer; if you did, it wouldn’t sell; if it did, it wouldn’t make money; even if it did, we wouldn’t be interested.”.There was no reason a Prime with a 286 could not have been a desktop system in 1983 or so and used for computer aided design (where Prime had a good market position), technical computing, academic computing, small office computing, driving daisy wheel, dot matrix, and laser printers, on and on, for all the work that made the successes of Sun and Microsoft.I mean, why the heck would anyone want anything from Microsoft before Windows XP or Windows Server 2003 when they could get a Multics on the same hardware?Computing has been ontogeny recapitulates phylogeny, and Multics remains one of the best of the phylogeny the Microsoft ontogeny is still trying to equal in all respects. The potential? “Look, Ma, lots of users writing and running any software they want with no security problems!”. Multics was designed for secure, multiuser computing.On all the hardware Microsoft’s software runs, until fairly recently a Multics, yes, kept up to date with a windowing system, TCP/IP, etc., would have been much better. All the growing pains Microsoft has put itself and the world through relearning the lessons of Multics is astounding and outrageous. Prime could have had ALL of the Microsoft success along with much of that of Sun, Dell, etc. It was sitting right there with the 286 in 1982.For a graphical user interface, Prime could have proceeded much as for X-Windows.Gates saw the future of a microprocessor on each desk, and Prime didn’t.Later with Sun, Linux, and IBM’s AIX, the UNIX world exploded in popularity and now is the main challenge to Microsoft.The Linux world has a double edged sword: Yes, it’s cheap and open and is awash in users and developers. But then it has no good organization or big bucks behind more development as must be needed in coming years. So, Microsoft has a chance: Have better software.On mobile? Apple’s strategy has always been Jobs’s — A walled garden with its pros and cons. The Linux world has been wide open. The Microsoft strategy has been to do the operating system, middle-ware, and some applications but leave the rest to developers.At present I see as Microsoft’s big, HUGE problem: A Windows user needs to be able to download and run literally any software at all with absolutely, positively no security threats at all, and Microsoft is not nearly there yet. Did I mention Multics?I can’t believe that the mobile world won’t encounter some really severe software and other computer security problems and need a solid solution. Mobile, small, cheap, whatever, it’s still a computer with essentially all the old challenges along with some new ones.

        1. Pete Griffiths

          I was intimately involved in the seismic change that UNIX visited on the midrange market. We did a huge amount of consulting in this area to most computer vendors in Europe and the US and saw at first hand people struggle more or less successfully to absorb the implications of the move to open systems. The X/Open group was initially just European companies who banded together to create a common platform to attract software developers so that they could have a hope of not being crushed by IBM and DEC. The other mid range vendors were not really a strategic threat – they all had the same problem – the big two. The key change was not the hardware – the point was that UNIX was portable to pretty much any hardware and hence could (a) become a common platform, and (b) take advantage of hardware changes much more rapidly than any os that was tightly coupled to proprietary hardware (written in assembler).I don’t think Prime had a hope. Even if they had run something like multics on intel – so what? The whole point is addressable market size. A better Prime OS wouldn’t solve that problem it would just be another (good) proprietary OS. VMS was a good such OS and it went down the toilet too.(btw There is no way you could run real Multics on 286, 386… Multics was a very demanding OS. UNIX and C were heavily influenced by Multics but were cut down and were implementable on mid range hardware.)The irony of what happened is that the companies who let the genie out of the bottle by supporting UNIX had no choice but to do it and yet for many of them it just meant another year or two of life because if the underlying platform is standardized how do you differentiate yourself? DG tried a faster box – didn’t work. Others added proprietary software to try to lock users in whilst preaching open systems but users got increasingly sophisticated at recognizing such strategies.The lessons of that transition are repeating themselves now. Don’t get caught between the market leading proprietary product and its open competitor. RIM got caught between Apple and Android just as so many got caught in a proprietary hell hole between Apple and Wintel or DEC and UNIX. It’s a terrible strategic place to end up.Lots of lessons from the past. 🙂

          1. sigmaalgebra

            It appears we are not talking about the same time intervals.I’m concentrating on year 1982 when DEC and Prime were very successful and the 286 just came out and UNIX didn’t yet amount to much.For “assembler”, Prime’s operating system was first written in a slightly tweaked Fortran.  Later they often used a simplified PL/I compatible at the object module level.  How they did it was fairly simple:  They had a fixed, maximum number of users, say, 32, and, then, a Fortran COMMON block with space for the ‘state’ (registers, etc.) for each user.  Then they were running nearly just a single threaded system!  It is true that the Prime hardware had more than one set of registers and did the ‘task switching’ in firmware and, thus, had very fast task switching.The Prime 400 of about 1978 ran in 40 KB of virtual memory address space in a machine of, as I recall, 128 KB.  It was maybe a 0.1 MIPS machine.On Prime 550, much the same as the 400, we were supporting 15 users at a time, 20 in a pinch.  A 286 box could have been cheap enough to be dedicated to one or just a few users.  For a file system shared among several computers, Prime had that in good shape in 1981.So, porting Prime’s operating system to new hardware, the 286, with very similar architecture (pages, segments, addressing, virtual memory) should have been relatively fast and easy.The 286 was likely faster than several versions of the Prime single board, bit-sliced hardware back to 1977 that did run Prime’s OS, that WAS basically Multics, just fine.  During the 1980s, soon the Intel microprocessors were faster than the DEC and Prime computers of 1980 or so.  E.g., the DEC VAX PDP 11/780 was supposed to be a 1 MIPS computer, and the 286 chip eventually had a clock speed of 25 MHz.  I ran OS/2 on a 16 MHz 286 in about 1987.  Soon the Intel 368 was at 50 MHz.I don’t know UNIX technical history well:  The original AT&T UNIX was a toy for an 8 KB machine with no virtual memory, etc.  By the Berkeley Software Distribution (BSD), UNIX was getting serious, but I don’t know the date of the first, good BSD version.  My guess is that the first good Sun product, on a Motorola 68000, was via BSD, but my guess is that that was about 1990?You are losing me on just what hardware and when UNIX did well before Sun on Motorola and Linux on Intel processors.  It seems to me there was a window of over 10 years from the 286 in 1982 until any version of UNIX did much to hurt the proprietary solutions of DEC and Prime.  During that window, Prime had a chance via the 286, 386, ….  That window is also what got Microsoft on their feet.For X/Open and Europe, I don’t have even a clue! You mean, someone in Europe built a computer system?  From all I saw about European computing, e.g., ISO/OSI, they had meetings each few months in Rome, Paris, London, Milan, etc.  with lots of Beaujolais or better, foie gras, etc. and eventually wrote standards documents only insiders could read! 

          2. Pete Griffiths

            Couldn’t resist this piece of the history:”In the 1960s, the Massachusetts Institute of Technology, AT&T Bell Labs, and General Electric developed an experimental operating system called Multics for the GE-645 mainframe.[6] Multics introduced many innovations, but had many problems. Bell Labs, frustrated by the size and complexity of Multics but not the aims, slowly pulled out of the project. Their last researchers to leave Multics, Ken Thompson, Dennis Ritchie, M. D. McIlroy, and J. F. Ossanna,[” The relationship between Multics and UNIX was intimate.There is some overlap with Unix and the period you refer to. V7 was available at that time and was making some headway but I agree that it was later releases – notably 4.2 in 1983 that critically incorporated not just virtual memory but tcp/ip.By 1982 UNIX certainly did amount to something. It was already the basis on which quite a few vendors were planning their common platform.At that time there were any number of mid range companies who had more or less decent offerings but they were all screwed (other than DEC) because they couldn’t compete for the software vendors.There was not 10 years from 1982 when the mid range guys were safe against Unix. Far from it. by the time that decade was over Unix had run away with the mid range leaving only proprietary legacy systems out there and many fatally damaged companies that never figured out how to reposition and recover. This, ironically, included the once mighty DEC.You are however absolutely correct about the foie gras and the meetings. 🙂

          3. sigmaalgebra

            As I recall, there was a Multics, REALLY popular for its security features, running in the Pentagon until it was finally pried out of the hands of its dedicated users in the 1980s or so.I don’t see that UNIX hurt DEC:  They sold hardware and, if you wanted, their OS.  People who wanted to run UNIX needed hardware and often ran on DEC.  If they ran on something else, say, Motorola 68000, then that was DEC’s fault for building hardware that was too expensive.  Supposedly the DEC Alpha chip was terrific, and I don’t know what happened to it.  Net, if DEC’s hardware wasn’t competitive, then that was the fault of their hardware, not an attack by UNIX.Yes, I knew that in some sense the AT&T UNIX effort borrowed from Multics.  My understanding, water cooler level, not solid, of the history was that Multics was in good shape in 1969-1970.But, again, UNIX started in an 8 KB DEC PDP 8 or 11 machine and was a toy.  Just when and from what organization UNIX got 32 bit addressing, virtual memory, and TCP/IP I don’t know.For the strategy issues you see, I’m mostly not seeing them:Sure, for many years IBM had its ‘good customers’, especially CIO managed, central ‘glass houses’ of banks, insurance companies, and manufacturing companies, all tied in, locked-in, etc.  These were all business customers.  Sometimes the CIO even got to play golf with the CEO and the IBM Account Executive!  The IBM prices were HIGH.IBM did a little in other areas.The DEC, Prime, Data General, Unix computing was long heavily technical, academic, and departmental where IBM’s pricing made IBM a joke.  In that computing I never saw any great fear of proprietary computing or much rush to open computing.Net, if I’d been CIO of a big bank, then I would have been an all Blue shop, paid what it cost, expected A+ from IBM, gotten it, and not tried to save money.But technical, academic, etc. computing was a whole ‘nother rodeo!Before at least the Motorola 68000, anyone wanting to run UNIX still need hardware, and DEC was a common, maybe nearly the universal, choice.  I can’t think that DEC was angry about people running UNIX on DEC instead of VMS or whatever!But also it seemed to me that when a shop bought a DEC VAX PDP 11/780, it didn’t mind also running VMS. Might as well.  Paid for it!  DEC was supporting it!For now, it seems to me that in the worlds of Apple and Android, basically the OS is Linux and the applications programming is in C++ or some interpretive language written with C++ or in Java.  So, to an applications programmer, the work is mostly C++, some standard classes and APIs, and some OS specific classes and APIs.Now, however, computing has changed again!  At least on Microsoft with .NET, that I’m developing on now, the programming is heavily just a little glue between some thousands of pre-fab software parts in .NET.  I have worked through over 3000 Web pages of documentation on .NET and suspect that Microsoft has over 15,000.  It’s like building a house out of pre-fab parts and never using a saw!  If the Linux world is not similar, then they have a long way to go to catch up!  If they are similar, then getting good with the collections of pre-fab parts is where the work is, not the actual code typed in!I don’t mind that Windows is proprietary:  Microsoft will be getting little or no cash from me until my business is nicely successful and can afford the Microsoft prices.  In the meanwhile I get to build on what is for me as a programmer the most productive platform for building a quite busy Web site.I have another point:  On Windows I writing ‘managed code’.  In part that means, essentially, “Look, Ma, no memory leaks!”.  So, I allocate objects right and left, never bother to free, and never have any memory that’s allocated but should be freed.Well, on making C++ managed code, Microsoft just gave up.  They should have.  Then Microsoft went language agnostic:  They concentrated on their common language runtime (CLR) and .NET and said, write in any language you want, whatever flavor of syntactic sugar you like, even write your own language, on the CLR, and we supply C#, Visual Basic .NET, and several more.  But I want the advantages of managed code and .NET and, thus, can’t use C++. RIP C++.What Linux has comparable with Microsoft’s managed code and .NET I don’t know, but they better have something good, WITH good documentation, and ready for 4, 8, 48, 80, or 1000 core processors soon, or the Linux penguin will be a dead duck in servers.For now, if a programmer wants to write a mobile app in whatever language and the app sells, terrific.Windows 8 on mobile may be more attractive to programmers and CIOs than Linux.For strategic issues, I don’t see them as very pressing now, and my guess is that the more important strategic issues are a bit distant from processors, operating systems, languages, APIs, and ‘open’ systems.  Maybe the strategic issues are how to negotiate cell phone connections with Verizon or movies with Netflix! 

          4. Pete Griffiths

            The reason UNIX hurt DEC is simple. Before UNIX if you wanted a mid range computer then unless you had a special requirement that could only be satisfied by some software that for historical reasons was available on someone other than DEC e..g HP & manufacturing software, then the smart buy was DEC because their large and growing market share in mid range was a virtuous circle. Bigger market grows bigger software portfolio which means the environment is attractive to more people etc etc. Furthermore, DEC was now big enough that if you were going to be locked into anyone then (as with IBM) be locked inot someone big – it’s safer. So DEC was extremely well positioned before UNIX. At this time they were expecting to overtake IBM in sales in the near/medium term! They wore badges with the date that they expected to overtake IBM at their conferences!!!The whole point of UNIX was that now all the UNIX vendors, and not long after 4.2 bsd there were many, were now a common platform and that platform was extremely attractive to software vendors – it was large and you weren’t locked into one venfor and at their mercy. This in an age of commodity hardware. UNIX broke DECs virtuous circle. They struggled to figure out what to do, never figured it out and crashed. UNIX was absolutely central to the destruction of DEC and many other mid range companies.

          5. sigmaalgebra

            You are saying that UNIX killed DEC because UNIX forced DEC to compete with commodity hardware and DEC was unable to do that. Okay.I’m saying that what killed DEC was that their hardware was too expensive. It wasn’t just UNIX: When I was shopping for computers, DEC’s hardware was significantly more expensive than, say, Prime’s.It seems that what was common to DEC, Prime, Data General (and there were a few others — e.g., it was fairly easy to use commodity bit-sliced chips to make a single board processor) is that they were all killed by microprocessors. Apparently the Motorola 68000 line, as I recall, eventually with virtual memory, was the first shot at DEC, etc. Then there was Intel and the Sun proprietary processors. Why the DEC Alpha microprocessor didn’t save DEC I don’t know.So DEC, etc. never successfully made the transition to microprocessors. So, they failed to do what Sun did. Just why Sun was successful selling desktop workstations and servers based on microprocessors and DEC, etc. were not I don’t know.Now to sell a server or desktop, have to start with processors from Intel or AMD.For a mobile device, can use ARM or whatever. Apparently Nvidia and QUALCOMM also know how to design mobile processors! So, AMD and Intel are under attack from Nvidia and QUALCOMM. It may be that VIA and Broadcom could get into the mobile processor business!Maybe one mobile strategic issue you see is RIM. Maybe they got attacked by both open Android and relatively closed Apple, both of which had the Linux developers writing for them. Okay.For Nokia, apparently they decided to go with Microsoft, and maybe they will be successful there.So a strategy is: Don’t paint the fence yourself but get everyone else in the neighborhood to paint the fence for you. Or, don’t write all the software your users will use but get others to write some of that software for you. Or, to get critical mass or network effects, attract developers and, then, more users by providing open APIs. Okay.Application: Poor RIM had to write, by themselves, all or nearly all the software their users would use while Apple, Android, and now Windows 8 have many thousands of developers ready to write software their users will use.One point I’ve made is that an operating system that, like Multics, would permit running any software without security problems would permit any user to make use of much more software without worrying about security. So, while Apple, Android, and Windows 8 are throttled having to ‘certify’ and/or ‘sign’ software as safe, a platform that could run any app safely would have an advantage.My ‘strategic’ approach is to let my users use just a standard Web browser and keep all the more complicated software on my servers. Then my users can be using any client device at all just as long as it supports a simple Web browser. E.g., my Web pages are all just 800 pixels wide! I don’t care what processor and operating system my users are using, and they don’t care what processors and operating systems my servers are using. Thank heavens for TCP/IP, HTTP, and HTML! And there should be good security for both my users and my servers.

          6. Pete Griffiths

            “You are saying that UNIX killed DEC because UNIX forced DEC to compete with commodity hardware and DEC was unable to do that.”I would rephrase it somewhat. I am saying the UNIX killed DEC because it changed the landscape to one in which DEC lost its competitive advantage (the large base of VMS software) and found itself in a world in which it had to compete on a level playing field where everyone had access to the same application software portfolio as well as commodity hardware if they choose to use it.The point about the commodity hardware is nothing like as important as the software portability. Sun was using a proprietary risc chip and still did extremely well for a long time. It wasn’t this chip that was the key to sun’s advantage and success any more than it was the 68K that built distinctive value for others.

    3. Mark Essel

      A former sys admin coworker never fails to remind me of VAX’s incredible OS. I wish it had survived, perhaps a few of its ideals can be rejuvenated.The manager maintenance issue increases in severity as a company grows. They will never all have simultaneous access to the inside of a CEO’s mind. In addition CEOs change, so if (or when) they leave with the secret sauce in their head, a company has an irreplaceable single point failure. Better to make large data sets and analysis your company’s competitive advantage/IP.

      1. sigmaalgebra

        Good point about a single point of failure.  So, if all the secret sauce recipe is in the head of only the CEO, then want to protect against the CEO getting run over by a truck.  So, have others in the company who understand the recipe.  Maybe some lawyers can work out how to do that without the recipe leaking out of the company. E.g., maybe Coke has been successful keeping its recipe secret! Maybe Google has been successful keeping the latest details of their ranking algorithm secret. Applied Materials likely has some secrets. Etc.

  20. gregorylent

    the distraction industry .. newness is its eternal salvation

    1. Emily Merkle

      we call it “exploration” … let’s go!

      1. sigmaalgebra

        You meant ‘exploitation’????????  :-)!!!!

        1. Emily Merkle

           not even close. together, hand in hand.

      2. Modernist

        What is life, if not discovery?

  21. Jennifer McFadden

    I’ve thought about this quite a bit and I think that the argument also has a sister form–is this a product or is it a company. I think that the rapidly decreasing half-life of a company is due to founders placing a greater focus on generating a lot of users early in the life of their company to either get buzz or get funding. This comes at the expense of thinking more strategically about how you will build a company, and not just one very hot product, for the long-haul. It would be interesting to take a deeper look at companies like Amazon, Intuit, Apple, and Google and really understand what was driving the product decisions. I would imagine that even at the early stages, the founders of these companies were thinking about a roadmap that stretched 3-5 years and included a number of different paths that could be taken based on the success of specific launches. This is a much harder exercise to do when you’re a start-up with 3 people and have $250K to burn until you achieve significant enough milestones to secure your Series A financing. The focus shifts to the immediate need to show traction on product A–not, how will product A be one part of an entire suite of products that we will build out over the lifespan of our company. The latter is a far more difficult exercise that requires time–a resource that is short when you are working to be first to the market to capture all of the early users in order to reap the gains of the network effects.I’m not sure if this is a result of a start-up ecosystem that promotes lean start-up methodology and financing strategies focused on short-term goal achievement over the development of a long-term operating strategy. It just seems like there is a pretty significant difference in the way companies–actual companies–are being built now vs. 20 years ago. And, yes, I understand all of the underlying factors that have lead to this shift–I just wonder if it’s worth stepping back for a few minutes and asking whether this is leading to optimal outcomes from both a company and financing perspective. It’s almost as if we’ve gone from having a few amazing restaurants to having thousands and thousands of McDonald’s. Which is better? I’m not sure that I know. The reality is that there are only so many consumers. And, they have limited time and attention. And, are drawn to the next shiny new thing. And, there are limited barriers to entry which are getting even lower every day. I’m not sure how you counter these trends to increase the half-life–or, whether you even can.

    1. Emily Merkle

       It is indeed an ecosystem that transcends time.

    2. Richard

      Could it be that most tech companies are contributing a deflationary secular market, reducing prices for everything. The valuations that we place on companies over the next 5 years are part of cyclical inflationary market.

      1. JamesHRH

        It is an awesome time to be buying impressions……price wise.

    3. ShanaC

      We know that limiting the cahs makes it easier to get a product out the door.Do you think it is worth it to wait through that process? what if the product timeline doesn’t matter if the initial product fails?

      1. Jennifer McFadden

        Shana–can you rephrase this? I’m not sure I’m understanding what you’re asking. Are you saying that you plan to delay raising until you have a product that you’ve been able to sell?

        1. ShanaC

          No that seems to be the natural right now with MVP/Lean methodologies. It seems to me that we should putting more cash in before the product gets out of the door, but that increases risks, according to what you say.

  22. Seenator

    Fred,I blogged about the same exact topic about web apps and TV shows and the difference in length between productivity apps (time savers) vs media businesses (tv sinks)http://cginsights.posterous…And I also talked about reverse network effects where I referenced Brad Burnhamhttp://cginsights.posterous…I feel so validated 🙂

    1. fredwilson

      #readlater

  23. Emily Merkle

    Fred can you kindly explain the game so those that wish to join can do so with knowledge?

  24. William Mougayar

    Wow. That’s a double-loaded post because you’re hitting on 2 themes that each could have had their own discussions: a) Web/Mobile Apps as TV shows, b) the longevity of network effects as a business.First, the Web/Mobile Apps as TV Shows: Yes, we need that and now. Why don’t we have that yet. I want to flip videos & content like I flip channels. YouTube is the closest, and they could own this if they wanted, but this space is waiting for some clever business that can aggregate and supplement what’s out there and deliver it nicely. Waiting …The second part is a heavy discussion. And I sense you’re playing a bit with us, testing to see what the community will say if provoked on certain issues that are close to them. My thoughts are:1. People are SOCIAL by nature. The Web is shifting from being Content-centric to People-centric. There is nothing stopping that.2. The big picture is the SOCIAL WEB, not Social Networks or Social Media. That’s the big theme we have to stay focused on.3. To say that “purely social apps, the ones that depend on having your friends on them, seem quite vulnerable to a mass exodus” is a bit of stretch. Go back to #1. People are social. Can you play games with yourself indefinitely? Can you have a discussion with yourself indefinitely? No. NATURALLY SOCIAL acts will remain and flourish on the Social Web. Ones that require un-natural efforts will die.4. One cannot make assumptions and extrapolations based on linear thinking. Linear thinking is like – it was like this 10-30 years ago, then it will be like that for another 10 years, or this happened before, therefore it will happen again that way. No. There are always disrupting events, technologies, themes that emerge which totally put a curve ball on that straight line. So, linear thinking has its limits and risks.

  25. testtest

    a few days ago i was hit by a meteor: http://meteor.com/ initially i didn’t realise the impact frameworks, which are used to produce web products, have. the edge they give the group using them. then it dawned on me……i couldn’t stop thinking about the implications: my body must have dumped a ton of dopamine in my system because i felt like i’d found buried treasure and won a bull fight at the same time.a few days before i was doing RESTful and CRUDy stuff like Rails. it’s great for a certain class of product. you start a project, it sets up the directory structure, sorts out how it all works together, then you can use the conventions over having to configure everything yourself–literally referred to as ‘convention over configuration’. this saves a tremendous amount of time, but is very restricting, which is fine if you want to create the type of application such frameworks were designed for (python’s django is another example).but what happens if you want a competitive advantage? incumbents have their advantages in spades. play by their rules and they’ll use those spades to dig a hole, dump you in it, and then fill the hole back up again. what they’re not willing to do is have untested technology, downtime, and uncertainty. even if that technology offers an order of magnitude greater efficiency. plus, larger companies have already committed and hired around a stack; the larger the company the more traditional the stack; they can’t hire enough developers in nascent technologies so pick another language that has a larger community with a more mature ecosystem (java, for example).stuck in an old paradigm, weighed down by past technology choices, incumbents have signed their own contract of obsolescence.moving to investors.the stock market is rigged. this isn’t a new idea. people do make money on the stock market, with the largest gains coming from value investments or momentum investments. value investments when a company is greatly undervalued: mistakenly so or during times of economic distress. and momentum investments, when a catalyst is present that will propel a stock price up, or during the inflation of a bubble where irrational exuberance takes hold and everyone piles in–leaving the ‘greater fool’ to hold the bag when the inevitable burst occurs.many of the players in the stock market are incentivised to push the stock price up as high as possible. executives are compensated based on stock price, brokers get commissions, websites want to sell advertising. all the while, high frequency trading is scalping money, inside information is being used to make more sure bets, access to additional information is granted to some and not others.i’ve been doing trading over the past couple of months and made some money. i’ve been a fool. constantly exposed to long-tail risk (the risk of something highly unlikely yet devastating happening), and as a newcomer ripe to get grin-fucked left right and center by other players with informational advantages.time better spent is time creating a company. where you have better information, and control over decisions. why be subject to the decisions of the CEOs from the companies that fred mentioned. after all, buying stock is just buying part of a company, so why not just put money behind your own company.   just because some people can’t put their money to work efficiently (which is too bad for them: boo hoo) doesn’t mean that entrepreneurs can’t make good returns on capital for themselves.

    1. Matt A. Myers

      Frameworks are great. They help the person learn faster, and make development faster.

      1. Emily Merkle

        No doubt. The thing about frameworks is that they are arrived at rather unpredictably.

        1. Matt A. Myers

          I think frameworks arise when information is evolved enough, where all of the knowledge that needs to exist does exist, where there’s perhaps an excess of information/noise from various modalities or ways of expressing the same thing – and then people attempt to create a noiseless / concise representation of what they see as the best pieces to highlight, best as to what they feel everyone else would most benefit from following.

  26. reece

    hhmm… how does this play with your theory on utilities vs. networks? or is the assumption that all competing networks can use/overtak existing utilities anyway?and how does a company stay in the game? while the core actions you mentioned (email, calendar) are set, are there ways to create new unique experiences that will outlast user hype cycles?i’d like to think so 

    1. fredwilson

      two sides of the same coin. i am evolving my thinking in real time publicly.

  27. Luke Chamberlin

    What if Ford only made the Mustang? What if General Mills only made Cheerios?They would not be billion-dollar companies, for starters.Companies make multiple product lines for two main reasons: to diversify income and to expand their market.The larger technology companies like Apple, Google and Microsoft have been able to successfully diversify their product lines (with varying degrees of success).But elsewhere, especially in the social space, we see a lot of products-as-companies. I agree that they probably have short life-cycles.Some of the really large networks like Facebook and Twitter could survive as utilities (think of the thousands of sites that rely on “log in with Facebook”).But I have a hunch that with all of the design and engineering talent at the top tech companies today we’ll start to see them diversify their product lines if they want to stick around.

    1. John Best

      I agree wholeheartedly.

    2. Rohan

      Do something small well.Get insanely good.Then do more.Well said.

      1. awaldstein

        Someone said. Do one thing completely. Exhaust your knowledge and creativity around it. Then everything else will be possible.

    3. ShanaC

      How do you diversify your network?  Are they closer to media companies (buy out another network, ala facebook) or to utilities (buy out the supply chain?)And what does that mean for feature based investing?Technology seems to be looking for a third path through the studio style short term investing and the long term investing that is normal to utilities.It is a utility that expires?  Maybe it is like selling produce or fashion?

      1. Luke Chamberlin

        What’s the difference between a network and a customer base? It’s a good question.

    4. Jasonpwright

      “Companies make multiple product lines for two main reasons: to diversify income and to expand their market.”‘Multibranding’ – a defensive strategy deployed by incumbents to deny competition access to ‘their’ market. Think Google. 

    5. fredwilson

      can the utility part of FB and Twitter survive if the consumer facing service wanes?

      1. laurie kalmanson

         as pluginsnewspaper word processing systems had messaging, which functioned like email, for a long time before the internet happened.if they had been smarter, they would have seen that other people would have liked the messaging, too

        1. LE

          newspaper word processing systems had messaging, which functioned like email, for a long timeI had a Unix system at my first company in the mid 80’s and there were Wyse terminals on everyone’s desk. We had email and even messaging. Infact if you open up a terminal window in Mac Osx (or any Unix/Linux) there is a utility called “write”. So we could message another user on the system in the office by just at a command prompt doing “write username”. Being young at the time I remember making comments about people being interviewed to one of my employees in another office. Anyway, the problem is there was no other companies that had similar systems. The potential was obvious intra company though.

          1. laurie kalmanson

             awesomeyesi worked in an office as an intern that had a wang word processing system; it was so cool at the time.

          2. laurie kalmanson

            we still use the old phone companies — but on our mobiles, not our land lines; they provide the signal and the billing service

          3. laurie kalmanson

            newspapers also could have done craigslist, but they didn’t

      2. FAKE GRIMLOCK

        NOT UNLESS BUILD LOTS MORE UTILITY.

      3. JamesHRH

        I think the utility remains but the revenue stream is less obvious: following someone you would like to learn about, be entertained by etc. has a core value that is very hard to replace. As does having a connection to everyone you ever met.TV & Movies are skewed in that their long term revenue streams piggy back on the cable industry’s need for more programming. Syndication made hit TV into a crazy good ‘hit media model’, long term business.Eventually, cable will become a connectivity service (with a huge amount of resistance obviously). Then TV will look like the web.One part utility; one part premium; one part noise.

      4. Luke Chamberlin

        Depends on your definition of “utility” (I’m still thinking about what mine is).I see twitter as a new category of communication platform. If twitter went away I would find some other service that did the same thing. It feels like more of a platform/utility to me and I think that gives it lasting power.Facebook maybe not.

    6. JamesHRH

      Luke – MS & GOOG diversified? hardly – both hit a moon shot and have done nothing else. Arguably, the same is true of the auto industry (combustion engine based transport).The last major, long term innovation in mass transportation was the SUV. Before that, the family minivan.I think the answer here is that social media is the second major wave of online advertising. The muscle cars of the internet?

      1. Luke Chamberlin

        Compared to one of the product-as-company companies above Google is much more diversified. Search, gmail, docs, YouTube, Android, etc.Microsoft entered the competitive video game industry with the XBOX and is now the dominant player in the console space.

        1. JamesHRH

          None of the products you listed for GOOG make money.MS XBOX I agree with. But the MS track record (MSN anyone?) one new products is abysmal. GOOG is worse.Check any GOOG filings…..’material revenue’ = Ad Words / Ad Sense. Nothing else matters.Don’t have the #s on MS, but I doubt that Xbox is 10% the size of Windows/Office (PC Tools business).

          1. Luke Chamberlin

            Diversification is a relative term of course. Diverse compared to what?Google used to only make money when people searched for things online.Now Google sells ads on YouTube, Android, email, etc.So if people stop searching for things on the web, Google can still make money when people watch videos, check their email, etc.I consider that diversification, even thought the business model (ads) is the same.

          2. JamesHRH

            LUke – I am being a crank. I have not seen any reports that say GOOG MAKES $ on Youtube or anything else …although I could be out of date

          3. JamesHRH

            Mike – the last paragraph states that GOOG ‘has yet to report making a profit’ on YouTube. And, they are making money by serving AdSense ads against it, so it is not really diversification (I could argue).

  28. William Mougayar

    There’s a difference between Network Effects and Friends Effects, isn’t there?FB trivialized the notion of Friends to the point that more was better. But in reality, studies have shown that 150 friends is the most number of friends anyone can efficiently handle.

  29. Richard

    Unlike something fragile, Technology and innovation “strengthen” the more they are stressed, the more we innovate the faster their advancement. Most businesses are on the other hand should be labeled “fragile” as their success is based on yesterdays technology.  

      1. Richard

        Ideas happen in bunches

        1. testtest

          wasn’t sure where you got the idea from. just liked it. he did an interview on econ talk about this as well. i listened to it twice.

  30. LE

    Said another way, if you knew that a company was going to earn $1mm a year for the next ten years and then be shut down, there is no way you’d pay more than $10mm for that company and certainly you’d pay something a bit less than that.I don’t observe the market as a whole operating with any rationality although a particular investor may. The market operates  on a “greater fool” basis if anything.After  all, if a company earns money it is worth more but not because the shareholders get a piece of that success in any significant way. It’s worth more because the stock will be bought by someone else for a higher price than they paid, in, at least, the majority of cases. Even if paying dividends, no big deal compared to the potential loss of appreciation of the stock given any number of uncontrollable or unknown events.Take GE as an example: http://finance.yahoo.com/q/…Current yield is 3.6%. Stock has a 52 week range of 14 to 21.  5 year range is over 40 to under 5.

    1. fredwilson

      i believe the market is rational over long periods of time and irrational in real time

      1. laurie kalmanson

         in the long run …

      2. Carl Rahn Griffith

        And we live in an era of ever-diminishing attention-timespan…

  31. LE

    Digtal Equipment Corporation was founded in 1957 and shuttered in 1998. RIM was founded in 1984 and in all liklihood will be gone before the end of this decade. Same with Sun Microsystems, Silicon Graphics, and many more iconic tech companies.If the company is in an industry that requires innovation and beating and keeping up with the competition, or if they have a cash cow that makes them fatter and lazier, they will inevitably fail. Simply because the workforce ages and loses their mojo. People get older, get married and divorced and spend their time doing other things.  Meanwhile they get paid more each year and they really don’t add to the creativity and productivity of the company. So they cost more and in relative terms do less. Think of your driving skills. Are you really better after driving for 30 years rather than 18 years? If so, by how much? Exactly. Not 3% more per year.This happens with both the workforce and the management. In a company with a large workforce, like, say Starbucks or McDonalds this can and is countered in a few ways.One way that McDonalds keeps its labor cost under control is by making the job such that people want to leave unless they move up.  It is so easy for McDonalds to train new people that there is no advantage to having someone working there for a long time. They would only have to pay them more to do a job that someone else could do for less.   

  32. LE

    RIM’s demise among my kids’ generation had more to do wtih everyone leaving BBM than anything else. For as long as all of their friends were on BBM, they all wanted to be on it too.Exactly what I observed in a different city with the same age kids. I wanted to buy my older daughter an iphone when it came out in 2007. She didn’t want one. Now both my daughters have iphones. Same reason – their friends didn’t use iphones.

  33. John Revay

    “RIM’s demise among my kids’ generation had more to do wtih everyone leaving BBM than anything else. For as long as all of their friends were on BBM, they all wanted to be on it too”Wasn’t RIMs (& BBM) demise more attributable to the rise of iOS and DriodConsumerization – Enterprise use to control the CPU and apps, now it seems like companies are more open to giving employees a set $$ amount and letting employees pick their CPU of choice – same w/ phone.

  34. LE

    Network effects are powerful in both directions. They can help you grow exponentially. But when they are going against you, they work just as fast. Myspace’s decline was mind blowingly quick. RIM’s has been as well. Who is next?Canary in the coal mine like the attached “wtf?”.  When companies that sell jet cards and jet shares want their target market to “like them on facebook” in an ad no less on an expensive inside page of the Wall Street Journal I think that a good sign of the part of the curve we are on.

    1. Dale Allyn

      I agree, Larry. It represents (I started to say “borders on”) the absurd. That’s an extreme example, but it’s showing up in many other “weird” areas. Why does Brian Williams care if we “Like NBC News on Facebook”? My guess is that he thinks it’s absurd too, but some producer has convinced a mid-level manager that they’re “current”.

      1. LE

        I like to give things like this the benefit of the doubt (lest be the fool at the table) but in the case of the print ad they don’t even have the colors right so right from that I know the person in charge of “marcom” (marketing communications) must suck.The use of reversed out type in a newsprint ad and the way you can’t even see the logo is just plain wrong for this print venue. (I was in the printing business and had this figured out in the first month..). So somebody is asleep at the switch in marketing. Here’s the person almost certainly responsible (courtesy of linkedin), heis in charge of flight options creative services:http://www.clevelanddesigne…Notice the same “white/black” style consistent with the ad.  Remember how Jobs got involved in all the minutia at Apple regarding advertising and product design? This is an example of what will happen at your company if you just hire someone that is qualified and you don’t know anything about it yourself.  You have to ride herd on everyone giving you advice which can be difficult if you don’t have a clue yourself obviously. Now I will be the first to admit that the success of Flight Options does not hang on the quality of their WSJ ads. But the ads are expensive and do cost money. And they should be done correctly. It’s not that hard.As far as Brian and Facebook I think that is merely an attempt to catch the halo around facebook. They are doing the same with twitter as well (with the big logo placement I mentioned). All this will pass. 

        1. Dale Allyn

          Thoroughly valid comment, Larry, of course. It all matters: typography, placement, colors, tones, gradients… And then there’s is that “little” thing called market fit and knowledge of one’s audience. :)As for the NBC/Williams/FB thing: we agree on the “halo” aspect, and that it will pass. These things are indicators of commoditization and eventual trivialization of such methods/elements. When sincere meaning is no longer a part of it, the social value is drained from it. 

      2. LE

        The attached WSJ ad, courtesy of Oracle is an example of a good print appearing on the page one. Oracle has purchased a large quantity of ads in this position where they attack IBM.Notice how minimalist it is while only driving home two key points:- 20x Faster- Replaces IBMWhat they don’t say is that the 20x figure is compared to old hardware. Pretty smart.  Don’t have time to go to school, study, take courses and listen to lectures to formally learn about marketing, advertising and design? Just pay attention to what others do and learn from that.

        1. Dale Allyn

          Exactly.

        2. ShanaC

          I find it funny that we’re all talking about print ads….

    2. ShanaC

      I see insurance companies running expensive ads asking customers to look at what other customers are saying on facebook (dangerous territory).Most firms, even social media firms, don’t really know what to do with the fact that they have customers chatting in front of them about them in front of their faces.  It is hugely difficult to use that data and that situation in order to affect branding/sales.  Most of the solutions that Facebook et al offer is based on buying ads to grow your presence (not always necessary, as seen by that ad).  I have yet to really see a deep set of tools to guide corporate communication and brand feedback quickly from this data.Really a pity, because that guidance is the thin edge of a very very expensive wedge.If jetshare wants to monitor its customers discussions, I’m all for it in context.  Sharing a jet is an expensive business, and knowing where you are failing could make the difference in profits and losses.

      1. LE

        Most firms, even social media firms, don’t really know what to do with the fact that they have customers chatting in front of them about them in front of their faces.The transparency can be totally distonic. Take the case of Starbucks deciding to discontinue a particular ingrediant. Why? Because:The insects, often found in a woolly-looking mass that covers prickly pear cactuses in Latin America, are also commonly used to color fabrics and cosmetics. But the pigment they produce is not vegan. It’s not kosher. It’s also just kind of gross, according to the more than 6,500 Starbucks customers who decided they weren’t adventurous enough to stomach the ingredient and signed a Change.org petition decrying the practice.There is no due process with all of this. A few people make a stink and then the company caves in.  Maybe they have a point and maybe this was the right thing to do. What I don’t like is the mob effect and how one story on “pink slime” can put companies in bankruptcy. Starbucks article:http://articles.latimes.com…Pink Slime bankruptcy (based on reports on ABC news).http://www.msnbc.msn.com/id…Older example of this, what happened to Audi after the 60 minutes story many years ago:http://en.wikipedia.org/wik…So what we have now is a way that the ordinary man can essentially do the same thing and start a movement that before took at least a major network or local reporter to get going. It’s so much easier it’s scary.

        1. ShanaC

          Mob effect is still a form of market behavior. Audi survived, and I now don’t have to eat pink slime, which makes me happy.

  35. LE

    “Who is next?”Oh yeah, linkedin as well. When you start getting multiple connection requests from people you’ve never heard of or care about that don’t take the time to send anything but the standard message the network will start to go in the opposite direction and becomes significantly less valuable.

    1. Alexander Close

      Too much clutter on Facebook makes room for Path.Too much clutter on LinkedIn makes room for ?.I’m not yet sure if a parallel lines can be drawn, as the two services are used in very different ways.  Maybe.

  36. kevinmarks

    TV shows these days are made by production companies, that are often controlled by individuals, or small teams. In some ways classic Startups are more like Movies, where a company is formed for the project (Film), they get backing from the Studios (VCs), have a big release, and hope to make enough money from the public to make a sequel.TV Shows have the sequels implied as episodes, so are more iterative. Of course nowadays we have YouTube, where episodic media can be created on a whim, with little need for Studios (VCs) and often don’t even need Producers (Angels). Get a camera, shoot, edit, post, repeat. Lean Movies? 

  37. Siminoff

    Would have been so interesting if RIM had opened up BBM at its peak. I wonder what would have happened to them then?

    1. fredwilson

      that would have helped. building a better OS & browser and a truly open app market would have helped too

  38. takingpitches

    Fred and all,Almost everything that Facebook has tried over the last year in expanding into others’ markets has been a bust, confirming that its hipness is long gone.  I love your use of the word “utility” because taking advantage of the fact that users check out Facebook periodically by habit and positioning Facebook as a platform for actual hip social apps to launch is really how FB should position themselves going forward.  I have a post from a few weeks ago on this, which I would love your thoughts on:http://takingpitches.com/20… I also love your point about reverse network effects. The corollary of this point is that we have a tendency to exaggerate user switching costs because of sunk investment.  I also posted about this last year:http://takingpitches.com/20… 

  39. Guest

    Do you mean every thing has a lifespan and it ranges from long to short for pure utility to pure entertainment?

    1. Emily Merkle

       you got it.

    2. William Mougayar

      Not necessarily. I meant that these companies are to fault, not the trend or product category they were participating in. Product cycles do evolve and sometimes new ones obsolete the previous ones. Incumbents that don’t move with that trend can disappear.

  40. William Mougayar

    But the demise of MySpace had nothing to do with Social Networks going out of style. FB built a better Social Network experience & marketing around it & it took off. DEC, SUN & RIM shot themselves on their own. HP went on to own the mini-computers space & we know what Apple did with smartphones.

  41. LE

    Retirement just means being able to do what you want, which is where I want to position myself as best I can – to enjoy life with what I am doing.Mike Wallace answered that issue well in the show they did on him and his career.  He had no interest in retirement and it was amazing how it isn’t obvious to people who would ask that question why he wouldn’t want to retire from the job that he had.  I agree with Wallace. To me working is a means of being able to do what I want to do every single day as much as I want which tends to be “all you can eat”.  My dad didn’t particulary like his import business when I was growing up, but after he sold that business he put time into making money in other ways that he enjoyed. And he is still doing that to this day and his mind is more active and alive than it was 30 years ago.

  42. Tom Hughes

    I attended the Amazon Web Services Cloud Summit in New York on Thursday and had a similar thought. Traditional IT management and product development is (was) like managing a factory: the essential problem is to wring the most value from your assets. Modern IT management and development under cloud computing is, for better or worse, more like running a movie studio: most assets are liquid, and the challenge is to pick the right opportunities and match them to the right assets. In the old model, there was some correlation between size of investment and return: bigger factory = more widgets at lower cost = market success. In the new model, that correlation is gone, or it comes much later and is called “scaling.”In this new world — to your original post — when failure comes, it comes quickly. It’s not like “John Carter” was an OK movie that stumbled along: it was a big flop, and obviously so from the day of release. 

  43. Andy

    “Utilities like search, email, calendar, document store, etc feel less likely to be subject to this issue.”They may be less subject to the issue, but they’re definitely not immune. For example I use to have a Yahoo Mail account.

    1. Tom Hughes

      Good point. For a utility, the route to security is to become the provider that everyone uses because everyone else uses it, like Google Search and Facebook. Decisioning products like ratings can capture this: everyone still uses Moody’s and S&P, in spite of everything, because their network lock-in pervades credit markets.

      1. ShanaC

        Some of that is legal.  It would be much easier to switch if certain kinds of pension funds were legally allowed to.

    2. fredwilson

      good point

  44. loteq101

    To my way of thinking, apps are the new channels at least where video consumption is concerned. For apps like say pbs, npr, cbc, abc etc these media outlets have an app, and they push vide and content through it, to me this is is now a channel. For other more focused experiences, like words with friends or angry birds etc, these are still just games… They dont offer much in the way of broad experience. Even network connected games such as words with friends are still a focused experience and i know what im getting when i go there so they cant really be a channel.Ive tried to encourage execs i know in the tv business to put more push behind their apps, as realigning their businesses around this will help them get their feet wet with where this is all moving (and fast at that) and think of new content that will drive adoption of those apps. If they dont figure this out, theyll just be another channel on youtube, and be leaving a lot of revenue on the table for google. At least if they have a strong app ecosystem and are moving ads through that they can do youtube etc as well, but it will focus people towards their own apps.

  45. george

    Investing in the Internet wilds requires tacit knowledge, and providing some cautious optimism is great advice! I believe that the $1B purchase of Instagram illustrates several of your points and how  susceptible any platform can become with variable change and/or a quick strike on the right pressure point.I think the fatal trap today is that products/services are less resilient; change in attitudes and behaviors happen much faster. I agree, the lifecycle on the S-Curve model is accelerating and many companies are struggling to adapt, which as you point out, recast valuations.What’s inevitable here is that outcomes are not a sure thing and sustainability is where I place a premium.   

  46. Dale Allyn

    This is such an important topic, Fred. Great post (again). In my view, there are distinct and different models in the Internet sector and each can be valid, but it’s important to understand them and to know in which one is participating. I’m not a fan of “build it to flip it”, but it’s easy to see the attraction to both entrepreneurs and investors. Same goes for building to be acquired as an acqui-hire. There is value to be had by the “feature” projects or the quick flips, but what excites me is projects designed with a vision of and for sustainability and meaningful value. Of course, one can define “meaningful value” in many ways based on one’s own value system and metrics. I want to see companies go 10, 20, 30 years and contribute to improving conditions for all, but at the very least provide improvement in the lives of the direct users of the product or service. Love the topic and the comment thread (as usual). [edit: typo]

  47. John Best

    To my mind, the “channels” are the serial entrepreneurs. They’re the Ford, Apple etc. People like Sean Parker, who can take an idea and make it fly. Then do it again, and a third time.The programmes are their Facebooks, Spotifys.I think the TV show comparison is apt. If you look at some of the biggest TV of recent years, its captured the imaginations of millions. Similarly tech businesses. However, every programme has its shelf life. Some last a few seasons, some decades, but without iterations, they can become stale and economically nonviable.

  48. Matt A. Myers

    This confirms a bit for me as to why Zynga paid so much for Draw Something.I had a belief that due to the userbase being mobile, and Draw Something looks super fun, engaging and is a very creative, interactive, and quick to use application – that Zynga most certainly saw a drain from their userbase / hours played, etc.. There are certain ‘games’ online that really function as more than games though for people, creative outlet-related games are definitely more attractive, and they serve a need of letting people expel their mental energy somewhere. Perhaps their runway is just a bit longer before people move onto the next though.I do hope and see the future moving towards real social engagement more and more though.

  49. ShanaC

    So does that mean you’re the harvey weinstein of it all?Technology behaves somewhat in between to utilities and entertainment.We all need technology, and we all need it to get better, which means technology expires.  But at the same time, it works closers to a utility in the terms of “WE NEED IT” the same way we need water, or air.I see this as walking a third road where technology that works will continue to stick around, and technology that doesn’t will fade like a bright star getting kicked off the air.And it is a fact irrespective of communities.  people may hate the object in question, but if they need to use it, they will, forever and ever.

  50. Brad Lindenberg

    My favourite line was this one:Utilities like search, email, calendar, document store, etc feel less likely to be subject to this issue.I wrote a blog post about this recently:http://www.lindventures.com…What this comes down to, is that these apps are NATIVE on all devices. Email, Calendar, Twitter now on iOS, SMS… Look at Twilio – they’ve built a business on what is a “native” app on every connected device. Same with Sendgrid. If you can build web services that harness the native apps of smart phones then your users dont need to be educated to download an app or do something before you can use their service. It removes a very significant layer of friction, however it also depends what your service is. You can’t run Draw Something on email… However Draw Something will come and go. Email, Twitter, Camera, SMS, Calendar etc will not.

  51. Peter Tippett

    This is such a great piece about how transient some areas are. When you apply this to any other market, you see this same pattern happening, but at 1/10 of the speed which gives many time to adjust. Utilities and passion are the constants that you need to tap this for a long tail business and design for the expected change which is something that the startups of today are all about.Many get caught in the trap of “I have made it” and move to protect instead of embracing the change.  Instagram can be looked at from the angle of Mark knowing there is change about and decided to embrace it by buying someone who is already in it with success. This adds that to their own DNA that they can leverage much more effectively and knowing that this is a fast changing area where they don’t have time to solve it themselves due to the speed of change.  Acqu-hires is another way of looking at this, but that is more a early stage strategic decisions that may have an outcome.So, to succeed today in any thing related to technology, your core vision may stay the same, but the execution will have to change is a fact of life and if you don’t, you will go out of business or become a “dead man walking.”

  52. Pete Griffiths

     Clayton Christensen’s ‘The Innovator’s Dilemma’ is a book to reread every six months.He is also charming and articulate in person.  A nice overview in a Techcrunch interview here:http://techcrunch.com/2012/

    1. fredwilson

      i’m a huge clay fan

      1. Pete Griffiths

        Yes. Classic invaluable work.

  53. SCKD

    Given BBM’s demise is largely due to the iPhone and their messaging service essentially does the same thing, why are you so positive about Android which will never be able to have that same option (or free calls), for fear of losing their customers (providers)?

    1. fredwilson

      i don’t understand your question.

  54. PhilipSugar

    I think the quick answer is in technology if you don’t kill your existing business somebody else will.So worrying about extending your current successful line in technology is not where you want your best people.You also don’t want to invest all of your resources there because the profits can support the expense.You want to milk it for all the profits you can.  Extract 80% net margins, not gross net. Don’t grow to fit the size of the aquarium.You want best people and most of your resources trying to kill your existing line.That is the brutal world of technology.Some days that’s why I want to wake up and be a plumber.

    1. William Mougayar

      Exactly. It’s better to shoot yourself in the foot rather than having someone else shoot you in the head.

    2. fredwilson

      indeed. but it would not be as fun.

  55. Esayas Gebremedhin

    fb is next. i said “be tool not media”. recently i was thinking apps are like music, you hear it, you love it and than you want the next smashing track and the old tracks just sound empty. i can imagine that the evolution of apps from one generation to another can be avoided by being inventive in very short iterative loops. cause people love the spring with it’s flowers and it’s fresh green leafs.

  56. deancollins

    So can any app really be worth $1b if its going to live and die in a decade? (though most tv shows get 3-4 seasons if that).Shouldnt we start pricing investments and exits based on sure you are hot today but what are you going to do for me tomorrow?

    1. fredwilson

      that’s my point Dean

  57. John Revay

    Email as a Utility – several years ago a @aol.com email address was big, then @hotmail.com   Now it is @gmail.com  with some occasional me.com addresses.  What will be next – I think it was Sheryl Kara Sandberg from  FB that predicted the end to email some day, even though they seem to still rely on it.Fred – thank you again for your time last week –  hoping that a Networked Utility will still have a chance.#IfWeBuildItWillTheyCome

  58. Jasonpwright

    You invest in the web for tomorrow. People use the web for today, which then becomes yesterday.

  59. Jasonpwright

    I’m not sure that the evolution of the web is linear. It may be a spiral.

  60. Jordan Elpern-Waxman

    Great post, but I take issue with the idea that DEC, Sun, SGI, and RIM went to nothing. – DEC was sold for $9.6Bn to HP in 1998 in what was hailed as the largest merger in the computer industry at the time. – Sun was sold to Oracle for $7Bn. – RIMM may have declined a lot, but it still has a $7Bn market cap and will go for more if acquired. These are hardly insignificant terminal values for a DCF valuation. SGI really did go to zero (effectively), but so do many 28 year old companies in all industries.

    1. fredwilson

      great points. you can catch a falling knife!

      1. Jordan Elpern-Waxman

        Or at least find someone willing to pay you for the right to catch it!

  61. Ryan Bloomer

    I totally agree and I tend to cringe anytime someone says they have a mobile app, but at the same point cringe morewhen that same mobile app then sells for a billion dollars.  Catch 22 but I agree with you Fred it sucks nothaving a chair at the end of the song. But what is worse not having a chair or never playing the game?

    1. fredwilson

      never playing

  62. Dan Sweet

    This reality check that business cycles, companies, and entire industries come and go is especially interesting in this era where more and more people are talking about wanting to remain private and never sell.Interestingly, an awful lot of deal economics just plain wouldn’t work without that terminal value being 40-60% of the total deal’s valuation.  Even probability-weighting the terminal value with a 25% “space no longer exists” variable would probably slow a lot of people down as well.  Food for thought.

    1. fredwilson

      yup. that’s my point

  63. FAKE GRIMLOCK

    ANYTHING HUMANS USE BECAUSE WANT, NOT NEED, VULNERABLE TO FAD.FACEBOOK, TWITTER, BOTH IN THIS CATEGORY.

    1. laurie kalmanson

       hula hoops, yoyos; but they seem eternal at the time

  64. jason wright

    Try eating soup…with a fork.

  65. Carl Rahn Griffith

    It’s a whole new economy subset, a whole new product lifecycle. Old rules go out the window.It’s a reflection of our disposable mindset.It’s also a great opportunity.It’s also a lot more reliant on zeitgeist and luck…

  66. Mark Simchock

    Perhaps the turnover is a function of size? As they say, the bigger they are, the harder they fall.Also, maybe the ideal model looks more like Branson’s Virgin than a single massive vertical trying to optimize and synergize everything? Sure, keep the parts working together where and when it makes sense but look to spin them off and let them fend for themselves, etc. Maybe there’s danger in being too focused / optimized?In other words, sticking with your TV theme, could the Instagram purchase be FB’s jump the shark moment? Things should spin-off Facebook at this point, not be sucked in, maybe?

  67. Jeff Schrock

    After building, buying and financing tech co’s for last two decades, i am firmly of the opinion that  most emerging tech companies today should be thought of (and financed) like entertainment products.There are clear differences between the markets.  But from a durability standpoint, there are just not many tech companies (particularly internet media based) that demonstrate the longevity or distribution of previous tech titans.  Put another way, Madonna has been monetizing longer than any internet company.I also think most investors would be much better off if they thought of start-up companies as oil wells, movies, or even tv shows instead of as future public companies.  For the most part, they are speculative enterprises with compelling product level economics and questionable sustainability as a corporation.  

    1. fredwilson

      that madonna line is classic. i may use it myself!

    2. Mark Simchock

      Perhaps in the upcoming (?) film “The Venture Capitalist” Simon Cowell could / should play Fred Wilson? 🙂

  68. Brian Hayashi @connectme

    The rapid upswing and downstroke of internet businesses requires managers to ask different questions. Emerging Internet startup syndicates must do things that are similar to yesterday’s nascent studios like Desilu and Paramount. Seven years ago, then-CEO Lloyd Braun had some interesting comments about program economics:”…I’m giving great thought that as we construct these models, that in success we are not going to create a system that does exist in television now where there are such crushing upfront costs that the whole weight of the system makes it feel like it’s going to break. We have to be thoughtful not to just look at this business where it is now but where it will be five, 10 years from now. Because once we start establishing these precedents, they get very, very hard to break.” The studio model allows FOX to efficiently wind down a series like “Terra Nova”, even though it pulls in an otherwise-respectable 7.5 million weekly viewers. Look at the internal dynamics of a Y Combinator or 500 Hats — they are starting to organize themselves in ways very similar to a movie or TV studio. Even Zynga had its own “upfront” two years ago and found success. Where does this end up? I predict that the companies that are most successful at obsoleting their own services will be the most profitable.http://connectme.typepad.co

  69. Youssef Rahoui

    It holds some truth I think and what I like in that is that it keeps the big names awake at night: they know their time is limited and if they ever want to be on top for more than 5 or 10 years, they need to innovate, to buy startups, etc. It’s like knowing you are the president vs you are the king.

  70. fredwilson

    They are coming in

  71. Matt A. Myers

    Nope, though I’ve heard things from it.They won’t ever be back at their level at prime – the market will change too much by then.And it may take 3-5 years from changes now really show their new value.