Finding Signal In The Noise Of Demo Days

I'm taking a break from MBA Mondays today because Christina Cacioppo just posted a terrific blog post on the USV blog on her takeaways from a summer attending most every demo day imaginable.

Christina notes:

Over the past few months, I've seen over 160 companies come through eight different accelerator programs. It's a skewed group, but it captures the zeitgeist of a certain segment of the tech industry – and, I think, looking at these companies is one of the best ways to get a sense for which opportunities compel internet entrepreneurs today. Here's a look at some of what these entrepreneurs are thinking about – and where we all might be headed.

Over the course of the summer and into the fall, my partner Albert's office whiteboard started to fill up with sticky notes of various colors. Christina was mapping out all of the startups she was seeing. And in the end she came away with two big megatrends.

Go read her post and find out where they are.

#VC & Technology#Web/Tech

Comments (Archived):

  1. Aaron Klein

    It seems like a lifetime ago, but earlier in 2011 when we raised our friends + family seed round, the big questions from these non-veteran tech investors was “how can you possibly do what you’re talking about for this amount of money?”And it was incredible to tell the story of how much the world has changed between 2006 and 2011. We used to spend thousands of dollars a month on servers; now it’s all virtualized and metered like the power bill. We used to spend thousands of dollars upfront on software; now we rent it at scalable pricing.These changes clearly won’t make startup ideas any more successful than they were before; they’ll just let us figure out what works and what doesn’t a lot faster and more efficiently.

    1. Rohan

      going up the S curve. ūüôā

    2. awaldstein

      True.It certainly takes less resources than it used to (thankfully!) but does it actually take less time to build your market?It feels like it does because time is compressed on the social nets now that we all live on a 24 hour clock. But with all these thin sliced pieces of capabilities, the competition for markets and users and true engagement is fierce.I wonder if the reduced cost of building and maintaining gives us the time we need and as that time to capture market is not really reduced. Building a market and a brand takes time.

      1. Anthony Ortenzi

        It’s not just cost that’s avoided, but also opportunity cost.The flexibility afforded by not having capital costs sunk into hardware and software allows you to chase the market when you see that you aimed wrong.

        1. awaldstein

          Opportunity cost aren’t really avoided as anything you spend your time on churns that core resource of yourself.I agree with you in concept, but I think it’s incorrect to think that pivoting isn’t a huge expense.Being flexible in infrastructure is a gift we have as entrepreneurs today certainly. But focused selling and discovery to find your market and iterate it is not the same to me as shifting gears, shifting communities and finding a new path if the market is not there.We are lucky to be able to do that but I’m a believer that focused direction finds the market. Not a big proponent in the pie against the wall theory of product or market development.

          1. Cam MacRae

            Agree – Pivoting is more like pulling out of an unintentional dive 300ft from the ground than avoiding the dive in the first place.

        2. FAKE GRIMLOCK


      2. Aaron Klein

        I completely agree. The time cycle has not compressed much at all.Nine women can’t make a baby in one month, no matter how cheap the infrastructure gets.If anything, the lower cost of infrastructure creates the illusion that you can try anything and place a million bets when you really can’t.

        1. andyswan

          Bingo.  I know at least 8 guys that have tried at least 6 ideas in the last 2 years.No one seems to know that the real work begins AFTER the idea is turned to reality.  And it takes a long, long time.

          1. Aaron Klein

            So true.¬†@stevecheney:twitter¬†had a great post on TechCrunch about the “overnight success” of Pinterest that actually took about 22 months.That’s what has me telling my team, no matter how excited we see people getting about our product, “the hard work is still ahead of us.”Not sure how much we like hearing it, but it’s the truth.Edit: Whoops, forgot to post the link.

          2. FAKE GRIMLOCK


          3. Rohan

            ‘It takes ten years to become an “overnight sensation”‘ – Robin Sharma¬†@aaronklein:disqus¬†

          4. jason wright

            It’s like a cell culture in a test tube. Start with one cell, ¬†dividing once every second, and the test tube is full in exactly 24 hours and no minutes and no seconds. After 23 hours 59 minutes and some odd seconds there appears to be ¬†next to nothing in the test tube. The visible action all comes in the final 30 seconds and the test tube is then full to the top.¬†I prefer this analogy to the iceberg 1 and 9. Icebergs melt.

          5. laurie kalmanson

            true with babies, too

        2. Christina Cacioppo

          only sortof related — but i read the “nine women can’t make a baby in one month” quote for the first time, last night, in one of buffet’s essays. nice to see it show up again so soon after!

          1. Aaron Klein

            I think Buffet is where it originated from, but it’s one of those word pictures that has stuck with me. I wish I was better at inventing them myself!

          2. John Clyman

            I suspect the origin is actually The Mythical Man-Month.Buffett is a wise man to be sure, but it’s amazing how much of what Brooks wrote 35+ years ago is still relevant today!

          3. Rohan

            That’s a fantastic one!

          4. paramendra

            I have seen that quote misused in too many non baby situations. 

        3. FAKE GRIMLOCK


          1. Aaron Klein

            well put!



        1. awaldstein

          So true…Dino wisdom!

  2. Rohan

    Interesting. Thanks Fred, Christina.Agree completely on technology becoming peer to peer.¬†I think it’s a part of the whole phenomenon of people empowerment. As the world is becoming more connected, we are going back to ‘tribes’ in a funny way – except it’s not geographic, but based on interests.Fascinating time, this.

  3. Rohan

    And continuing on the quote of the (week) day feature. Today’s quote is lifted from Fred’s tumblr.‚Äé’It’s not happy people who are thankful, it’s thankful people who are happy.’¬†Lots to be thankful for, indeed.. :)Have a great start to the week, all! ūüôā

    1. fredwilson

      You liked that too, eh. I am very thankful and quite happy too

      1. Rohan

        I am thankful for that and happy for you as well..Hahahaha

        1. William Mougayar

          4 Ha’s? You’re totally throwing me off now…Are your Ha’s on a scale of 1 to 4?¬†HaHahaHahahaHahahahaI don’t mean to pick on you. We often forget to laugh, thanks for reminding us Rohan.

          1. Rohan

            ‘Never be afraid to laugh at yourself, after all, you could be missing out on the joke of the century.’haha.Not feeling picked on at ALL William.What’s life if we don’t laugh, eh? ūüėÄ

          2. Carl J. Mistlebauer

            Humor, the tonic that cures all….Suffering from a hearing loss and finding hearing aids an annoyance it took me a while to develop the confidence to laugh at myself….Now, if I answer a question and I get a bunch of blank stares I just say, “…well, obviously I didn’t hear the question but you got to admit that my answer was awesome and obviously the question that I answered should have been the one that was asked.”

          3. Rohan

            @tao69:disqus¬†Nothing shows a man’s character more than what he laughs at..¬†I feel your story is testament to that. ūüôā ¬†Thank you for sharing!

  4. ShanaC

    I have a large question about component based startups: ¬†Esentially, they commoditize the building of software. ¬†Will more higher tier services that use these component startups be more easily built and rebuilt causing audience fractures (along the lines of cars, there is a car for everyone, and they are built essentially the same way)And again with work- will this push prices into funny directions when it comes to salary if we’re all replaceable mechanical turks?What does it mean for innovation if we take people out of the picture?

    1. Aaron Klein

      I don’t think it’s taking people out of the picture. It’s just moving the value chain around.Go back to the 1920s. Car companies didn’t lose people because components got standardized – it allowed huge value creation, growth and hiring.

      1. ShanaC

        Commodotized people, then?

        1. Aaron Klein

          People have always been commoditized, if they fail to differentiate and improve the skills they can offer an employer…

          1. ShanaC

            I feel very meh about this. Very meh

    2. LE

      “And again with work- will this push prices into funny directions when it comes to salary if we’re all replaceable mechanical turks?”Prices will be driven down. Loyalty allows you to raise prices. Making the work a commodity will do the opposite. ¬†It becomes a “labor unit” that can be replaced. ¬†There is a saying that is common to many businesses and exists in various forms.”price quality speed” pick any two.¬†

      1. ShanaC

        well it also sounds like that if your standout, the price you’ll be able to demand is going to be very very high. ¬†Sort of like a lemon market.These sorts of trends could make doing a startup very weird, your market got much poorer

        1. LE

          “if your standout, the price you’ll be able to demand is going to be very very high”Typically what you see in any personal service business is the following behavior (if the player is smart enough to maximize profits) because the labor is fixed and you gain clients over time.a) Start off taking any and all business.b) Weed out low margin or high maintenance clientsc) End up with top loyal clientele (allows “d”)d) raise hourly ratese) pass “b” to someone else (an associate for example).You see this happen with lawyers and accountants as they build their practice. In the end either supply goes down or prices rise or the practitioner doesn’t care about money. Make sense?Traditional business (say a supply house):This is one of the reasons to buy an existing business rather than starting from scratch. They have already weeded out the customer base that would take you many years to build. Typically a new business most easily picks up the clients that are a pain in the ass or not worth the time of the established player (if they were getting value for what they gave and in equilibrium they wouldn’t be looking for a new vendor.) Gross oversimplification but that’s the way it works in many cases.

    3. sigmaalgebra

      “Esentially, they commoditize the building of software.”You mean like organs, violins, and choirs made Bach unimportant, pianos and orchestras (with ‘components’) made Mozart unimportant? For some of the magic:…Computers are still dumber than paint. The only way computers appear to be ‘smart’ is when there was some programming by a smart human.It will be a very long time until we can “take humans out of the picture”.

      1. ShanaC

        No, ¬†I think this will be like cars. ¬†Think back to the photosharing app that is being talked about above. ¬†In a “piece by piece” startup world when ti comes to the back, and a very open labor market for the front, how much will building 5 more photosharing apps for different types of people cost? ¬†A lot less when both the people and the software are essentially hot swappable.

        1. sigmaalgebra

          Shana,To be clear enough, I will have to give a longer response!I claim that ‘components’ are necessarily inevitable. Then commonly old apps get ‘commoditized’. Then the components help us get some progress.For components: As we write apps, we use ‘divide and conquer’ and write the app in ‘pieces’.Surprise: Commonly the more narrow the work of a piece, the wider its applicability. Call this the Iron Law of Component Generality (trivial idea but roughly correct).So, ‘pieces’ from our use of divide and conquer become reusable ‘components’ whether we so intended or not. So, the components are necessarily inevitable.Then the components get used and, commonly, ‘commoditize’ old apps.This is an old story and goes back to the wheel: It made carrying fire wood easier. But then the wheel also made carrying water easier and made possible chariots, trains, cars, etc. The same holds for a large fraction of our tools from metals, electricity, …, computers and software components.But, there is a bright side! One ‘app’, YouTube, was maybe mostly just for kitty cat antics? But it can also be used for:St. Paul Chamber Orchestra…Heifetz — Bach Chaconne……Presentation of the Rose…Renata Scotto. “Un Bel Di”…That’s a lot of music to get from something for showing kitty cats!Note: The ‘Chaconne’ is not really ‘easy listening’.Yes, there may be some ‘totally available’ former photo-sharing app developers. Next time they should aim higher and, then, they will face less danger of being ‘commoditized’.This situation is part of the path to progress!I submit:(1) We’ve made progress for thousands of years.(2) The results in total have been fantastic.(3) Mostly the progress was made in difficult steps.(4) Maybe most of the best progress was made by fewer than 1% of the people.(5) Still, the 99% had their lives improved, and maybe now 89% live better than the 1% 100 years ago.(6) There is more progress to be made, MUCH more, and the components will help — in simple terms, and nearly necessarily and inevitably, we want several factors of 10 in increased economic productivity and that heavily via computer-driven automation.(7) Currently computing and the Internet are the best foundations for progress ever.(8) The progress won’t all be easy or a ‘commodity’ and may be mostly just from the 1%.(9) Still the progress will help at least the 89%.I’ll submit one more: The good stuff is necessarily rare and, thus, also necessarily difficult to plan accurately or even to evaluate in advance, especially for the 99%.Such is part of the way of the world.

  5. Carl Rahn Griffith

    Nice and succinct – I love the physical manifestation of thoughts/analysis/ideas – the lack of a Whiteboard in a company always alarms me. Old school, maybe, but there is no substitute.Excellent, very useful! Thanks, Christina!

    1. awaldstein

      whiteboarding is a frame of mind.

      1. JimHirshfield

        Waterboarding is not.

        1. Aaron Klein

          Or pepper spray, for that matter…

          1. JimHirshfield

            Pepper spray is _definitely_ trending.

          2. Aaron Klein

            Sadly.I don’t find myself agreeing with the protestors for the most part, but I can’t believe the stupidity of UC Davis leadership with that incident.Their PR people were very thankful for Thanksgiving itself this year…it’s the only reason that Chancellor is still in her job.

  6. gregorylent

    “granularity” is common to both those trends … individuation, is perhaps another word. of both tools and intention.

  7. JimHirshfield

    Great to hear an answer from VCs to the question “What are you seeing out there?” that isn’t just “Mobile, Social, Location based…..”Thanks Christina

    1. Aaron Klein

      If “photo sharing” was still a mega trend, I think I would have hit my head on my airplane tray table this morning. ūüėČ

      1. JimHirshfield

        Ha! Reach for the air sick bag?

        1. Aaron Klein

          Fortunately, we decided not to build location-based social photo sharing into a tool to revolutionize how we make investment decisions…so no, I didn’t need the air sick bag. ūüôā

    2. Christina Cacioppo

      I think “Mobile Social Local” is only an acceptable response if you tack on a “Bacon” at the end.¬†Related:¬†…

      1. JimHirshfield

        Ha! I don’t tape bacon to anything; I’m vegan. :-p

        1. JamesHRH

          You just fell off the #NOEATFRIDAY list.@FAKEGRIMLOCK:disqus fresh meat!

          1. JimHirshfield

            Ha! That’s too funny!!I’m gonna guess fakegrimlock prefers meat-fed meals. Herbavore not so tasty.

          2. FAKE GRIMLOCK


          3. JimHirshfield

            You callin’ me a garnish.

        2. ShanaC

          you’re a vegan – how do you get your b12?

          1. JimHirshfield

            Abundant in nature, contrary to popular belief. Nonetheless, everything from soy milk, cereal, to Smoothies are pumped chock full vitamins and minerals.On Nov 29, 2011 6:47 AM, “Disqus” <>

          2. ShanaC

            Note to self: if I ever want to cook a vegan meal, talk to you before going about so I make sure everyone gets their vitamins.

      2. William Mougayar

        Very funny deck. 

  8. laurie kalmanson


  9. LE

    The three companies that stood out to me after reading¬†@cristinacaci:disqus¬†post are: – magazine quality writing – proofreadingand to a lesser degree (freelancers)(Side note the only company that comes close to having a ok name is “contently”. The others are totally random and would have no meaning if I tried to recommend them to a customer or someone I met at Starbucks. They have near zero pass along value. )Anyway, I like these companies because they connect via the “peer-to-peer model” vs. the “30 of the companies” that Christina saw that developers are the target model for. I don’t like the companies that target developers because I see that as particular to a point in time (now) when there is high activity in startups that I don’t feel will last ¬†and create a sustainable business long term.¬†The problem that nobody has solved though is the age old one. How do you get people to know about these services? ¬† And to remember that they exist when you need them (I have a wiki that I use to make note of all the things I trip on and read about but my guess is most people don’t do this. Maybe they make a bookmark if at all. And still others do nothing.)Solve this problem.In the real world someone can come to your door or send you postal mail saying that they do plumbing. They give you a refrigerator magnet so you can find them when you need them. Or they give you a sticker to put on your HVAC so you can find them if you need them. There is no equivalent thing with all these great peer to peer services.

    1. Christina Cacioppo

      i went back and forth on whether companies targeting developers are Good or Bad.¬†the answer is, of course, “it depends.” (or perhaps: “neither Good nor Bad.”)though if you believe the world’s increasingly going to be dominated by software, i think you also have to believe that serving developers will be an increasingly large and attractive market.

      1. Carl Rahn Griffith

        With the prospect of so many new developers out there, the market will sure need a lot of QA tools! ūüėČ

        1. JamesHRH

          Testing tools have along and profitable role in tech, for sure!

          1. Carl Rahn Griffith

            Indeed. Shame more contemporary devs don’t seem to use them – or at least go through good old fashioned Alpha > Beta > GA phases before being unleashed on the poor unsuspecting public ūüėČ

          2. Matthew Tendler

            True QA testing (like what was used in waterfall models) is passé these days. Just like revenue models.

      2. LE

        Developers are fickle and eager to hop on the new thing. Sure they can be brand loyal but then a new kid comes along and takes things in another direction or has a cool twist. ¬†(And, agree, of course it depends on the particular product/service they are offering..)Anyway, fickle. I’ve seen this happen over the years with programming languages. The new language always avoids the problems of the old language and the adoration is non-stop. ¬†It’s this. It’s that. It will do everything elegantly. It will help you get a date.¬†Perl->Php->Ruby; ¬†Mysql->Mongodb | Nosql (or put whatever you want here).¬†Anyone remember dBase and Ashton Tate? That was the shit back in the 80’s. ¬†Conferences, books everything.¬†End users, consumers and business for that matter are more slow moving and definitely more brand loyal until they get knocked out of orbit by something amazing that solves a problem.Not just something that makes things a little better or is fun or cool.They don’t have time (business) for fun and cool.¬†One company town.Finally, if the startup gets its business from other startups primarily (as opposed to enterprise or small business) and we know that the majority of startups are going to fail (and will fail at a higher level than a typical business failure rate) where does that leave the developer centered business when that inevitably happens?

        1. Sebastian Wain

          “Finally, if the startup gets its business from other startups primarily (as opposed to enterprise or small business) and we know that the majority of startups are going to fail…”Our company experienced that issue exactly in the 2007 crisis. We were selling software components and our main customers were startups, so the company was forced to think harder and search more stable customers.

      3. Cam MacRae

        If you’re a service you probably need to give it away or offer something so compelling that the devs are willing to spring for it themselves until some critical mass is achieved. Both gigs seem pretty tough.



      1. LE

        efridgemagnet.comIf the startups weren’t sooo freemium someone could make a living being a manufacturers rep for a variety of startups.…Gasp. A middleman.¬† Oh we are getting rid of those. See if there is no money in what you are selling there is no money to spend on advertising and marketing to let people know about the product or service until, if ever, you achieve critical mass.

  10. andyswan

    Spot on.If the future isn’t all about individualism, I’m gonna die a broke man.

    1. Aaron Klein

      If the future isn’t all about individualism, we will collectively die broke ūüėČ

    2. fredwilson

      i’d bet against that last part

      1. andyswan

        You might be able to!

    3. Mark Essel

      I wouldn’t mind going out with a neutral balance sheet.Can’t take it with me, and the Feds will tax what doesn’t go charity. Giving to charities is more fun while you’re alive, much like everything else ūüėČ

    4. Carl Rahn Griffith

      We’re all individuals, Andy… ;-)…

    5. jason wright

      Tribes have a long history. You just need to find the right one.

      1. JamesHRH

        @andyswan:disqus¬†Maselow’s Hiearchy applies here.

        1. jason wright

          I don’t know what that is.



      1. Christina Cacioppo

        we really are all snowflakes.

        1. FAKE GRIMLOCK


    7. sigmaalgebra

      “All”, no.¬† A lot, and often a crucial part, yes.Now, how to do that?

  11. William Mougayar

    That’s a very astute analysis by Christina pertaining to P2P work distribution as a business opportunity. What are the business/revenue models behind this segment? Is it mostly transaction-based SaaS fee model or do network effects of value creation also come into play there?¬†

    1. Christina Cacioppo

      i might be a leeeeetle bit biased, but a network model seems the most natural fit with a p2p business to me.

      1. William Mougayar

        I’m biased like you…for a lot of good reasons ūüôā

  12. LE which is for this purpose:¬†“Commission magazine-quality writing from experienced journalists or bloggers.”¬†would be good to test out the suggestion I had the other day in a comment. That was to hire someone to summarize all the blog comments and create a 1 or 2 page overview that could be distributed to politicians/news organizations. If I remember correctly the post was regarding “Copyright Infringement vs Theft”.

  13. Andrew Hillman

    there is something to individuation, I think this will become more apparent in 2012. 

  14. Carl J. Mistlebauer

    What comes next?Well, I just got my Twylah page, and so I have something new to play with!  I only signed up because Fake Grimlock had mentioned Twylah so I went and checked it out.Now, I can add Twylah to all the other products that the AVC community has brought to my attention and to which I signed up for and have absolutely no clue what to do with!Does any of this have a purpose?



      1. Carl J. Mistlebauer

        Bacon Candy Canes? Oh, FG, that sounds so nasty! ūüôā I hope you realize that you are corrupting the youth of the world with these candy canes!But at least now I realize that I need to start doing more interesting things!That explains a lot! Move over Ann Landers the 21st century belongs to FAKE GRIMLOCK!

        1. Carl J. Mistlebauer

          Oh, and FAKE GRIMLOCK…my twylah page…has a section labelled “EGO” and there are two tweets about you!

  15. DanielHorowitz

    Why does the “Next Post” button go to the previous (chronologically) post? And, why does the main USV blog (where Christina’s) post is, not have any navigational buttons?¬†

    1. fredwilson

      because we assume you are reading in reverse chronological orderthat’s what the default behavior/layout is on blogsgreat point on we will fix that

      1. DanielHorowitz

        Hey Fred – Don’t want to derail the comments here, but I don’t believe this to be the default behavior. Do you have any links to resources describing this, or other blogs that follow the same convention. Every blog I’ve checked does the opposite of you.¬†Here is an example of where your UX is confusing. I open an old post and it’s part 1 of something and I want to read part 2, I need to go to the previous post instead of the next post? This is not logical.¬†Perhaps it’s better to have the title of the blog post rather than using previous or next.

        1. fredwilson

          i’m curious what others think?i prefer it this way but if the majority of AVC readers don’t, we’ll change it for sure

          1. LE

            Last = previous = olderNext = newer

          2. William Mougayar

            Next means previous, and previous means next. That makes a lot of sense. But AVCers aren’t conventional, so what do I know.

          3. Aaron Klein

            I’ve run into this a number of times and have become convinced those buttons should say “Older” and “Newer”…

          4. Rohan

            I like older and newer. ūüôā

  16. Mark Essel

    So who’s building Skynet?That’s who’s getting my dollar.

  17. Ela Madej

    Great post by Christina. What personally ¬†I love about it is that it’s yet another small validation that what me+team are working on makes sense (yes, we do have a great vision, we know it make sense… but in all those ups and downs we ned those validations, don’t we?).¬†This won’t add much to the discussion but made me smile!

  18. Peter Sullivan

    Props to Christina, did you guys finalize your new hire? 

    1. fredwilson

      nope. we are in the first round of f2f interviews.

      1. Peter Sullivan

        Too bad I was trying to change the world or else you’d have a kick ass candidate ūüôā

  19. andyidsinga

    It is a really well done post. (as in content, not bbq).

  20. Mordy Kaplinsky

    This discussion¬†hearkens¬†back to the old best of breed or full solution vendor. ¬†I love that individual companies can now create and deploy “niche/best of breed” products successfully, but at the same time it seems that very little thought has been put into how all these products work together.¬†Even when one looks at the Google AppStore there is limited cross-application capabilities.With the advent of this new environment I think it is prudent to figure out a unified framework that truly takes advantage of all these products combined capabilities.

  21. EMS1944

    Christina’s post was insightful, but the single scariest thought of the entire piece was about the commodification of labor. I think she’s utterly right about reeputation. I’m terrified she’s equally correct about the rest of the work force.I’m reminded that investments depend upon political stability. Political stability often depends upon the possibility of opportunity, of earning enough to maintain dignity and maintain hope.As I said, I’m terrified……..

  22. sigmaalgebra

    Ms. Cacioppo,Nice survey of many startups from several accelerators.While in political polling we want a ‘representative’ sample, in business, we want something quite exceptional.¬† And the future we want to predict is for these exceptional cases.¬† So, for finding the exceptional and predicting the future, a representative sample is not very useful.¬†

  23. jason wright

    …and you should have pulled out your Android and taken a photograph of the NDA…and posted it here.

  24. ellen

    I have read some of the comments in this post.¬† I really thought some of your readers were a bit more intelligent.¬† They blame the patients or the lawyers when that is not the case.¬†I recommend your readers take the time to read “Secret History of the War on Cancer”by Devra Davis and then they can also see that large industry has some culpablity in causing some major health problems.¬† Cancer treatment costs a whole lot of money of which big pharma and the health industry¬† adores.Your readers often talk about quality of care issues.¬† When a hospital discharges a patient before death, they are not counted as a mortality so how do we count their quality of care?I do believe doctors¬† should not get a yearly lump sum for treating a patuent because the patient will suffer.

  25. fredwilson

    Yes and no. I can take a picture in instagram, post it to foursquare and twitter and generate engagement in three network with one push of a button

  26. awaldstein

    Does the largest part of the market really evaluate them all? I think the consumer chooses quickly and changes slowly.That why a thoughtful approach to understanding and iterating your core value is imperative. The idea of what a launch is changing dramatically as a result.

  27. Sebastian Wain

    Probably in the future we will finally have agents that discover and use applications by themselves.But the problem again is about distribution. If the agents only discover applications or APIs on the top, only the offerings in the top ranks will win. Not the ones that fit your needs.We need better discovery engines to reduce the “long tail”. It’s very easy to say if you like “Star Wars” or not based on a list of known movies, but very difficult if the items are outliers.

  28. Christina Cacioppo

    the proliferation of photo services is its own interesting thing, i think. there’s so many out there, most of them launched in the past ~18 months, and yet nothing seems to have solidly bested all the rest. (instagram, of course, is closest; facebook’s dangerous when/if they get mobile together.)there’s so much competition amongst photo services right now though that it almost seems “worth it” to launch another if you’re hell-bent convinced you’ve got a killer many photo sharing services do we think will be around 18 months from now?

  29. fredwilson

    start somewhere and go from there.

  30. JamesHRH

    @fredwilson:disqus¬†@christinacacioppo:disqus¬†‘start somewhere and go from there’ is a world class early stage investor mantraThat’s the other side of the Chasm from the mainstream consumer who holds the mantra ‘where does everybody start from and I want to end up where they end up’.

  31. JamesHRH

    @christinacacioppo:disqus¬†tons. Its a nice to have, not a need to have. People don’t care enough to switch, because a late entrant cannot add enough value to a nice to have.The habit outweighs the benefit. Cripes, a lot of people still have Yahoo as their browser home page!

  32. Albert Hartman

    I think the best photo option would be much clearer if there was a company out there who clearly, really loved photos.

  33. awaldstein

    Depends on the reason for the share.Just to connect and create context…don’t know.Photo as fingerprints for verticals, certainly. With wine, seen some early demos of an app that reads the text on wine labels. Rather than clunky QR codes or a universal label database, they are reading the words and cross referencing them back.The photo is the share but the reader and database is the connecting data relevance.

  34. Aaron Klein

    and then we’ll have startups buying mass agents to “engage” with their product for the sole purpose of issuing growth statistic press releases to TechCrunch ūüėČ

  35. andyswan

    And if it is, we all will.

  36. Sebastian Wain

    There are several ways to prevent that. For example it’s very easy to measure where are you in the normal curve. If your retweets are mainstream retweets the agents can move to other information sources.

  37. andyswan

    It was a joke Charlie….keep in mind I am co-founder of an awesome community (now owned by TD Ameritrade). I get the power of (voluntary) community.A joke….much like OWS and this list of one-guy’s demands that read like a Chavez wet-dream.He forgot #12: Everybody gets a participant trophy and a pony!

  38. frabcus

    The genius of Heroku is its add-ons, which is basically an app store for sysadmin startups, where Heroku get to take a cut.