Jun 15

Tim O'Reilly

Tim O'Reilly

Why the Future is in South Korea

Business 2.0 has an interesting article arguing that the future is in South Korea. There's so much food for thought in this article that I had trouble deciding whether to make one long entry about it or a half dozen separate ones. (I've chosen the former.) Here's the premise of the article:

South Korea has become the world's best laboratory for broadband services - and a place to look to for answers on how the Internet business may evolve....Cyworld, for example, is a social network owned by a subsidiary of SK Telecom, the country's largest wireless provider. To an American eye, the Cyworld service looks like a mixture of some of the hottest US properties: it's MySpace meets Flickr and Blogger and AIM and Second Life.

Users have avatars that visit and can link to each other's "minihompy" - a miniature homepage that's actually a 3-D room containing a users' blog, photos, and virtual items for sale. Cyworld's digital garage sales include music, ringtones, clothes for your avatar and furnishings for your own minihompy.

Cyworld has penetration rates that would make Rupert Murdoch, CEO of MySpace parent News Corp. (Research), green with envy: An astonishing 90 percent of South Koreans in their 20s use the service. Celebrities and politicians set up their own minihompies, and the way to get ahead in twentysomething Korean society is to found a popular Cyworld club, or chat room.

This story came to my attention via a posting by Dewayne Hendricks on Dave Farber's IP List. There's also a lot of interesting followup on the list. For example...

Bart Stuck wrote:

"In 1990 Korea Telecom initiated a network study, KT 2000, to take their network to world class in the year 2000. I was engaged as a consultant, And suggested the following: South Korea has 80% of its population in Seoul and five other cities, so deploy optical fiber rings into each city, And trunk them together to provide high speed digital transport for the chaebol. Once that is done, the rings can sprout new optical fiber tentacles: 1) for cellular telephony, 2)for CATV, and for 3)high speed data communications via xDSL and CATV modems (no World Wide Web in 1990). Each of these Rollouts had pro forma income statements, with a revenue model and an expense model (including equipment depreciation, operations, marketing and sales) And each of these rollouts had to PAY FOR THEMSELVES. The study was completed in mid 1991, the management of KT said thank you (HQ has copies Of the final study, four inches thick, in Hangol), and the rest was history. The bulk of the investment went into thenetwork in 1992-1995, and at the very end, deploying PCs was a relatively small amount of total CAPEX."

Bruce Kushnik then wrote:

"The exact same type of study was done by Deloitte & Touche for the Bell companies --- though instead of phone based it was 'state-based'. Opportunity New Jersey, Pennsylvania, Illinois, Indiana and Ohio all used the same $1 million cookie-cutter study... also with revenue and cost models and all showing how fiber optic services would dramatically change America's economic growth, --- Tele-Learning, Tele-conferencing, Tele-anything you want.

This same type of study was also done by other Bell groups ----

The primary difference -- once the laws were changed in favor of the Bell companies, they took the money and ran... Virtually no PUC ever held the companies accountable for the extra cash and tax breaks. Pennsylvania's PUC tried in 2000 but was shot down, and the Ratepayer Advocate of NJ wrote a
scathing review in 1997 --- didn't matter. (I believe Indiana is the only state that got real concessions for some of the problems.)

Well documented, ---According to the New Jersey state order, signed in 1993, that granted the phone companies large sums of customers' money for upgrading the state with fiber, by 2010 the ENTIRE state is supposed to have been upgraded with 45mbps services, capable of high definition video, as well as open to competitors and ubiquitously deployed.

Here's the actual info from the New Jersey Order, created in 1993. This "Order", including this timeline and commitments is still on the books.

Here's California's timeline for 5.5 million homes by 2000

We wouldn't be 16th in Broadband had the states enforced the laws and the FCC analysis/reports actually examined the funding issues. (it left out every state commitment in their reports for Section 706.

Here's a summary of the FCC's data problems. [And here's my article] $200 billion broadband scandal.

That's a sobering discussion of the failures of corporate responsibility in the U.S. Even today, the telcos are far too reluctant to bet on the future.

The discussion on the list also addressed the issue of connectivity to the outside world. The article had mentioned that the most successful Korean services were all homegrown. Udhay Shankar wrote: "This might also have something to do with the fact that (when I was in Seoul last) accessing the outside internet, while adequately fast, is nowhere near the multi-megabit speeds you can get from servers located in Korea. Trying to access a streaming multimedia service located outside Korea (yes, I tried) was not a very responsive experience."

Brett Glass replied: "This may have something to do with the fact that Korea is among the #1 sources of spam. One of the unintended negative consequences of providing ubiquitous, extremely high speed access to nearly an entire country is that others actually have to throttle or block traffic from those networks to prevent abuse."

John Levine expanded on South Korea's bad net-behavior:

"South Korea has made itself into a thoroughly unpleasant net neighbor. When they wired up the country, they gave no apparent thought to security. You could tell each time a new school came online because it had a server with the same unpatched version of Windows which was taken over by worms in about 15 minutes and started spewing spam and worse. They compounded the damage with an ill-considered spam law that made spam legal so long as the subject line had the Korean equivalent of ADV:, so the blast of illegally relayed foreign spam was soon joined by an equal blast of home-grown spam. Complaining about the spam was useless for a combination of reasons: the nominal managers of most of the computers were completely untrained, they usually had no language in common with the recipients of the spam their networks were sending, and the sheer volume of spam was so huge that even the few competent ones were overwhelmed.

Several years ago I got so tired of Korean spam that I got the list of IP ranges assigned to Korea and set up a DNSBL for my own use called Despite having no publicity other than word of mouth, it's now used by mail admins around the world and fields hundreds of queries a second. I keep statistics, and will remove networks on request if it's not seeing spam from them, but I have only
removed a few small industrial and education networks and the big ones, most notably the national phone company's Kornet just blast away. Now and then someone writes and says "how dare you block our whole network, it's a gross overreaction." I write back and say "what would you do if a network were sending a thousand spams for every real message? That's what I did."

The government changed the spam law to be reasonable, and I have talked to people from the Korean government who are trying to deal with their network security problems, but they have an impossible task. I gather that most people in Korea consider e-mail to be useless, and if they use it at all, they discard an address after a month or two because of the spam load. I don't follow other security issues as closely, but I see a lot of Korean addresses wherever trouble arises.

So it's true, wiring up the entire country that fast was a technical tour de force, but before you wish for your own country to do the same thing, be sure you understand what you're asking for."

(On a contrary note, though, Slashdot just pointed to a Register article noting that 64% of spam comes from Taiwan. The US is #2 with 24%, and China is #3 with 3%. Korea wasn't even on the list. So this looks like a good time to remember the biblical admonition against looking for the mote in your neighbor's eye....)

Back to the positive enthusiasm of the original article, here are some of the other things that caught my attention:

  • "Cyworld is a license to print money. The service itself is free (and available on cellphones as well as online), but to buy all the extras - like ringtones and virtual furnishings - will cost you "acorns," the service's virtual currency. Cyworld sells its users $300,000 in acorns every single day." Do the math: a free service, with a commerce marketplace owned by the vendor that generates well over $100 million a year in revenue. In Korea, with a population one sixth the size of the US. Not too shabby.

  • "Users have avatars that visit and can link to each other's "minihompy" - a miniature homepage that's actually a 3-D room containing a users' blog, photos, and virtual items for sale." Love that neologism, "minihompy"! But what I really found interesting here was the idea that the 3D web has reached the mass market not through wide-open worlds like Second Life, but by a constrained model that creates the 3D equivalent of MySpace.

  • "Cyworld is expanding fast. It launched in China and Japan last year, and a US launch is slated for later this year.... SK Telecom's ace in the hole is its experience with running a social network on mobile devices." It has always seemed to me that social networking and the phone are made for each other. Current social networks require you to recreate your communications network. But applications like telephony, IM, and email, tools for communication between your real social network, are ripe for reinvention in the Web 2.0 world. The question is whether the phone companies have the genetic makeup to let go of their walled gardens. Cyworld may work in Korea because SK Telecom holds all the cards, much as AOL did before the introduction of the open world wide web. What I really want to see is this kind of thing built on the open web, with 3D minihompy servers that can be put up by anyone as easily as they can put up a normal web site, and with open interconnectivity to social networking on phones, via interoperable IM networks, or any other mechanism that people use to communicate with each other.

Lots of food for thought.

tags: web 2.0  | comments: 21   | Sphere It

Previous  |  Next

0 TrackBacks

TrackBack URL for this entry:

Comments: 21

  Kevin Farnham [06.15.06 08:58 AM]

If anything, the momentum in social networking seems to be increasing, especially when you look at the entire world. See, for example, this post:

With economic transactions involving purchase of virtual goods using virtual currency that is exchanged for "real-world" currency, it almost looks like the social networking virtual world is becoming another facet of "real world" life.

Email was great for business, but it remained just a tool. Social networking, with its provision for creation of roles and characterizations of oneself, is an entirely different realm.

Social networking worlds are "places" where people relax (just as they might do by reading fiction), but they are also places where people "play" and interact socially. Economic transactions imply a future where social networks will also be places where people work -- they'll become integrated into certain non-virtual workplaces before long, I think...

  Brian Aker [06.15.06 06:21 PM]


I was in Korea last week and a local showed me Cyworld. The article gets it right... it is very targeted for Korea and its very much the thing. It was the site people brought up again and again to me.


  Gen Kanai [06.15.06 07:43 PM]

Many, many companies/people confuse popularity of services or products in Asia with similar popularity elsewhere.

Vodafone failed in Japan for various reasons, chief among them not respecting the differences in culture.

NTT's i-Mode service is only now beginning to take hold in the UK, but is hardly a "success."

Japanese console games, popular in many markets, are not popular in S. Korea for historical reasons.

Cyworld is a very Korean service, something that is tied very closely to Korean culture. Cyworld is trying to do the same in Japan and elsewhere in Asia, but has yet to succeed.

My point is that it is never as easy as it may seem and that all too often cultural differences are overlooked because they cannot be easily controlled or accounted for.

There are lessons to learn from other markets, but it has been my experience that they are not what they might seem.

  CK [06.15.06 09:06 PM]

Another area Korea has been fairly strong: Wireless services. It will be interesting to see how successful Wibro and HSDPA will turn out to be.

  Molly Webb [06.16.06 01:44 AM]

I agree with Gen Kanai, and posted about CyWorld recently ( after spending a month in Korea looking at science and innovation.

What makes social networking successful is getting people to continue using it, constantly coming back and feeding the interactions. There are different ways of compelling this in different places, because of how our offline social and working lifestyles are arranged.

CyWorld seems to follow theoretical models of mobile phone usage - you interact online with people you see everyday and your close relationships get closer. Will it take off in the US? We've seen friendster give way to MySpace already, will CyWorld be next?

  Johnny Won [06.16.06 07:39 AM]

The most amazing thing about the article is what it leaves out, it doesn't talk about all the other significant milestones that Korean people are doing day to day. This includes and is not limited to the prominence of online gaming, communication networks based entirely w/o wires, and endless bandwidth. Before ABC had their shows on streaming, Korean soap operas were being sold for 25 cents for 5 years.

Also, let this be a record to show that CyWorld is going to be a massive hit in the US. There will be imitators, in 18 months but us.cyworld is going to be huge because its not a webpage but a true web application. Ease of use will be incredibly simple and people will start using it in droves while the blogger community gets in a tiff about the lack of standards and openness. For the people who think blogging is hard (a lot of people), cyworld will make everything gloriously easier.

  TumbleOne [06.16.06 07:59 AM]

One aspect of the Korean internet that was not mentioned was the fact that most Korean sites require your Korean citizen card number to sign up for an account. From news sites to web based game sites they all require this ID number to get access to video streams or to play games.
Even Cyworld requires your citizen card number. There is a Foreigner sign up page on Cyworld where you don't require a number but I haven't gotten an account and I've been trying on and off for over a year. The error message says "The service is full. Try again later." or something to that effect.

  Brian Williams [06.16.06 08:28 AM]

My wife is Korean, so I have my own Cyworld Minihompy. It's rather annoying to use, since it only works 100% in IE, and you have to disable popup blockers,; all the content is in the popups.

  Some Joe [06.16.06 10:42 AM]

So they have bandwidth. Historically it has always happened that a scarcity of a resource inspires more invention around that resource than glut. Consider the technology development of tropical civilizations. Compare it with that of civilizations that experienced serious winter scarcities, i.e., Europe. The scarcity forced Europeans to innovate to stay alive. Why would anyone in Korea want to innovate in, say, AJAX when even the slowest CGI program loads in less than 2 seconds? If anyone will capitalize on bandwidth achievers like Singapore and S.Korea, it will be the world's biggest bandwidth consumers, the international mass media conglomerate. Media companies (the real ones, not their dead-trees spinoffs everyone assumes to be out of the loop) will coopt existing technology groups in those nations and make a fat buck off them right there. And most likely it will happen before anyone here in the US knows about it. If it hasn't already happened, that is. So go on wishing that you had a bandwidth glut without the foggiest idea what you'd do with it 99% of the time. All Korea is using it for right now is sending trojans everywhere. The article about Korea being advanced and most spam coming from Taiwan is correct, because Korea delivers trojans via Windows Messenger Service. I've done enough baremetal restores to know all about Korea's advancement, with their IE-only websites and such. You'll find the real innovators in the burned out industrial districts and backwoods of your hometown, making things happen with 56k.

  Search Engines WEB [06.16.06 01:21 PM]

Thanks for this story - gave it a digg a few hours ago - it struck a chord, it made page one

  xbird [06.16.06 03:02 PM]

I think the premise that bandwidth limits innovation is frivolous to be honest; AJAX is not used because it simply saves bandwidth, I would not call any 2.0 website 56k friendly. AJAX takes advantage of high bandwidth by offering a dynamic experience rather than the drab, static, and flat design of a website optimized for dialup. You fail to mention many Korean social internet innovations, the digg concept is far from new and existed in some form in Korea for years, and it is mainstream by all definitions. 56k only makes the internet unappealing and unused.

The fact that you mention that 99% of bandwidth would be wasted also shows that our lack of bandwidth stifles our innovation and expectations. Most foreigners in Korea are overwhelmed with possibilities for streaming on demand TV shows in HD, nearly 100% HSDPA high bandwidth coverage (pathetic that portable cell based internet in Korea offers more bandwidth than the average US broadband customer), and the benefits of not having to bend down for the lowest denominator. Why else is our HD broadcasting so limited?

  Jinsoo An [06.16.06 05:30 PM]

I've been tracking Cyworld for many years now as it grew in Korean community, as well as Japan, China, and Taiwan. In fact, I was waiting for Cyworld to make the move to US when Friendster and MySpace were beginning to grab public attention. And they've finally done it!! According to Cyworld US beta testers, Cyworld has quietly let go of its IE-friendly ActiveX, and has decided to implement AJAX instead. In terms of social complexity and features, Cyworld is a definite winner against MySpace. The truth is that MySpace will lose the appeal from users as they�ve became the late-comer for the next stage of user interaction - meaning that "I've made all these online friends, shared something in common, and posted hundreds of comments. What can I do with them now?" Besides, MySpace community has grown so much in the recent years that in some cases, users believe MySpace has gone too mainstream and it's becoming a short-lived fad for teenagers. A large number of these users are turning to other similar services hoping to experience another level of social interaction or find smaller niche groups. Many Web 2.0 analysts also predict that users will become more demanding and want more sophisticated user interactions as Web 2.0 start-ups are receiving huge amount of attention for their innovative ideas. Unless MySpace offers new service or experience that keeps the users involved in the community, services like Cyworld may well take over its place.

Meanwhile, Cyworld US also has to resolve its own set of issues before it can compete against MySpace. First of all, they need a better understanding of the American demographics. Cyworld US beta has received many complains about its visual � for being too childish. This may require the designers to redo the visual so that it will be more accepted by broader range of age groups, instead of just kids. Unfortunately, Cyworld probably won't be undergoing major face lift as it destroys their highly appraised global brand appeal and recognition.

Here are some facts about Cyworld.

In Korea, Cyworld currently has approximately 15 million users who generate 90,000 hits per day. Currently, there are 47.5 million people in Korea. If you do the math, that's about 31.6% of population using the service. Three of the largest target age groups are 15-20, 20-30, and 30-35, and they make up about 80% of the community. MySpace currently has about 80 million users from US and outside of US. In addition there are approximately 295.7 million people in US. Out of 80 million users, approximately 20% of users are from outside of US, mostly from UK, Canada, Australia, France, and Singapore. So realistically, there are about 21.6% of US population or 64 million Americans who use MySpace.

Cyworld generates large amounts of profit from dotori online currency and ads. A typical cost of a large banner placement costs $30,000 for a specified time (which I do not have the information for). In addition, they offer mobile services through SK Telecom and Samsung/LG/Pantech phones which have Cyworld application pre-installed. In conjunction with NateOn IM service, they were able to central SMS, MMS, online community, e-mail, and video/photo upload.

Guess what Helio (venture of SK Telecom and Earthlink) is doing in America? Obviously, they�re hoping to take advantage of the same strategy that has been proven to work in Korea. They're currently offering high-end mobile phones (from Pantech of Korea) with multimedia contents and built-in MySpace. However, bringing Cyworld into the US changes the whole picture. In fact, it's matter of time when Cyworld US starts offering mobile contents to its users Helio. Operating a social network with power of Internet and mobile network could definitely give Cyworld a definite advantage over MySpace. Now there are two important questions that need to be answered before anyone can mock at SK Telecom for its multimillion dollar investment on Sprint�s 3G network: "How long will the exclusive contract between Helio and MySpace last?" and �When�s the launch date for Cyworld?�

  framore [06.16.06 06:11 PM]

I m Italian and I live in Seoul and I have a cyworld minihompy. I m writing a thesis about this particular community to publish for my graduation in Italy.
It is really original way to think the web.

  Duncan gough [06.17.06 08:15 AM]

Interestingly, a lot of the growth in south Korea has been driven by Casual Games and MMOs, specifically, the 'item model' which has many similarities to the 'freemium' services we talk about in the West.

More here, as I've been following this kind of thing for a while now:

  Justin Mason [06.19.06 02:59 AM]

See also the related "selca" phenomenon, which knocks Flickr into a cocked hat:

  Tomi Itkonen [06.21.06 07:57 AM]

After a quick glance, Habbo Hotel is in many ways similar to Cyworld. My kids (girl 9 and boy 11) used to rush to the local kiosk to convert their weekly pocket-money to Habbo Hotel credits. Then they bought virtual pets and furniture to their rooms.

The next fever was Runescape. Large portions of the game are free - membership costs and opens several levels and other features.

Now they are playing pool and tic-tac-toe -type simple classic games, which run as Java applets. The opponent is some other guy/gal on the net. They buy some additional features (game statistics, avatar modifications, ...) with SMS messages.

Some things to notice:
They tend to change their attention quite quickly from one game to another. This is quite natural, they are kids after all.
Most of the games are not advertised anywhere, they spread largely from word of mouth.
The girl plays all the games also. In this case and age, the gender is not a dividing factor.
E.g. in Runescape, there's lot of global buying and selling going on. I think it is quite good practise for their future jobs. They get to learn some tricks and laws of the international trade... ;)

  Ellie JY [06.26.06 07:02 PM]

I have my own cyworld hompy, and my mother also has it too, through which she drops a line almost everyday(other than that, she doesn't use it though). Cyworld became life for many Koreas.

I happened to get interviewed by BBC, which came to Korea for special survey on technological advances in Korea. I stood up in front of blue screen and acted as if me and a BBC reporter ARE in my virtual avatar world. It was a great fun.

People do the cyworld because it is, by far, the easiest way for young Koreans to stay in touch with friends. :-)

  SJ [06.28.06 11:02 AM]

It's very interesting to read articles about Cyworld, and I'm quite surprised at the depth of comments and postings. I guess I'll have to talk to SK Telecom and Cyworld officials and what's the key issues, especially concerning their latest moves to expand its service overseas.

  marble2 [09.18.06 09:31 PM]

I believe that the future is in ultraband, not limited to South Korea. Speeds greater than 30Mbps and you start to think about a massive next wave revolution in internet usage. What I want to know is how fast is the connection and how much it will cost me, so that I can get on a plane and move to the fastest connection in the world.

I've been following CyWorld for awhile and agree with the viewpoint that it's both a cultural communication difference coupled with a leading company on a new type of social structure driven by this new ultraband experience. We're at the very front end of rich social networking experiences in other markets and the CyWorld, South Korea model is the tip of the iceberg.

  Anonymous [01.31.07 10:09 AM]

I thought Radar left South Korea when his Uncle Ed died.

  Bro [03.03.09 10:24 PM]

I agree with "Some Joe". if you have more than 56k then you cannot innovate. End of story.

Some Joe's logic is lucid

Post A Comment:

 (please be patient, comments may take awhile to post)

Type the characters you see in the picture above.