Author / Clay Shirky

by Clay Shirky

The WTO and the Seattle Protests

How the Seattle protests may end up strengthening the WTO.

First published in FEED, 11/99.

What do the United Auto Workers, the Sierra Club, and nameless vandals in black ski masks have in common? Not much, actually, which is why it’s surprising to see them all together protesting the current round of World Trade Organization talks in Seattle. The amassed protestors represent a broad spectrum of views: The WTO should be given the power to enforce labor issues; it should have its power to lower tariffs stripped away; it should be made more transparent; it should be abolished. The central fact currently uniting these incongruent positions is the WTO’s secrecy and lack of accountability. And yet, if the protestors manage to force the WTO to become more accountable to the citizens of the countries it represents, two things will happen: The WTO will become more a part of the emerging world government, not less; and the differences between the protestors who want to improve the WTO and the protestors who
want to destroy it will break out into the open. Both of these things would strengthen, rather than weaken, the WTO.

The WTO has consistently maintained that it is not part of a world government, but that claim is just an expedient fiction. Cross-border trade is a classic governmental issue, and an organization with the power to set and enforce rules on world trade is a de facto governing body. Like the European Union and the International Olympic Committee, both of which endured crises this year occaisioned by their back-room dealings, the WTO has a crisis of legitimacy on its hands. It began life as GATT, the
General Agreement on Trade and Tariffs, with a mandate to set rules but no enforcement powers. (Unsurprisingly, laws with no cops didn’t make for much of a system.) In response, the WTO was formed, giving it both the ability to make rules and some ability to enforce them, though without any real transparency. This may be an efficient way to proceed, but of course efficiency without oversight is a recipe for pure self-interest
to carry the day. The challenge now presented in Seattle is to open the process to input from the people whose lives will be affected.

But democratizing the WTO will not provide all the protestors what they want, because they want such different things. Reformers only make common cause with revolutionaries when things are so bad that the differences between their position are blurred, and if the WTO addresses the protestors’ demands, it will quickly become clear that the central difference between the groups in Seattle is between those who accept that globalization is not going away and those who don’t. Globalization is not a program being advanced by a cabal but a by-product of changing costs. When shipping costs fall in a manufacturing economy, international trade rises because the number of suppliers is less bounded by geography; when communications costs fall in an information economy,
international trade rises even faster because geography becomes even less important. The WTO did not start this process, it does not control it, and it could not end it even if it wanted to.

International trade can only be shaped — not stopped — and arguing about what that shape should be is a task of Solomonic complexity. The protestors who want to improve the lives of Honduran workers making sneakers bound for the US are not likely to find common cause with labor unions who don’t want US sneakers (or steel, or cars) made outside the US at all. Environmental groups might not like increased popular input
into the WTO’s decisions, because most of the people affected by the WTO care more about raising wages than drowning turtles. The people talking of abolishing the WTO will be least happy of all, since reform and compromise are the enemies of revolution. The WTO is at a crossroads: If it does not open up, the rising tide of protest will strip it of its legitimacy and perhaps derail the current Millennium Round. However, if they don’t panic in the face of public expression of real grievances, and if they
find some way to give the protestors a seat at the table, the WTO will enter the Millennium Round as a strengthened institution rather than a weakened one.

by Clay Shirky

Internet World ’00: A View From the Floor

What the ’00 Internet World says about the state of the industry.

First published in SAR, 11/99.

Just got back from Internet World 2K in New York. I have gone to every Internet World in New York since 1993, and my approach is always the same – skip the conference and the keynotes and go straight to the trade floor, and then walk every aisle and look (briefly) at every booth. Today was several hundred booths, and took 7 hours, because the show not only filled the Javits end-to-end, but spilled over into a temporary structure erected north of the Javits to take the overflow.

I do this because it doesn’t matter to me when one company thinks they have a good idea, but when I see three companies with the same idea, then I start to take notice. As such, this write-up isn’t about individual companies so much as about industry themes.

In addition, the floor of Internet World is a good proxy for the marketplace: if a company can’t make its business proposition clear to the person passing by their booth, they aren’t going to be able to make it clear in the market either. Most memorable tag line this year: “Web Experience Management Solutions for Organizations that must deliver successful web interactions.” Oh good, my boss was just asking for that. Another goodie: “It begins with ‘e'” (for a company called eon). That probably tested great with the focus groups, but to my eye it looks like a commentary on how predictable naming strategies for internet businesses have become. Herewith my impressions of the IW2K zeitgeist:

What is Internet World for?

In years past, Internet World has had an identity crisis, as a tradeshow for everyone from router engineers to media buyers to end users. That is over now — INterent World, at least in NY, is now a tradeshow for companies that provides goods and services to companies using the internet for their business.

There were only a handful of booths for sites for end users — Atomfilms, MyFamily.com — and they looked completely out of place amidst the content syndicators, web site audit tools, multi-lingual translation services and HR services. Internet World has become a strictly b2b affair.

The New New Thing…

…didn’t show up. I didn’t see anything really new at the show, for either or hardware or software. There were lots of new products and services, of course, and lots of new companies, but no new ideas and no new categories.

Internet World is now a trade show for a maturing industry, where the emphasis is not on creating new kinds of functions but in making improvements on existing ones — CRM, ASP, content management, hosting, et al.

The only really new thing at IW was Groove, Ray Ozzie’s audacious attempt at living up to the dream of the two-way web, and Ray was holding court in meeting rooms off to the side — Groove wasn’t even on the show floor.

I lost count of the number of times companies told me they were the leading this or the leading that in their space. No one, in 7 hours and many conversations, ever told me that their offering was unique, or that they had no competition — everyone wanted to be part of a well-defined sector, and then to be the leading company in that sector. I detect novelty fatigue, and a sense that its safer to be in a validated business sector than to be pioneering a new one.

No big theme jumped out this year either. In other years, there’s been Java, or push, or some other bandwagon people were eager to jump on or distance themselves from. This year though, the only broad theme I could detect was fairly broad, and which I would call “Anxiety Alleviation”.

Gone are the days when convicing businesses to go online in the first place was the mission of IW: to a first approximation, every business is now using the net. Many of the exhibitors in this show were linked by trying to provide these businesses not great new opportunities, but trying instead to solve particularly thorny problems, like –

“How can I manage all this content?”

Content management is back as a category. As the home grown solutions groan under the weight of constant updates, and more businesses have to move to dynamic content in order to be able to update the site without breaking it, the competition for web publishing/content management systems is heating up again, with Allaire introducing its first offering in this space, and Vignette, the clear leader in high-end web publishing, trying to outpace the competition by re-positioning itself as an ‘ebusiness solution’, instead of merely being a content tool.

– “How can I let my employees publish to the web while bypassing those recalcitrant sysadmins?”

Lots of competitors to Microsoft’s FrontPage as well, but designed specifically for business environments. These products are more workflow systems than full content management solutions, and are more positioned around the idea that a companies Web site can’t be allowed to suffer from the bottleneck of an organizations IT group. The general solution for this class of products is to allow certain employees the ability to change web content, and then to give them tools to let them update pages without requiring them to know HTML.

– “How can I use the Web for groups of employees?”

Many offerings for intranet/groupware, designed to “enable collaborative workgroups” or “capture group intelligence.” The days when internet use was a largely personal affair seem to be ending, at least for businesses, as concern for creating shared Web environments grows. (The aforementioned groove.net is in this space as well.)

Another concern here was in “knowledge management”, companies offering ways to capture the collective wisdom of a sales force or work group.

– “How can I talk to/interact with my customers?”

Lots of companies offering to answer this question. Of all the “Anxiety Alleviators” of the show, this was the category with the most obvious anxiety around it: why are so few of the visitors to my site buying anything? There were lots of related solutions here — CRM, internet call centers, real time chat with customers onsite — and it is all targetted at raising the the number of people who buy and number of people who return to ecommerce web sites.

– “How can I get them this Zirconium ring they just bought?”

There were several companies offering to help with shipping and tracking, from a start-up that is taking a page from Kozmo and building package dropoffs in grocery stores, to UPS, who is trying to make up for the ground it lost in letting FedEx become the ecommerce shipper of choice for several high profile brands.

– “How can I monitor my customer base?”

The anxiety in the executive suites about how little they know about how customers are using their web site was also in full flower. WebTrends was there, of course, but most of the offerings weren’t stand-alone log file analysis but full-on suties of “ebusiness intelligence monitoring” — customer experience audits, site performance analysis, data mining. All these businesses share an idea that by their behaviors, customers are trying to tell you something, and if you listen, you will be able to improve their experinece and your bottom line.

– “How can I get a kicker on ecommerce revenues?”

I was amused to see several companies offering customized wrapping paper, logo printing, and other ways to generate an extra $1 on ecommerce transactions. Look for a lot of this sort of novelty/one off upsell this Christmas.

– “What if I get sued?”

There were several insurance companies, a category I haven’t even been aware of in previous Internet Worlds. How’s that for the net going mainstream?

– “How can I make my site more responsive?”

Many products or services designed to speed up the responsiveness of a web site, by caching, streaming, or optimizing the way it is served. Akamai has clearly raised the standards for web site responsiveness, and in light of the slow roll-out of broadband, there are a number of companies rushing into the breach.

– “How can I keep my servers running?”

As uptime becomes a mainstream business issue, it is driving growth in the number of companies offering remote monitoring, testing, and reporting on network health and web site responsiveness/uptime.

If they pay for you, they own you.

The only big surprise to me this year was the explosion of partner pavilions. I have always been suspicious of partner pavilions, since the real beneficiary is the sponsor, often at the expense of the nominal partner: “BEHOLD THE MAJESTY THAT IS ORACLE!!! (and here’s our partner, MumbleCo….)”.

The real effect of the Partner Pavilions is chest-beating. Companies that sell services or solutions (Oracle, MS, Real) use the size of their Partner Pavilion to signal how wide industry adoption of their offering is. As a result, I found it impossible to take any of the actual partenrs in the booths seriously — they are the corporate equivalent of booth babes. The bigger message, however, is that the giants we have now are largely going to be the giants we have in the future, and for smaller companies, their affilaitions with these giants are going to become increasingly critical.

Even odder are the country pavilions, the “Made In Israel/Germany/ Sweden” booths with companies from their respective countries. These companies are likewise impossible to take seriously, because the message behind this supposed show of national pride is “Not Good Enough.”

Not good enough to make it to Internet World on their own, that is, a suspicion heightened when you see how many companies on the show floor have dual headquarters in the US and Tel Aviv, or Berlin, or Hong Kong without needing any help from their respective governments.

If a country really wants to spend its tax dollars helping local businesses join the internet economy, it should buy 3 or 4 separate spots on the show floor and give them to the companies, without the self-defeating “Made in Sweden” label.

The power of the hosting providers

Hosting providers, those unsexy businesses that build network centers and buy bandwidth, are suddenly the industry’s new power brokers. Globix, Global Center, and Genuity all had mammoth booths.

Because hosting providers have long term relationships with their clients, unlike ad agencies and consultants, and because they get a monthly check, unlike product vendors, and because they are the only third party who has access behind the firewall, they have emereged as the ideal broker of additinal services to companies that already use them.

The global internet and the local internet

Geography has arrived as a major force. Language translation is everywhere as a service, as the days of the purely American-centric internet wane.

There are also a surprising number of services offering to aggregate content by local area, from a mapping company, to a search engine that maps URLs to geography to a company that aggregates news/weather/etc and lets you pick content modules for your website, tailored to your area. The web is increasingly being mapped, literally and figuratively, to the real world.

Content syndication

Many companies are offering content syndication tools, where they provide some sort of dynamic service — news, calendaring, “browser buddy” navigation tools — which you license for your website.

Along with the content management tools and template-driven publishing tools, the push towards syndicated contetn is all part of the general anxiety over managing a dynamic and constnatly updated web site without having to become a publishing company.

Mobile

Mobile has arrived, of course. The bad news for WAP is that everyone is selling “Mobile Solutions”, of which WAP is presented as a sub-set. WAP had hoped to become synonymous with the mobile internet, and that didn’t happen.

In particular, Palm is emerging as the alternate platform of choice, a move further dramatized by the side-by-side positioning of the packed Palm pavilion and the relatively sparsely attended WAP pavilion. Because Palm is a real OS, several companies are offering Palm browsers (YadaYada) or Palm servers (Thin Air) to extend Palm’s functions as a mobile device.

Talk is back

Voice over the intenet is back in another incarnation. Last year, there was lots of internet telephony — that was all but gone this year, to be replaced by several companies offering talk as a service to increase customer retention and loyalty, whether through sales agents, pre-recorded messages, or help functions.

VR/3D is back, with a vengeance.

Another hardy perennial is the idea of 3D shopping, sometimes with a 3D avatar, sometimes just to look at the product.

This has always seemed to me like the Videophone of the Web, an idea that is always being pushed by technologists but never adopted by users. In keeping with this year’s theme of Anxiety Alleviation, however, 3D is no longer being sold as a way to make a cool web site, but as just the ticket for closing sales and encouraging customer loyalty.

I’m not holding my breath.

Ontology is back, and so are search engines

After a few years in which AI and agents were largely ignored as impractical, and in which a few major portals had the search space locked up, specialized search engines, especially ones armed with “inference engines” (sets of rules about which words mean what and are related to which other words) or geared to smaller than Web-wide searches (only industry veritcals, or corporate intranets, or in geographic areas. Google proved that as the amount of data on the Web continues to explode, new search solutions can find a market.

The dogs that didn’t bark in the night.

– Digital Signatures. There were few companies trying to take advantage of the new Digitial Signature Act. By next year, expect services offering electronically enforceable signatures to be legion.

– E-books. There was only one standalone “ebook” product I saw. This is probably because ebooks don’t (yet) have any b2b applications.

– Peer-to-peer. Softwax was the only peer-to-peer company there, and no one was touting their peer-to-peer solutions. By next year this meme will either be ubiquitous or gone.

– Linux. Only one company specifically touting Linux. When it comes to operating systems, the concerns represented at the show were much more about uptime, remote management, and reporting, rather than being about which operating system to use.

– Napster, and the Entertainment Industry. Not a peep, except Atom Film and one lonely NAPTE booth in the North PAvilion. The entertainment industry’s issues with the internet are being aired elsewhere.

– Likewise, there were very few companies offering employee monitoring/copyright compliance software to companies, or Digital Rights Management services to producers of content.

Other random notes.

– Two years ago, ‘IP’ at Internet World stopped meaning ‘Internet Protocol’ and started meaning ‘Intellectual Property.’ This year, ‘POP’ stopped meaning ‘Post Office Protocol’ and started meaning ‘Point of Presence’ marketing.

– Lots of job recruitment sites and outsourced business solutions for tech employees this year. Despite the layoffs in recent months, the need for skilled labor is still strong.

– Magazines are a trailing indicator of a business sector. There were lots of magazines there, presumably hoping to validate themselves as good places to do all that spending on ads dot coms do. Even the NY Times got into the act, publishing a gratuitious “Ecommerce’ section on the first day of the trade show, featuring not a single article that wouldn’t have been at home in the ‘Circuits’ section.

Expect this to contract this year. As a trailing indicator, its will soon be the magazine world’s turn to get big, get niche, or get out.

– Associations are a trailing indicator as well. There were lots of associations you could join this year, whch blur in my mind into a “Worldwide Consortium of Affiliated Web Managers and Marketers Society”. This presumably is occaisioned by the desire to keep a new crop of upstarts from displacing us, the old crop of upstarts.

– AOL metamorphosis into an internet company is complete. The AOL booth, complete with partner pavilion, and the ads for AOL6.0 and AOL anywhere, put it at the center of the action.

– XML is no longer a service, now its a feature. Last year there were several companies offering to help you with your XML needs. This year, XML is simply a check-off on many products and services.

Like Java, XML won’t live up to the hype, but like Java, once its been widely enough adopted, it will find its niche and occupy it.

– The ASP model is now the norm. I had several conversations with companies selling standalone corporate software products, but almost all of them assured me they had an ASP alternative in the works. This is related to the new power of the hosting providers, who have the leverage to hasten adoption of ASP services.

– I did see two services that took bits of old models to make a new thing. The first was Navixo, which is taking the “Yahoo Browser Bar” idea, but allowing companies to populate it differently when users come to their sites. Its a way, in other words, to add real pull-down menus to site navigation, an idea that never really caught on with the launch of DHTML.

The other was a company called Enkia, which uses its own database to build “Find more like this…” features for sites using an ASP model, rather than requiring those sites to install local software on their database.

– The personal computer was nowhere to be seen. PCs are now as taken for granted as routers, and while Compaq and IBM were there, they weren’t showing PCs, while the pure PC manufacturers like Toshiba and Gateway weren’t there at all.

– Ominously for Apple, the iMac, last year’s booth acoutrement of choice, was also nowhere to be seen. – In general, this show was the most characterless Internet World I’ve been to. You would never know, walking the trade floor, that this industry had ever been considered threatening to corporate America, since by now it largely is corporate America. I saw one head of purple hair, but otherwise there was almost no one there who would be unwelcome in the halls of General Electric.

As if to underline the point, as I was leaving I ran into Ugly George, the Flying Dutchman of New York’s Internet Industry, lugging his albatross of a web cam with him, and not only did he seem bizarrely out of place, even his identity had been corporatized, as his access pass read “George Ugly”. Glad thats over. See you next year. -clay

by Clay Shirky

The Future of Europe Lies In Email

What a cyber-cafe on steroids says about European Internet adoption.

First published in SAR, 10/99.

The future of post-national Europe is standing in line in London. A new
shop has opened opposite London’s Victoria rail Station, with a line out
the door from noon to night, selling the most addicitve product on
earth: connection. The shop, a cybercafe on steroids called
easyEverything, takes the traditional approach to cybercafes (“Like a
real cafe, but with terminals”) and turns it on its head (“Like a
college computer lab, but with latte.”) In return for providing
travellers with hundreds of terminals and cheap, accessible internet
access, easyEverything has been rewarded with an unexpectedly huge flow
of the under-25 set travelling across Europe, and these travellers
are all using easyEverything for one thing: HotMail. Thats it – almost
no one is surfing, buying, listenting, or any of the other things the
Web is being hyped for, just thousands of people sending tens of
thousands of emails a day, 24/7. The messages are short — ‘Here’s my
phone number, here’s my itinerary’ — and the turnover of customers is
high, but the message being sent out via these digital jungle drums is a
‘Map of Cool’, a traveller’s atlas of Europe. The subtext of the
messages is “London is good right now. Glasgow is good right now.
Frankfurt is lame, but Prague is cool…”, on and on, a vast interactive
‘Let’s Go’ produced in real time, and with far more accuracy than the
commercial travel guides. This may seem like nothing more than a
holiday pastime, but this kind of information moves markets, and is the
key to transforming the European Union from a free trade zone into a
real country.

The EU is the test case for the effects of the Internet on government.
No other multi-national region of the world has gone so far to
dismantle national broders. Within the EU there are no passport
checks, no customs checks at internal broders, and no barriers to work

  • any citizen of any of the 12 EU countries can work in any other EU
    country without needing a visa. Things that Americans take for
    granted, like being able to move 3000 miles for a job, are available
    to the citizens of the EU for the first time. In other words, the EU
    has most of the trappings of a country except the citizens, and the
    citizens are being produced at places like easyEverything. The people
    sending their email there are Europe’s first post-national generation,
    its first Internet generation, the first group of people who can move
    from one country to another if they hear that life is better
    elsewhere. The willingness of this generation to ignore national
    identity is going to confound their elders, the people who have grown
    up convinced that sentiments like ‘The Germans are efficient and
    humorless, while the Italians are undisciplined and fun-loving’ have
    an almost genetic component. Nationality matters less than economics
  • the Internet generation is going to behave more like customers than
    citizens.

There used to be a story in the automotive industry that went this way:
in the 1980s, American car manufacturers made expensive gas guzzlers,
while the Japanese made zippy little bundles of automotive efficency.
Lesson? Japan good, America bad. Then in the 1990s, the Japanese got
sloppy and smug, while the Americans re-tooled and re-engineered, and
the advantage switched to Detroit. New lesson? America good, Japan bad,
the opposite of the old lesson. This is the official history, anyway,
but there is another, simpler interpretation, one that doesn’t require
belief in a pair of opposite lessons in subsequent decades. That
interpretation is this: people buy the cars they like best, and don’t
care about nationality.

The reason that this simpler interpretation is not the official
history is that a global market is terrifying to people, or at least
to people who make cars for a living. Detriot’s “Buy American”
campaign was proof that producers thought that nationality could – and
should – sway consumer choice more than quality and price, but those
days are ending. Patriotism is the last refuge of an unprofitable
business. We are so used to seeing markets through the filter of
national borders that its hard to see what the emerging global market
is going to look like, but the behavior of the Internet generation in
Europe will show us its outlines.

The current cohort of European under 25s have several important
characteristics that make them immune to cheap nationalism — they are
the first generation whose parents didn’t live through WWII, and they
are richer, more mobile, and speak better English (the official second
language of the 21st century) than any generation in history. Add to
this that they are comfortable with the internet, that they can work
anywhere they like, and that places like easyEverything are springing up
to satisfy untapped demand for communications, and you get a generation
of rootless cosmopolitans, people who are unimpressed with arguments
that they should tolerate unemployment, or high prices, or limited
horizons, simply in order to defend national characteristics that boil
down to little more than a preference for different kinds of cheese.

This emerging European generation will run into a great deal of
hostility from the status quo –a continent that has erupted into war
twice in this century will give up its borders uneasily and unevenly
— but the transformation is inevitable. ‘National culture’ is just
another way of saying ‘arbitrage opportunity’. The internet has
created the biggest referendum on the ability of government to satisfy
its citizens in the history of the world, and governments that fail
the test will see their citizens vote with their feet. As the world is
increasingly divided into EU-like free trade arrangements — NAFTA,
OSEAN, OAU, MERCOSUR, CARICOM — the free flow of information about
jobs and prices across those borders will accelerate the transfer of
power to supra-national entities, and the loyalty of citizens will go
with it. This period has only just begun, and it will have a long
time to go before the post-national generation is in power, but you
can start to see what the world will look like in a 25 years while
standing in line at Victoria station.

by Clay Shirky

Culture is Just Another Word for “Arbitrage Opportunity”

The threat cross-border ecommerce presents to national culture.

First published in FEED, 07/99.

Britain’s Tory party’s use of ecommerce as an argument for retaining the pound is a new high-water mark in the internet’s influence on politics. Britain has been sitting on the fence for years about whether to scrap the pound in favor of the euro, and there has always been a significant block of British “euro-skeptics” who fear that economic convergence equals loss of sovereignty and national character. The Tories would love
to embrace these euro-skeptics (and their votes) by pandering to their sense that a strong pound stands for a self-sufficient Britain, but have been wary of doing so because rejecting the euro looks like a vote for economic weakness. With the rise of ecommerce, the Tories now think they can have it both ways — by mixing “new economy” optimism
with flag-waving they can embrace economic expansion while defending their cultural traditions against international dilution, a position likely to resonate well with the euro-skeptics. They are certainly right that a borderless economy could help preserve the pound, but if their aim is to defend Britain’s cultural identity, they will rue the day they ever heard the word ecommerce.

National character is driven by economic barriers. What a citizen eats, reads, drives, watches, and wears is shaped simply by what’s available, and what’s available is shaped by borders. Tariffs and customs sharply restrict the movement of both goods and services, so countries differ from one another in part because each population has different
limits on their choices in the marketplace. Any cross-border commerce — McDonald’s, Louis Vuitton, Jackie Chan movies — breaks down these cultural limits on choice to a degree, but ecommerce makes national borders so permeable that now it isn’t even necessary to open a shop in another country to start doing business with its citizens.

In an age of falling geographic barriers, culture is just another word for “arbitrage opportunity.” Until recently, geographic borders protected local businesses from serious international competition, but cross-border commerce is changing all that. Britain is currently seeing this in the form of a price war in consumer goods, precipitated by Wal-Mart’s entry into a market that has never seen a low-cost retailer before. The euro-
skeptics have grasped that joining the euro will expose Britain to far more of Wal-Mart’s style of ultra-efficient price-driven competition, but they have not grasped that keeping the pound and embracing ecommerce is no solution. Differing currencies are no longer much of a barrier — with online currency converters, foreign ecommerce companies offering cheaper goods to British citizens can switch prices from euros to pounds
instantaneously.

The Tories are relying on the pound’s symbolism as a barrier to foreign competition. But what they don’t mention is that embracing ecommerce and rejecting the euro will increase international competition faster than embracing the euro and fighting ecommerce. There are of course uniquely British products which are safe from competition — blood pudding, bagpipes, marmite — but these aren’t much of a counterweight to those products where quality and price matter far more than national origin — computers, books, cars, not to mention airline tickets and stock trades. By vastly increasing the width of choice offered to British consumers (and therefore the depth of competition faced by British producers), ecommerce will make the pound as a symbol of an aloof Britannia irrelevant through other means. No matter what currency their goods are priced in, a borderless economy will weave their island inextricably into the fabric of the world.

by Clay Shirky

How Television Ratings Portend the Death of Mass Media

Television is becoming more like the Web.

First published on FEED, 10/14/1999.

The fall Nielsen ratings are now in for the new crop of prime time TV shows, and in what is becoming a predictable pattern, the results are underwhelming. Despite the promotional avalanche that accompanies every new series, this year’s new shows failed to attract as much attention as the hits of previous seasons. This steady decline in popularity of each successive season’s offerings is fairly dramatic. Not since “Seinfeld” has a show managed to reach one out of four American households. The top three (“E.R.,” “Frasier,” “Friends”) only reach about one household in five, six, and seven, respectively, and even that tops the current season. The most popular new show (“Stark Raving Mad”) reached barely one in eight. There’s no mystery about this process: Proliferation of choices divides attention, and every year sees increased choices on cable, the video store, and the web. The mystery is that the price to run a commercial on one of these hit shows is at an all-time high.

If the most popular shows reach fewer people every year, why are advertisers willing to pay more for commercial slots on those shows? The first and simplest reason is that the dot-coms are pouring a huge amount of money into advertising, and driving up prices. But the more interesting answer has to do with the large-scale splintering of mass media. Broadcast networks and advertisers are used to buying audience attention in bulk, and in prime time bulk means at least 20 million households at a time. These massive audiences for prime time shows, developed when the entire TV universe included just ABC, NBC, and CBS, provided the kind of reach that made it possible for the General Foods of the world to turn a new brand into a household word in a few weeks, by simply buying the attention of a quarter of the country at once. As the shows which can offer this kind of mass audience disappear, things are getting trickier for the networks — they can sell the advertiser two shows that each reach an eighth of the country, but how can they prove that you are reaching two different people and not just the same person twice? They can’t. The symbiotic relationship between mass media and mass marketers is being threatened by the increasing number of media outlets, which sub-divides the mass audience into smaller and smaller niches.

Broadcast TV can charge higher prices for fewer households because the mass marketers simply have nowhere else to go. Despite the increasingly anemic performance of the most popular TV shows, there isn’t another medium that offers the option of reaching 10 million households at the same time with a single ad — even giant portal sites fragment their reach across a number of offerings. Indie mutiny has been in the works for years, but these recent Nielsen numbers may be a cultural bellwether — among the first concrete data to confirm the decay of mass media. Like a small island with a receding coastline, the total area the broadcast networks cover shrinks every year, but for mass marketers, it’s still the only ground above water. It is obvious that both the networks and their advertisers are soon going to have to adapt to a fragmented media market where nothing regularly reaches 20 million people, and the only way to get mass will be niche plus niche plus niche. In the meantime, though, old habits die hard, and it is these old habits of looking for mass that are driving ad rates up for hit shows even as those shows lose the very audience that makes them valuable. If the truth in advertising law had any real teeth, CBS would have to change the name to “Almost One Person In Eight Loves Raymond.” Don’t expect even that to last.

by Clay Shirky

Kasparov vs. The World

What the Kasparov community chess match tells us about community.

10/21/1999

It was going to be acres of good PR. After the success of Garry Kasparov’s chess
matchup with IBM’s Deep Blue, Microsoft wanted to host another computerized chess match this summer — Kasparov vs. The World. The setup was simple: Kasparov, the John Henry of the Information Age, would play white, posting a move every other day on the Microsoft chess BBS. “The World” consisted of four teenage chess experts who would analyze the game and recommend counter-moves, which would also be posted on the BBS. Chess aficionados from around the world could then log in and vote for which of the four moves Black should play. This had everything Microsoft could want — community, celebrity, online collaboration, and lots of “Microsoft hosts The World!” press releases. This “experts recommend, The World votes” method worked better than anybody dared hope, resulting in surprisingly challenging chess and the ascension of one of The World’s experts, Irina Krush, into chess stardom. Things were going well up until last week, when Microsoft missed a crucial piece of email and the good PR began to hiss out of the event like helium from a leaky balloon.

While Deep Blue was a lone computer, here Kasparov’s opponent was to be the chess
community itself, a kind of strategic “group mind.” Since communication was the glue
that held the community together, it’s fitting that the game came unglued after a missed email. During last week’s end game, it was generally agreed that The World had made a serious tactical error in move 52, but that there was still the possibility of a draw. Then, on October 13th, Ms. Krush’s recommendation for move 58 was delayed by mail server problems, problems compounded by a further delay in posting the information on the Microsoft server. Without Ms. Krush’s input, an inferior move was suggested and accepted, making it obvious that despite the rhetoric of collaboration, the game had become Kasparov v. Krush with Kibbitzing by The World. Deprived of Ms. Krush’s strategic vision, the game was doomed. The World responded to this communication breakdown by collective hari kari, with 66% of the team voting for a suicidal move. Facing the possibility of headlines like “The World Resigns from Microsoft,” the corporate titan rejected the people’s move and
substituted one of its own. The World, not surprisingly, reacted badly.

Within hours of Microsoft’s reneging on the vote, a protest movement was launched,
including press releases, coordinating web sites, and even a counter-BBS which archived articles from the Microsoft chess server before they expired. Microsoft had run afoul of the first rule of online PR: On the internet, there is no practical difference between “community” and “media”; anyone with an email address is a media outlet, a tiny media outlet to be sure, but still part of the continuum. Since online communities and online media outlets use the same tools — web sites, mailing lists, BBS’s — the border between “community interest” and “news” is far easier to cross. The Microsoft story took less than a week to go from the complaints of a few passionate members of the chess community to a story on the BBC.

Microsoft made the same famously bad bet that sidelined Prodigy: By giving The World a forum for expressing itself, it assumed that The World’s gratitude would prevent criticism of its host, should anything go wrong. As Rocky the Flying Squirrel would say, “That trick never works.” What started as a way to follow on IBM’s success with Deep Blue has become a more informative comparison than even Microsoft knew. Two computerized chess games against the World Champion — one meant to display the power of computation, the other the power of community — and the lesson is this: While computers sometimes behave the way you want them to, communities never do. Or, as the Microsoft PR person put it after the game ended: “Live by the internet, die by the internet.”

by Clay Shirky

Sun’s Quasi-Open Source Model

Sun Microsystems’s misunderstanding of Open Source.

First published in FEED, 10/7/1999.

If imitation is the sincerest form of flattery, the Open Source movement should be blushing from head to toe by now. Its most recent admirer is Sun Microsystems, who has just announced that it will make the source code to its Solaris operating system available under something it calls a “Community Source” license. In making such a bold move (Solaris is their core product) Sun is embracing everything that has made the Open Source movement such a success. Everything, that is, except that bit about opening up their source code.

Sun is attempting a quasi-Open Source move because while they want what Linux has (an army of talented developers working for free) they also want what Linux doesn’t have (commercial control, patents on intellectual property, and a steady stream of income). They are trying to split the difference by creating a license which only allows you access to Solaris source code if you promise not to make any money from it, and only if you agree to co-ordinate any changes you make with Sun. In order to stave off criticism of these restrictions (which don’t exist under real Open Source licenses), Sun has wrapped itself in the flag of community — after all, what could possibly be wrong with a “Community Source” license?

In net-speak, “community” no longer means “a gathering of likeminded people,” it is just a reductive variable written into every dot-com business plan. The problem with Sun’s “Community Source” is that Open Source software is not written by communities, it is written by individuals — thousands of individual developers, each sharing their work with the others. The difference is a subtle but important one; Sun is betting that Open Source software works as well as it does because of community feeling and collaboration, when in fact it is mostly driven by individual selfishness. In an Open Source project, new features come not from what a developer imagines some hypothetical client might possibly want to do someday, it comes from what the developer him or herself wants to do right now. Like a game of “Which of these things is not like the others?,” commercially developed software is starkly different from Open Source projects: Linux exists because Linus Torvalds wanted a Unix clone that ran on cheap hardware, Apache exists because Brian Behelendorf wanted a good web server, Perl exists because Larry Wall wanted to make writing reports easier, Solaris exists because Sun wanted to make money.

Most of the valuable aspects of open source software — its cleanliness in implementation, its compact size, its ability to run on cheap hardware — are a direct result of this selfishness, since developers are making the software they want to use. Like the paradox of the free market, where reducing central management increases economic efficiency, the paradox of Open Source is that by reducing commercial control, software can actually improve faster. What Sun doesn’t understand is that this core aspect of Open Source is indivisible: The thing that makes Linux so desirable for developers is the very thing that gives commercial software companies indigestion — it transfers control of the software to the individual. There’s no way to split the difference, because there’s no difference to split.

Sun’s license guts the very freedoms that drive open source adoption in the first place, because without real transfer of power, there is no incentive to use open source software. And this problem is not Sun’s alone — giants like Microsoft and Oracle will also need strategies for dealing with Open Source, and sooner rather than later. As with so much about the internet, Open Source is about removing the middleman and transferring control to the individual: Linux is the net’s first disintermediated operating system. No matter how many iterations of its “Community Source” license it goes through, in the end Sun is going to discover that the only way to get the advantages which the Open Source movement enjoys is to open its source.

by Clay Shirky

An Open Letter to Microsoft

Software usability and what it means to be a media company.

First published on SAR, 09/99.

Dear Microsoft,

I tried to give you some of my money last weekend and you wouldn’t let me. As this sort of behavior might be bad for your long-term profitability, I thought I’d write and explain how you can fix the problem.

Last Sunday night while visiting friends, I remembered that I was running out of time for a discount plane ticket, so I opened Internet Explorer 3.0, the browser running on my friend’s machine, and went to Expedia to make my purchase. (I trust it will not be lost on you that both IE and Expedia are Microsoft products). After submitting my flight details on Expedia, the results page created so many browser errors that I couldn’t even see what flights were available, much less buy a ticket.

Do you understand that? I wanted to use your product on your service to give you my money, and you wouldn’t let me. Being a good citizen, I wrote to the Customer Service address, expecting no more than a piece of ‘Thanks, we’ll look into it’ mail. What I got instead was this:

“Thank you for contacting Expedia with your concerns. Since Microsoft has upgraded the website, they have also upgraded the version you should use with Expedia. The version will be more user-friendly with Internet Explorer 4.0 and above.”

Now before you think to yourselves, “Exactly right — we want to force people to get new browsers”, let me contextualize the situation for you:

I cannot use IE 3.0 to buy tickets on Expedia, but…

I can use IE 3.0 to buy tickets on Travelocity.
I can use Netscape 3.0 to buy tickets on Expedia.
I can even use Netscape 3.0 to buy tickets on Travelocity.

Let me assure you that I bought my tickets as planned. I just didn’t buy them with you.

I understand from what I read in the papers that your desktop monopoly has atrophied your ability to deal with real competition. Get over it. If you really want to start building customer-pleasing businesses on the Web (and given the experience I had on Expedia, I am not sure you do) then let me clue you in to the awful truths of the medium.

First, you need to understand that the operating system wars are over, and operating systems lost. Mac vs. Windows on PCs? Dead issue. Linux vs. NT? Nobody but geeks knows or cares what a web site is running on. There’s no such thing as a “personal” computer any more – any computer I can get to Amazon on is my computer for as long as my hands are on the keyboard, and any operating system with a browser is fine by me. Intentionally introducing incompatibilities isn’t a way to lock people in any more, its just a way to piss us off.

Second, media channels don’t get “upgrades”. You have made a lot of foolish comments about AOL over the years, because you have failed to understand that AOL is a media company, not a software company. They launched AOL 1.0 at about the same time as you launched MS Word 1.0, and in all that time AOL is only up to version 4.0. How many “upgrades” has Word had – a dozen? I’ve lost count. What AOL understands that you don’t is that software is a means and not an end. As a media company, you can no longer force people to change software every year.

Third, and this is the most awful truth of all, your customers are the ones with the leverage in this equation, because your competition is just a click away. We get to decide what browser we should use to buy plane tickets or anything else for that matter, and guess what? The browser we want to use is whatever browser we have whenever we decide we want to buy something. Jeff Bezos understands this – Amazon works with Netscape 1.22. Jerry Yang understands this – there is not a single piece of javascript on the Yahoo home page. Maybe somebody over at Microsoft will come to understand this too.

Thinking that you can force us to sit through a multi-hour “upgrade” of a product we aren’t interested in solely for the pleasure of giving you our money is daft. Before you go on to redesign all your web sites to be as error-prone as Expedia, ask yourself whether your customers will prefer spending a couple of hours downloading new software or a couple of seconds clicking over to your competition. Judging from the mail I got from your Customer Service department, the answer to that question might come as a big surprise.

Yours sincerely,

Clay Shirky

by Clay Shirky

Why Stricter Software Patents End Up Benefitting Open Source

Why commercial software controls will backfire.

First published in FEED, 8/18/1999.

As the software industry cracks down on its customers, the software itself is opening up. Open Source software, the freely available alternative to commercial software, is making inroads in the corporate world because of its superior flexibility, adaptability, and cost. Despite this competition, the commercial software industry has decided that now’s the time to make licensed software less flexible, less adaptable, and above all, more expensive. A proposed new law, called the Uniform Computer Information Transactions Act (UCITA) would give software manufacturers enormous new powers to raise prices and to control their software even after it is in the hands of their customers. By giving the software industry additional leverage over its customers, UCITA will have two principal effects: first, it will turn software from a product you buy to a service you pay for again and again. Second, its restrictions will greatly accelerate the corporate adoption of Open Source software.

All trade groups aspire to become like OPEC, the cartel that jacks up oil prices by controlling supply. The UCITA working group is no exception: UCITA is designed to allow software companies relief from competition (companies could forbid publishing the results of software comparisons), and to artificially limit supply (any company which was acquired by a larger company could be forced to re-license all its software). The most startling part of UCITA, though, is the smugly named “self-help” clause, which would allow a software company to remotely disable software at the customer’s site, even after it has been sold. This clause could be invoked with only 15 days notice if a software developer felt its licensing terms were being violated, making the customer guilty until proven innocent. UCITA’s proponents disingenuously suggest that the use of “self-help” would be rare, as it would make customers unhappy — what they are not as quick to point out is that the presence of “self-help” as a credible threat would give software companies permanent leverage over their customers in negotiating all future contracts.

Unfortunately for cartel-minded software firms, the OPEC scenario is elusive because software isn’t like oil. Software has no physical scarcity, and the people who know how to create good software can’t be isolated or controlled the way oil wells can. Where UCITA sets out to make software a controlled substance, the Open Source movement sets out to take advantage of software’s innate flexibility of distribution. By making software freely available and freely modifiable, Open Source takes advantage of everything UCITA would limit — Open Source software is easy to get, easy to modify, and easy to share. If UCITA becomes law, the difference between Open Source and commercial software will become even more stark. Expect Open Source to do very well in these circumstances.

Economics 101 tells us that people make economic decisions “on the margin” — a calculation not of total worth to total cost, but of additional worth for additional cost. For someone who wants a watch for telling time but not for status, for example, the choice between a $20 Timex and a $20,000 Rolex is clear — the $19,980 marginal cost of the Rolex knocks it out of the running. In the case of Open Source vs. commercial software, the differences in cost can be equally vast — in many cases (such as the Apache web server) the Open Source solution is both cheaper and better. Cartels only work if there is no competition, a fact the UCITA group seems not to have grasped. If UCITA becomes law — that could happen as soon as December — the commercial software industry will be sending its customers scrambling for Linux, Apache, and the other Open Source products faster than they already are.

by Clay Shirky

The Internet and the Size of Government

Challenges to governmental control.

First published in FEED, 8/11/99.

Fritz Hollings, Senator of South Carolina, and Zhu Rongji, Premier of China, have the same problem — the internet has made their governments too small. In Mr. Hollings’ case, ecommerce is threatening to damage South Carolina’s local tax base, while Mr. Zhu is facing threats to the Chinese Communist Party from dissident web sites in other
countries. They have both decided to extend the reach of their governments past their current geographical boundaries to attack these problems. The final thing the senator and the Premier share is that their proposed solutions will accelerate the changes they are meant to postpone — you can’t fight the effects of the internet without
embracing it, and you cannot embrace it without being changed by it.

Fritz Hollings’ problem is simple — states can collect taxes on local sales but not on ecommerce, because ecommerce has no respect for locality. His proposed solution is equally simple: a 5% national ecommerce tax (the “Sales Tax Safety Net and Teacher Funding Act”) to collect money at the Federal level and funnel it into state “subsidies” for teachers salaries. Unfortunately for him, “No Taxation Without Representation” cuts both ways — education policy is one of a state’s most important functions, and a Federal education tax opens the door for a Federal education policy
as well. If Hollings is worried about loss of state power, extending the states’ reach to the level of national taxation is a good short-term solution, but in the long run it will worsen the problem — states who defend their autonomy will watch the internet erode their revenue base, but states who defend their revenues will watch the internet erode their autonomy.

Zhu Rongji’s dilemma is more complicated, but no less stark — China’s communist party is vulnerable to international dissent because political web sites have no respect for national borders. Shortly after the Chinese government banned the Falun Gong sect for “jeopardising social stability,” China’s Internet Monitoring Bureau attacked Falun Gong web servers in the US and UK. The People’s Liberation Army newspaper called this action a “struggle in the realm of thought,” indicating that China now respects no national borders in its attacks on dissent. During the Kosovo crisis, China argued loudly for non-intervention in the affairs of other countries, but these attacks on foreign web servers tell a different story — non-interference in a connected world is incompatible with a determination to stifle dissent. The Falun Gong attacks will help
protect the Communist Party in the short term, but by demonstrating to future dissidents how afraid the Party is of the web it makes them more vulnerable to the winds of change in the long haul.

Governments, like companies, are being forced to respond to the increasingly borderless movement of money and ideas, but unlike companies they have no graceful way of going out of business when they are no longer viable. In place of mergers and bankruptcy, governments have wars and violent overthrows — governments will do anything to avoid closing up shop, even if that would be better for the people they are meant to serve. This makes governments more successful than businesses in fighting change in the short term, but in the long run, governments are more brittle and therefore more at risk. Zhu Rongji and Fritz Hollings have both adopted old men’s strategies — preserve the
present at all costs — but postponing change will heighten its force when it does come. In another five years, when the internet has become truly global, the damage it will do to things like South Carolina’s tax base and the Chinese Communist Party will make it clear that Messr’s Hollings and Zhu are trying to put out fire with gasoline.