In a typical month I get more questions about Web 2.0 than about Ajax. The Ajax questions cover a broad range of topics, the Web 2.0 questions are mostly “what the heck is Web 2.0?” I don’t really have a good answer, but these frequent questions have caused me to ponder why Tim O’Reilly’s definition seems so flabby. Here then is one of my infrequent meditations on Web 2.0.
It is well over a year since Tim O’Reilly laid out his ideas on Web 2.0. At the time he wrote his seminal article, “Web 2.0″ had already been in common parlance for a year and a half “with more than 9.5 million citations in Google.” Since that time, Web 2.0 has become synonymous with Ajax, social bookmarking, user generated content and other buzzwords of the day. It is my guess, however, that Tim didn’t intend his 7 characteristics of Web 2.0 to be taken ala cart, or “like a fruit salad.” But I do think that as the concepts of Web 2.0 have been proven out in the real world, some of his 7 characteristics — especially the process ones — are turning out to be best practices for any sort of web application.
And Then There Were Five
So which of the characteristics should we peel from Web 2.0? My first nominee is #4, “End of the Software Release Cycle.” Also termed “the perpetual beta,” it is the idea that systems will undergo constant change and that users should be treated as co-developers. Further, companies must have the operational chops to support this constant change. There is nothing wrong with #4 in principle, but there are a number of reasons why it isn’t really a part of Web 2.0:
- Let’s be honest, we’re talking about short release cycles, not no release cycles.
- Early adopter will accept and even embrace quickly changing software. For the great mass of consumers and established businesses that depend on your service, rapid changes can be problematic.
- Web based businesses have gone through cycles of innovation and consolidation since the early 90’s. Small, nimble companies (including Yahoo!) pushed features into their Web sites and applications in short release cycles. Once these companies became larger, innovation took a back seat to stabilization and growth. Release cycles became longer and, as happens with so many large organizations, many of them stopped taking risks. But since innovations in browsers and other technologies change the terrain of the web so frequently, and the cost of distribution has remained so low (the price of a good colo remains quite reasonable, unless the sluggish behemoths that are against net neutrality get their way), the cycle of innovation and consolidation has remained consistently short. We’ll rediscover the end of the software release cycle each time when Web 3.0, 4.0, and 5.0 all roll around.
- Small, nimble startups are usually better able to push new functionality in short releases. In fact, it is often an unfortunate side effect of a lack of process. Big companies have to work hard on shortening their cycles. As such it is a triumph of governance and process rather than technology. Claiming this fundamental competitive advantage for Web 2.0 is overreaching. It is neither a necessary or sufficient condition for Web 2.0.
Characteristic #7, “Rich User Experience” is also a candidate for eviction. The concept here is that using Ajax you can deliver GUI style applications. The implication for Web 2.0 is that rich user experience will enable all sorts of apps not seen before:
We expect to see many new web applications over the next few years, both truly novel applications, and rich web reimplementations of PC applications. Every platform change to date has also created opportunities for a leadership change in the dominant applications of the previous platform.
While we are seeing new applications, such as Google’s collaborative spreadsheet, we are seeing many more retoolings of traditional shopping cart and other Web 1.0 applications. Just like #4, #7 will not distinguish Web 2.0 from Web 1.0 apps — i.e. it is not a necessary nor sufficient condition for Web 2.0.
Why bother trimming away at Web 2.0? I think there is some merit to the Web 2.0 meme, but making its characteristics so broad obscures its true nature. I think other concepts — the wisdom of crowds, the long tail, and software above the level of a single device — are necessary and sufficient conditions for Web 2.0.
The Wisdom of Crowds, the Free rider Problem, and the Paradox of Choice
Of all of the Web 2.0 concepts, I think the Wisdom of Crowds is the most import. In my opinion, all of the other principles are in service of enabling user generated content and more generally the “noodle of the mob”, err, the wisdom of the crowd. But while it is the most important principle, it is also the most fragile; online communities have been around since before 1979 (Usenet, Compuserve, etc.). They all seem to have a lifecycle, however — they rise and fall in a predictable arc.
That something so fragile and perishable should be at the heart of Web 2.0 is worrisome. I’m not sure that anyone has the answers as to why all online communities wither away, but I have a few ideas.
- In knowledge management, there is the concept of tacit knowledge capture, i.e. a community does something together (a project, for example) and the artifacts that they produce — discussions, documents, FAQ’s — are valuable knowledge. When you try to gather knowledge without “doing something,” you have to insert more and more management — librarians — to make things work. Communities where users simply share testimonials about how great a product works is going to wither sooner than most or maybe never live at all.
- When the incentives of all the participants are aligned, communities work well. When they are not, such as when commercial interests intrude, the community degenerates. This is how USENET eventually succumbed to a flood of spam. These are bad actors, as opposed to the free riders below.
- At their best, online communities are like communism: From each according to his ability, to each according to his need. But why did communism fail? Because of the free rider problem. Free riders are actors who consume more than their fair share of a resource, or shoulder less than a fair share of the costs of its production. The free rider problem is the question of how to prevent free riding from taking place, or at least limit its negative effects. Unfortunately, it seems that online communities suffer from a massive free rider problem, with lurkers, occasional contributors and frequent contributors having about a 90-9-1 ratio.
- Communities don’t die because they are fads, like pet rocks. See this analysis of the Friendster collapse to see why it was more about #2 above.
The wisdom of crowds is the methane that fuels the torch of Web 2.0. As such it is important to solve some of the issues that plague the effectiveness and longevity of online communities.
Anyhow, that’s it for my meditation on Web 2.0. Sorry I wasn’t able to come up with a pithy Malcolm-Gladwellism to summarize my thoughts.