The Global Web Development Community

I got an email from my friend Robin Chan this morning. He's in Beijing and attended a hackathon sponsored by our portfolio company 10gen (developer of the open source datastore MongoDB). He said the developers in Beijing were showing their love for Mongo and building some great web apps with it.

That caused me to run a Google Trends analysis on MongoDB and this is what it shows (click on the image to see it in a larger format):

Mongo trends

The top regions for searches on MongoDB are Russia, South Korea, China, Ukraine, Sweden, Israel, India, and then the US. The top cities are Beijing (Robin was right), San Francisco, Moscow, Shanghai, Stockholm, and NYC. The top languages are Korean, Russian, Chinese, then English.

MongoDB is becoming a very popular datastore for new web and mobile applications. So this is a crude proxy for where the new web and mobile applications are being built.

When people ask me what the biggest trend on the web is, I tell them globalization. Web services are being used globally and they are increasingly being built globally. There is no place in the world that has a lock on talent, ideas, and execution. Web development is a global phenomenon.


Comments (Archived):

  1. Guest

    Very interesting proxy. Wondering when USV would open a office in Bangalore, Shanghai

    1. fredwilson

      we are not likely to do that. but those locations are ripe for more local VC investors.

    2. Panchabuta

      A lot them in India now, both Indian and US firms with an Indian arm. Though a lot of the team members are still financial  experts and few operational experts, but very few with actual entrepreneurial experience.My experience in working with serious entrepreneurs in India, is that it is much easier to raise money here in India

  2. William Mougayar

    Definitely,- global talent is diversifying and multiplying. Technology is a universal language, a common denominator that binds people together.We’ve recently outsourced some work to Australia, India and Lebanon, and the quality was comparable across the board.

    1. Luis Correa d'Almeida

      Interesting to think how far away we are from outsourcing ideas and concepts. Building software is – in many cases – trivial.

      1. Dave W Baldwin

        @wmoug:disqus Amen

    2. fredwilson

      Lebanon – yes!

      1. William Mougayar

        Got a chuckle when I read that the 13th Sony hacker was an 18-year old from Lebanon, a couple of weeks ago.…

  3. JimHirshfield

    More evidence that we need better immigration and visa policy here in the USA. Let’s welcome these innovators to our shores.

    1. JLM

      I agree that the USA needs a rigorous review and revision of its immigration policies and I have a personal bias for admitting folks to our country who are going to be able to create jobs.Having said that, this is also evidence that the world is not viewing geographical boundaries as business boundaries.  The magnitude of global Internet/software development is quite staggering.The Internet has created an entire new state of real estate.  You can be in Moscow and be in Austin, TX at the same time.Yesterday, I was finalizing a decision as to a piece of test development and administration software to use for testing employees who had completed on site, remote, refresher and recurrent training.I wanted to be able to design the tests ourselves and to be able to do everything over the Internet even if we had to host the application on our own servers.The finalist I was dealing with was located in Tombov, Russia, just 100 + miles southeast of Moscow.  I was talking to them on a chat line as clearly as if they were in Dallas.Point being, THEY did not need to be in the US to take care of me.I am just amazed by what is happening in the world right now.

      1. kenberger

        totally agree with both of you. And I also believe– because I keep personally seeing it– that the existence of competent talent offshore actually helps US firms well, rather than threaten them. See my other comment.Yes, I’m highly biased and self-serving in having such a view, of course.

      2. Frankie Sez

        “Yesterday, I was finalizing a decision as to a piece of test development and administration software to use for testing employees who had completed on site, remote, refresher and recurrent training.”Jesus, who are you, a character out of a Dilbert comic strip?

        1. Dilbert

          Great stuff.  Having a big Saturday night, are we?  

      3. fredwilson

        me too JLM

      4. markslater

        we have happily built Getabl in the Ukraine. Its the second company our guys have done for us, and we could not be happier.

  4. JimHirshfield

    As for mongodb…who knew? Congrats @dmerr

  5. WoosungAhn

    Totally agree with you about globalization as the biggest trend on the web. This is also true on the mobile:…I guess the only challenge from a business perspective is how to monetize better since both CPM and ARPU from direct payment are fairly low in international territories. That might be economic issues after all, if not talent or digital divide issues any more.

  6. Luis Correa d'Almeida

    Pure technology plays are found in many places around the globe. It is harder to find the combination of skills and experience that it takes to take products to market and drive adoption. While some areas of the globe (Eastern Europe, India, China) are known for their ability to innovate on the lower end of the stack, there are very few places  where we find a healthy combination of technical expertise with the skills needed to develop, take to market and scale a successful product.These ecosystems are unbalanced and fail to provide the social and economic infrastructure needed for innovation to leave the basement.The broader question is how do we equip those that are not ‘lucky’ enough to have access to marketing and productization expertise, capital, a sound legal framework, or local markets that will adopt their product.

  7. Aviah Laor

    This is one of the US biggest economic mistakes. While manufacturing declines, the US has gained a lot of experience in scaling and handling large consumer markets via the web.Still, many many sites (including Amazon for a lot of products) are willing to supply only “to a US address”.This is a huge loss. America can easily become a global hub in the physical supply chain for nearly every market.

    1. fredwilson

      sometimes this is not the developers fault. content licenses can often be the issue

  8. kenberger

    Agree with you x1000 on the globalization trend. Our dev labs are seizing this with a vengeance. Our main shop remains in Vietnam, our newer shop thrives in Bangalore (albeit competitive there these days), we might work soon with resources in Argentina. We’ve yet to see a US or European client complain about the distance, and they welcome our diversity.Some are quick to complain and point to immigration issues or our jobs being sent offshore– we are actually seeing that the existence of competent remote augmenting services *create* jobs here in the US. In a tight skilled labor market, these services enable essential engineering work to get done, allowing innovative but cash- and talent-constrained startups to grow unhindered.

  9. Semil Shah

    Fred, have you heard of “Appsterdam”? Mike Lee (@bmf), formerly of Apple and a co-founder of Tapulous, left Silicon Valley after the sale to Disney and is organizing an iOS developer community in Amsterdam. Check it out:

    1. fredwilson

      that is awesome

  10. Douglas Crets

    I agree that web development is global, but I would rather be in China doing this than in America. Much more scale, and much more freedom to experiment.

    1. fredwilson

      more freedom to experiment?

      1. Douglas Crets


  11. sigmaalgebra

    “When people ask me what the biggest trend on the web is, I tell them globalization.  Web services are being used globally and they are increasingly being built globally.  There is no place in the world that has a lock on talent, ideas, and execution.  Web development is a global phenomenon.”But at least two recent threads on this blog provided some serious numerical evidence that US information technology venture capital is not doing very well on, say, exits over $100 million or on IPOs.So, yes,”Web services are being used globally and they are increasingly being built globally.”And, essentially, venture investing in such things is not very promising financially.”There is no place in the world that has a lock on talent, ideas, and execution.”Yes, there is:  May I have the sealed envelope, please?  Yup, here it is:  The US has a lock on the really powerful, valuable talent for new ideas and execution that can yield good investment opportunities.Is this point widely understood?  Nope!  Necessarily how to build a company worth over $100 million and an IPO cannot be widely understood!”Web development is a global phenomenon.”Yup, but that “Web development” is a shot in the dark for good ROI.The information technology promising for good ROI is not in ‘Web development’ such as is ‘globalized’.Net, the ‘Web development’ part is routine but low in value.The big bucks are elsewhere.For a start, we should set aside “Ideas are easy, and plentiful”:  Yes, ideas that make venture capital returns so low are.  Good ideas, however, are difficult and rare.How to find and evaluate good ideas?  There are ways, and parts of the US dominate the world. 

  12. Dave W Baldwin

    @sigmaalgebra:disqus What you left out is the fact a product does not have to do the 100 million to deliver 100x earnings.  And the number of products able to do that will grow.To everyone talking about their places around the globe, good job… and IMHO, the combo of Brazil and China works in triangulation with the US…..What I started with will be a big enabler for everyone since the economics of production to retail matched against rising wages (yes that is happening) elsewhere builds the bigger marketplace.Last, to go back to the lingering concern over ideas… it is a matter of smarts and creativity, knowing when the pieces fall into place and be ahead of the game.  Some know that.  As the creativity side takes hold globally, the smart and creative idea that fits into a better picture will increase in number. Do not let the big guys tell you what is in your mind is worthless because that simply leads to the same problem as allowing the government decide what is worthwhile and who gets all the cash…Like this for instance, one of the reason some of us work with kids rather than bitch about Education: These Car Prototypes Built By Kids Can Do Close To 2,000 Miles Per Gallon

    1. sigmaalgebra

      “What you left out is the fact a product does not have to do the 100 million to deliver 100x earnings.  And the number of products able to do that will grow.”I might not understand your “100x earnings”.  Maybe you meant that a product does not have to lead to a company worth $100 million to generate comparatively high earnings.  If so, then sure.But venture capital wants companies that can ‘exit’ for $100 million or more; they believe that exits of $50 million or $20 million, terrific for a sole proprietor with a Sub-chapter S, won’t ‘move the needle’ or give good ROI (return on investment) to the limited partners, and the arithmetic for this point is solid.So, how to find and evaluate ‘ideas’ for a company that can exit for $100 million?Over the past 10 years, on average US venture capital has not been very good with this question.  Further, and I omitted, with the two recent threads on this blog that venture capital returns have been low, some of the data came from M. Suster who also had a graph showing the numbers of venture capital (VC) general partners (GPs) year by year for the past 10 years or so, and the numbers fell off worst than housing prices.  The ‘unemployment rate’ for VC GPs is much worse than the 9%, 15%, 20% or whatever for the general population.  We’ve got to be talking a lot of Ivy League Bachelor’s and Harvard or Stanford MBAs out looking for real work!So, this thread mentioned Web development outside the US.  Okay.  Maybe the suggestion was that there would be VC investment opportunities there.Well, maybe.  If get lucky, sure.  If some country is really behind in Web development and has a long way to go and make an investment early on even in something fairly routine, say, copying some success — Facebook, LinkedIn, FourSquare — in the US and easy enough for good Web developers in another country to copy, then maybe.  Maybe in some country there is something special in the culture or economy that makes something routine in Web development suddenly very valuable?  Okay.But more generally, I have a tough time believing that there are good chances of good ROI investing in anything that is close to routine Web development.So, there may be good Web developers around the world, but this is also true for bicycle repair people, goat milkers, shoe makers, welders, cast iron fabricators, etc., and there’s not much chance of high ROI from investing in any of those.That is, now Web development is just too routine.  Sometimes tedious?  Yes.  Sometimes involving some ‘graphic arts’ talent?  Okay.  Frustrating due to security concerns or scaling problems?  Sure.  But special enough to lead to $100 million exits?  A long shot.With the low ROI numbers, somehow US venture capital is going to feel a really big, swift kick in the back side as more GPs leave the field, LPs (limited partners) quit sending money, etc.Yes, I’m talking VC industry averages here since it is accepted that Union Square has had very high ROI per dollar invested.This stuff of US VC on average not making money is a really big bummer for the LPs, GPs, entrepreneurs, US economy, US more generally, and even civilization, especially as the VC field has been heavily ‘pruned’ of its ‘poor performers’.  A bummer.And from 50,000 feet, the low ROI numbers for information technology venture capital make no sense:  Processor cycles, main memory bytes, disk storage bytes, Internet bandwidth, infrastructure software, etc. are all factors of 10 better in price and factors of 10 better in performance, for factors of some thousands in price/ performance, compared with, say, the ‘golden age’ of venture capital in the 1990s or whenever that was or when Arthur Rock was investing, etc.  There was a time when one of Jim Clark’s Silicon Graphics workstations would really set one back, but now can get more in, what, a $20 video adapter card?Intel is saying that they can put 1000 cores on a chip and have them work together effectively.  Last I heard, could get 96 GB of main memory in a server 1 U (1.75″) high — UNbelievable.  HP has a server with 32 cores that can accept 2 TB of main memory — beyond belief!  Microsoft’s .NET?  It’s enormous.  And then the Internet as a resource for more in information technology:  Have a question about .NET, SQL Server, IPv6, etc.?  Then Google it!  So, we should be seeing information technology feeding its own growth for exponential growth and spectacular ROI.  But we’re not seeing the ROI.  Also UNbelievable.My first diagnosis is that the attitude “Ideas are easy and plentiful” is much of the problem.  Instead, it is just crucial to be able to work effectively with new ideas and, in particular, have a quite good estimate early on of the financial value of proceeding with those ideas with just routine execution.  Venture capital doesn’t believe this.  Bummer.It appears that in the days of Arthur Rock, etc., it was common to fund a project described ‘on the back of a napkin’.  If so, then what was crucial were just the ‘ideas’, for the market and the target customers, the problem to be solved, the solution to the problem, the marketing and ‘channels’, any technology for the solution, any advantageous proprietary technology, any joint deals, etc.Now (in information technology, here and below) the solid theme is “Ideas are easy and plentiful”.  So, basically VCs essentially just ignore any and all ‘ideas’.  For a substitute, the main criteria are to play with the product and estimate how well people will like it and ‘traction’ and its rate of growth.These criteria, the associated data, and a little spreadsheet work can be okay for estimating when the project might become profitable.  Okay.  Now we’re talking business success on the order of one or a few profitable pizza shops.But the criteria and the spreadsheet are still poor evidence for the coveted $100 million exit and, net, bluntly, just looking at the ideas, as might have been done even before the first line of software was written, has to be much better.With these criteria, no wonder the number of $100 million exits are so rare and VC ROI is so low.  Who expected something else?I smell the putrid oder of a dead mouse behind the refrigerator or naive, scared, salaried MBAs at the larger LPs.  It looks like the LPs wrote the criteria, maybe after the Dot Com bubble burst.Here’s an obvious reaction of the entrepreneurs:  All across the US, pizza shops get funded, started, and profitable, all routinely, thank you, without any attention at all from VC.  Well, starting a serious Web site, with the software written by the unpaid founder, takes much less pre-revenue and pre-earnings capital for equipment than a pizza shop.  E.g., one of those pizza ovens has got to cost more than a nice 4 core server with a 2.5 GHz clock assembled from parts from, say, Tiger Direct.  Then for the pizza shop, there’s the plumbing, HVAC. fire issues, rest room, flooring, electric power, natural gas, etc.  — really adds up.  Could put together one heck of a little server farm in a spare bedroom with an AC unit in the window for what a pizza shop costs.Then, here’s an advantage of a Web site over a pizza shop:  If the project is good and the site starts to become popular and if the site gets ads from, say, Google or some ad network and has some good user ‘demographics’, then ballpark the site should be able to get revenue of $1 per 1000 ads displayed.  Put, say, 3 ads per page.  Say a page is 100 KB and can be sent as 1 Mb. Say that the upload bandwidth is 15 Mbps and half fill that 24 x 7 for 7.5 Web pages a second.  Then get1 * 3 * 7.5 * 3600 * 24 * 30 / 1000 = 58,320dollars a month in revenue.  So, one guy, the founder writing the software, a few servers in a spare bedroom, a willing ISP, and the project is much more profitable than any one pizza shop.So, I have to guess that many owners of good Web development projects, by the time they meet the VC funding criteria of ‘traction’, etc., won’t be much interested in equity funding, a C corporation instead of a Subchapter S, lots of time and money with lawyers, a four year vesting schedule for just part of what they already own 100% of, a board that doesn’t understand their business and can fire him, etc.For an example, there’s Plenty of Fish, at one time one guy, two old Dell servers, ads just via Google, $10 million in annual revenue, and no VC funding.  Heck, in a few months he has more in pre-tax earnings than a relatively large Series A. I have to believe that if a VC firm called the founder and offered equity, then the founder might say, “Sorry, Sir, I don’t know if you do that or drink it, but I’m not having any.”.For how many more such examples there are, I don’t know.But if there are many such examples, even ones with just a paltry, little, trivial $1 million in annual revenue, for one guy, with servers in a spare bedroom, then what the VCs are getting to fund are basically the much less good projects.In this case, VC begins to look like small scale private equity — be the buyer of last resort for projects nearly in trouble.To get a $100 million exit, which fundamentally on average across all entrepreneurs can’t be easy, making a decision early on has to involve seeing the future quite accurately, and, thus, it has to be just crucial to pay close attention to the crucial information which is overwhelmingly just the ideas.  Sorry ’bout that.The Little Red Hen in Mother Goose could see the future this way.  So can owners of pizza shops.Net, the criteria the VCs are using just are not promising for selecting projects that will both take equity funding and also yield a $100 million exit.If LPs are the problem, then some LPs need to listen up before they completely kill the goose that can give them their coveted golden eggs.I’ve done a lot in technology projects and helped start one of the world’s most successful venture funded startups, and when I hear the criteria of the ‘early stage’ VCs, I do a ROFL.  Hopeless.  No wonder the ROI sucks. 

      1. alphacalculus

        “I might not understand…”Yup. Well said.

      2. Dave W Baldwin

        I wrote that fast, really on the run this weekend… should of phrased beginning with the 10x moving over to higher.The whole thing regarding ideas is knowing what is coming.  Yeah, you can have Twitter and FB, but it isn’t so much me too in the US, it is a matter of similar versions appear in other countries, i.e. China.You make really good points… from my side, I see where thoughts are drive up costs, get the big funding and then force the 10x+.  Problem is, to have those high costs brings in danger on 2 fronts- platform makeup sucks and the marketing will be subpar compounded with shelf life.Good post.

        1. sigmaalgebra

          I can’t address any of my thoughts to Fred:  Apparently he’s making a LOT of money, and that’s certainly the goal, and I can’t argue with it or hope to improve on it.Maybe Fred just got lucky, but I don’t believe it.  For me, yesterday was to go to Home Depot in Wappingers Falls and get some monofilament line for my Weed Eater trimmer to begin to clean up my overgrown yard, and across the street at the McDonald’s just inside the door I saw on the floor a big poster from FourSquare with something about ‘check-in often enough and win some free French fries’ or some such.  NICE.  Congratulations to Fred, Union Square, and the FourSquare team.Just how Fred understood the FourSquare opportunity from a few foils and a demo beats me.  NICE work.Here’s one reason I would have missed the opportunity:  For each customer at Sam’s Club, Sam’s has astoundingly good information including the time and date of each checkout and the full list of everything bought.  FourSquare can help retail merchants track repeat customers, but Sam’s has that data and much more.  Yet, I haven’t seen Sam’s make any use of all the data they have on me.  Similarly for customers at A&P who have an A&P card, at TruValue Hardware for customers with a TruValue card, at any merchant for any customers who usually pay with a credit card, etc.  Still I haven’t seen much exploitation of such data.  So that McDonald’s wants to work with FourSquare is another case of Fred seeing something not obvious.  Nice.For ‘ideas’ for a new business, I take a view that is somewhat limited but at one time was nearly standard:  Find a problem that is causing a lot of ‘pain’ and get a solution that is a ‘must have’ instead of just a ‘nice to have’.  Then for a “moat” (as Buffett mentioned recently) or to create a high ‘technological barrier to entry’ and an ‘unfair advantage’, build on some of the best, appropriate stuff from the best US research universities and add on some original work that, net, gets an especially good solution, a better solution than the other guys have much chance of duplicating or equaling.This approach to the ‘solution’ has been standard in technology for decades.  Indeed, in simple terms, the reason Congress so generously funds the US research universities, lets the US have nearly all the world’s best research universities, the world’s research ‘dream team’, is to provide the foundation for better ‘solutions’ to problems in US national security and, now with NIH, health care.  Congress is correct here.So, when I claimed here, in different words, that “Web development” such as Fred was describing around the world was “routine” and not a promising basis for the coveted $100 million exits and that the US is the world’s unique source for the right stuff for such exits, I meant the stuff from the research universities.”Unique”?  It took a large collection of the world’s brightest, hardest working people many years to build the best stuff on the shelves of the research libraries, and, thus, there’s no way anyone without careful study at such a university will reinvent anything important on those library shelves.  Doesn’t, won’t, can’t happen.  Sounds like a nice business advantage to me!So, my recipe for getting $100 million exits and beating the pants, loin cloths, robes, burkas, copies of US blue jeans, or whatever off the rest of the world is to do the same as US national security:  Start with the best of the US research universities and do some original work.Protect the work?  Lock it up.  For the code, run it on the servers only.  Indeed, mostly don’t even tell people about the crucial, core ‘secret sauce’; instead, just say it’s just ‘software’.I would go farther:  Such work has to be the key to getting US average information technology venture funding returns way up.  No amount of cleverness in the stone age will equal routine work in the iron age.Yes, US biomedical technology believes that research has a role.  That US information technology doesn’t believe the same about research is a bummer, a huge mistake, a self-inflicted, serious wound in the gut.  No wonder the returns are low.In particular, this starting with research and doing some more yields only ‘ideas’.  So, the attitude that “Ideas are easy and plentiful” shows US information technology venture capital shooting itself in the gut.  Similarly, what appears to be Fred’s admiration for Web development skills around the world strains over routine gnats and forgets the very special elephants of research where the US already blows away the rest of the world.So, the US research universities are blowing away the rest of the world, and US information technology venture capital, with low returns and an astoundingly high general partner ‘unemployment rate’, insists on ignoring research.  Hmm.I can hear the questions now:  “Can research be relevant to information technology and exploiting the Internet?”.  The one word answer is simple, “Yes”.  For two words, “Hell yes.”.”Where are the examples?”.  Well, there haven’t been enough $100 million exits!  Also we’re talking an opportunity, that is, in entrepreneurship, largely unplowed ground.  But there are examples:  A. Viterbi did, what, QUALCOMM?  This has a role in, say, ‘mobile’?  Somewhere in the UV optics in a new $2 billion Intel plant that can get line widths down to 22 nm (about 200 atoms) has to be some examples.Then we come to what you mentioned, Facebook and Twitter.  Yup, they have made money for their investors, and that’s the main goal.  Good.  That they are ‘social’, maybe heavily for gossip, is fine with me.With their success, apparently they do solve a ‘big problem’ and, for many millions of people, are a ‘must have’.  I’m surprised but learning something.  I was assuming that a ‘must have’ had to have closer connection with things more directly in the economy.But I would note, as you did, that Facebook and Twitter might be relatively easy to copy by people around the world with the good Web design skills Fred described.  So, in the US the “moat” is ‘communities of engaged users’ but that moat does not always work well for US companies outside the US. 

          1. Dave W Baldwin

            True… just remember the end run customers and maturity gained using tech.  FB and Twitter have shelf lives and in the end what wins is what fits the customer’s wishes. That is not meant as a slam on either company.As far as the big groupings of research, that will spill over too.  It already has in some ways and will increase.Never think I’m belittling programmers, for I’m not.  We have good stuff going from several sources as mentioned… just sometimes folks forget I’m the Evangelist…Hang in there, for you are right.  @fredwilson:disqus , never think I’m slamming, because you know if I was, I’d say so.Otherwise, gotta go. Have been helping with search of a missing lady (since June 1) and am pretty wiped out.

  13. Fred Stevens-Smith

    Totally. It’s just a damn shame that the States remains pretty much the only place in the world where investors ‘get it’ in terms of pathbreaking new business models / untested concepts.

    1. David Chu

      I think your opinion here may be short sighted.The three attributes of American culture that I think have lead to its success is less fear of failure, a direct speaking language and greater emphasis on empathy.  These things are all changing and I would not be surprised if within the next 10 years a huge multi-national Google-like company did not come out of Asia.  Things are changing fast and people out here in Asia.

      1. Douglas Crets

        The only reason a “huge multi-national Google like company” has not come outof China is because of the language issue. In fact, more than one have comeout of china, but typically Americans don’t know it, because they don’tspeaka tha language. Look at Baidu and several of the other web companies,and even Facebook, LinkedIn and Amazon / Ebay like challengers. There aremany Chinese language speaking nations in Asia, and even nations outside ofAsia that speak languages other than English, which the Chinese engage withnormally in trade relations.I think where we fail in America is in our ability to see what has alreadyhappened — China has cracked the trade code.I will repeat the most profound advice anyone has ever given me, afterspending a good portion of my formative business years in Hong Kong, India,Shanghai and Beijing.Ready? “LEARN CHINESE!” Said with effect. Said with insistence. Said aseven a reprimand, as it was said to me.

  14. David Semeria

    What exactly does “globalization” mean in this context?Are we talking about the location of startups / development / users ?Whatever the case, that wouldn’t be news as far as I’m concerned.For me globalization is users interacting with each other regardless of geography.Twitter got the ball rolling, but realized that music was cross cultural.This trend is still in its infancy – and has really got legs…

    1. fredwilson

      turntable had to turn off all users outside the US today for licensing issuesthe one part of the internet that isn’t global is the premium content part

      1. David Semeria

        That’s the great advantage of UGC – no licensing required.Turntable (or someone else) should repeat the model but only allow content from artists who have released the rights to the platform.A bit like a cross-cultural MySpace done properly. 

        1. fredwilson

          that’s how soundcloud works

    2. Robert Thuston

      Good question. “users interacting with each other regardless of geography” is a good definition.  It seems like a good “end” to begin with.  Music and art may be the starting point for this.

      1. David Semeria

        Indeed.Turntable really blew my mind when I found myself in a “Japanese” room and they all started writing in English so I could understand & interact.I see a big educational angle here. Using music as the “glue”, kids could teach each other languages, for example.

  15. John Hinnegan

    Sorry, Fred, but I completely disagree that “this is a crude proxy for where the new web and mobile applications are being built.”  From Google Trends:… “he data is scaled based on the average search traffic of the term you’ve entered. “  That means that the stats by number are scaled relative to the average amount of traffic occurring in the region.  So 10 searches in China for Mongo with 1 million searches will appear to be 10x the searches as 10 searches in China for Mongo with 10 million searches.  It’s still just 10 queries.The ranking is a percent of total queries — so good for Israel (pop 7 mil) and Sweden (pop 9 mil), but they’re just too small on the scale we’re talking about not to be discarded for reasons of sample size.  (Even if they really are that much more technical, you can just attribute the fact that the US is much more economically/industrially diverse.)Now consider that Google has relatively poor market share in the countries topping your list, Russia (25%), South Korea (7%), and China (15%), each of which has their own dominant home-grown player.  Add to that the language bias — English is a second language for all 3 of those.  I now go out on a limb and say that Google will search English sites better than and have a more complete index than Yandex, Naver, or Baidu — it’s just not relevant to their core constituents’ needs.  (The implicit assumption here is that most technical websites, docs, and avenues for community support are in English. Having worked as an engineer in Japan for a year, though anecdotal, I believe that to be true.)Now you’ve got the case in which you’ve got developers self-selecting themselves into a minority search engine — Google.  They might only use Google for their work or technical stuff, leaving their normal (home-life) searches on the dominant local player.  Now you have the case where technical terms dominate the overall search volume — because it’s all relative.To picture this more clearly, lets suppose that developers discovered that Bing indexed tech blogs, stackoverflow, and email list records better than Google.  So by this time next year, all the software engineers had started using Bing over Google when looking for answers to their MongoDB questions.  What conclusions would we draw about the technical innovation of the US if we used only Bing’s relative search volume?  Technical terms like Mongo would dominate because the industry in general had self-selected into the group, and proceeded to dominate the search terms in relative weight.I think you’re going to have to do a lot more research before you can start drawing any conclusions.

    1. ShanaC

      This is probably the best analysis of how to use google trends I’ve seen in a while, thanks!

    2. Robert Thuston

      Maybe a crude “crude proxy” (considering your figures in paragraph 4) … but it doesn’t negate the value of the discussion.  

      1. John Hinnegan


    3. fredwilson

      as i said, “a crude proxy”

      1. Mark Essel

        Think crude is not quite right.If the numbers are normalized they’re only good for looking at local trends.May want to update the original post, saw this in my feed and thought these numbers don’t look right. No way, we have less Mongo hackers than 3-4 other nations, some far smaller than the US.Imagine a crude proxy for traction metrics like this. Hey we went from 4 to 6 users, we grew by 50% and are bigger than twitter. You’d quickly jump in to resolve this for any portfolio startup that claimed this.

    4. leigh

      so shouldn’t have looked at this on a sunday morning.  now you have made my head hurt. 🙂

  16. Mike Su


  17. Guest

    Spot on Fred. Our company ( has been built from the ground-up using global collaboration (based on relationships my co-founder and I made during our times living abroad). It works really well for us because it is based on strong existing relationships (for example, our head of technology is one of my best friends in India, where he lives due to Visa reasons). I don’t think our approach would work without such strong existing relationships though.I am amazed by the Web talent I’ve seen while living in India and South America (Colombia is a hub of amazing designers). We complain about not being able to find and hire capable devs and designers here, yet there’s literally an entire world of them being under-utilized. I’ve seen it first-hand. But it is also an incredible opportunity for globally-minded companies.

  18. awaldstein

    To me the big change is that socialization is now possible on a global real-time scale. This is not a technological revolution, but a social and behavioral one.That flattening is a ramp for cultures to meet and communicate over cross cultural, uni-behavioral drives like music or sw development or celebrity obsessions or just passion points like wine or art or even, in my work, marketing.The power for communities and businesses is not for cultural niches but for global populated communities around more specific points on the interest graph with huge possible memberships that span the cultural and geographical divide.

    1. fredwilson

      yeah, that’s what david was saying too

  19. Wells Baum

    “Think global,” learned that from an important boss.  I also learned it in foreign affairs school.  Anyone interested in starting a global music fund? 

  20. Noam Makavy

    As a founder of a web-based start-up in Israel, I wish this was true. Israelis tend to read all about bleeding-edge web development technologies, then develop using .NET as their military pedigree has accustomed them.Fortunately it doesn’t take long to train good developers in Ruby on Rails.

    1. fredwilson

      my friend saul was telling me that one of the challenges in israel is to get the best developers to start using non microsoft technologies

  21. nikiscevak

    Fred, completely agree but I’d love to know how this thread of conversation jibes with some earlier points you’ve made around ‘outsourcing’. With many of the most talented developers in the world not in the US, wouldn’t it make sense to have a globally distributed team, even if the company was based in the US?Especially considering if the market for startups all has the same view and all tries to hire the same developers in Palo Alto, wouldn’t simple investment theory suggest that those engineers are being dramatically over valued and that hiring people overseas in cities like Shanghai (or even rural cities) will be dramatically undervalued?

  22. KrishS

    Love this post on Globalization. I just moved to NYC after spending 12+ yrs across Asia/US. Based on my experience talking to various companies here, here’s my take:(a) most digital/mobile companies here don’t understand emerging market at a deep level to make significant inroads (i know there are exceptions). The audience demographics and behavior is different. What works in NYC or California doesn’t always work in India or Indonesia. Why ? God knows ! That’s the way it plays out.  (b) When they do staff international executive roles out of US HQ to hammer out an “international strategy”, it’s important to hire talent that has deep local in-country experience. I don’t see that happening a lot.The speed and agility with which Asian companies move today is far greater than what it used to be pre-Internet. US companies will need to start looking at global markets with a different lens if you really want to dominate those markets.

  23. Chopper

    Developing something doesn’t make it globalized unless they are selling the result in that country. My guess is it’s being developed there and sold back to western markets. Or even developed by western companies in the first place, which is just technology being used by outsourced labor run by national/international companies. Either way, I’d hardly say something going on for nearly 20 years (even 10 years for web services alone) is a trend…It’s been here longer nearly than google and will be many years after google.

  24. Alejandro Marin

    Hello, what is special about MongoDB, what differentiates it?

    1. fredwilson

      simple to setup, easy to scale, less complexity, more power

  25. jmcaddell

    On my site for the past month Italy is #2 in readers, after the US. UK is #3. Per capita, Italian readership is 3x US readership. It’s amazing to me how content transcends borders, how global platforms allow people everywhere to discover interesting things.I’m trying to figure out how to write a post in Italian to honor those readers!regards, John

  26. paramendra

    “There is no place in the world that has a lock on talent, ideas, and execution. Web development is a global phenomenon.” Tell me about it. 

  27. ellen

    I am older and adore  I love the chat rooms and adored the international aspect ofit as well as the domestic socializing. There was so much excitement at my house from  thechildren  when turntable allowed “sir mix a lot” to dj a room. I hope that turntable will be able togo global again soon.

  28. Cheap android tablet

    I can’t agree with you any more. 

  29. Roert Brookes

    It is really an amazing information and gives the lot of knowledge about web development community. It is the actual thing that i was looking for. It helps me a lot.Joomla Developer Kent