I have come to avoid talking about this stuff with people. The firstquestion anyone asks me is “what is Web 2.0?” Unfortunately for theensuing conversation, it’s a little tricky to provide a straightanswer. Every time you find a neat expression for summing the whole Web2.0 thing up, I immediately think of an exception, or three, or waysthat the definition doesn’t really get us anywhere.

In the list that follows, I’ve taken a lot of these characteristics or definitions from Tim O’Reilly’s What is Web 2.0?, and also Paul Graham’s Web 2.0 and Jason Fried’s user survey about the term.

1. The wisdom of crowds: We’re thinking here ofthings like digg that harness collective judgements to decide theimportance of news stories. People talk about the power of ‘networkeffects’ when they’re keen on this definition. Google Search works likethis by using the number and quality of inbound links to decide apage’s importance. But the whole idea does not apply to Google Maps, orany of the other Software-as-a-Service (SaaS) crowd e.g. Basecamp,Writely, 30boxes, etc., which are nonetheless thought of as being Web2.0. Nor does it apply to social networks that are just aboutdeveloping and maintaining friendships, like MySpace, though they dobenefit from network effects, of course.

2. Shared Web Applications. One of the definitionsfrom Jason Fried’s list and quite promising. Almost the opposite of ourfirst definition, since it quite clearly applies to things likeBasecamp, Writely and 30Boxes. However, there are some Web 2.0applications that have no social element whatsoever, e.g. Pandora,Google Maps, Orchestrate, goowy. I’m also struggling with the idea ofweb applications. I can see why digg and Google Search areapplications, but to have this as a defining feature of Web 2.0 wouldmean classifying MySpace as an application. And if I allow that, thenalmost any web site becomes an application.

3. Web as platform: It’s hard to know where thisone starts and ends. In some sense, every web page is using the web asa platform. For Tim O’Reilly, who came up with this explanation, itmeans services that could not exist without the web, and he’s thinkingof things like eBay, craigslist, Wikipedia, del.icio.us, Skype andDodgeball. For me, that means that every online community could fallinto this category. Are message boards and usenet Web 2.0? Most peoplewould say not. Too broad.

4. User Participation: This is about the pointingout the differences between old-fashioned newspaper and magazine sitesand new services like YouTube, flickr, and OhMyNews where the consumersare also the creators. The expression ‘Read/Write web’ crops up amongproponents of this definition. Again, it’s rather too broad, so itcould equally apply to message boards, but also too narrow in adifferent way, since it misses the SaaS sites.

5. Rich User Experience: Web 2.0 sites use CSS,AJAX and other technologies to enhance usability and create dynamicpages able to display more information in the same space. But hang on,the default MySpace page is probably one of the least “rich”imaginable. Oh, apart from craigslist. And until they introduced searchterm prediction earlier this year, Google Search didn’t use any fancypresentation technologies at all. Also, the presence of anAJAX-enhanced shopping cart on an etailer site doesn’t really capturewhat people mean by Web 2.0. Dell.com, for example, has had a ‘live’shopping cart for years. It’s a good cart, but Web 2.0?

6. Marketing Buzzword: This is what all thesceptics say. So Google Search and Amazon and eBay and craigslist, allof which are believed to be Web 2.0 applications, because they matchsome of the other characteristics I’ve described here, are just somesort of modern fad that’s going to fade away, are they? The same thinggoes for anyone who wants to describe Web 2.0 as “the new stuff on theweb”. I do agree, incidentally, that Web 2.0 has become a marketingbuzzword, it’s just that I think that it’s also more than that.

7. Data is the next Intel Inside: Though it’s a bitof a mouthful, I actually quite like this one. Again, it’s from theO’Reilly paper. Data management is a core competency of Web 2.0companies. “SQL is the new HTML,” is another quotation from the paperalong these lines. All the Web 2.0 crowd, and we can go from giantslike Amazon and Google to startups like 30boxes and Orchestrate,operate mainly from databases to contain and present personalised viewson that data. There’s two problems here: (a) data management isn’tquite such a sexy idea as people would want and (b) a lot of the Web1.0 companies were also about finding clever ways to use databases e.g.Altavista, Lastminute.com.

8. Permanent Beta: Web 2.0 applications arere-released, re-written and revised on an ongoing basis, putting paidto the yearly release cycle that characterised earlier softwaredevelopment. Most Google applications, for example, are still in Beta.flickr is rumoured to sometimes be revised every 30 minutes. MySpaceand the other social networks add extra features every couple of weeks.I think that this is a clear characteristic of Web 2.0 apps. But it’salso become a feature of mainstream applications. Windows and MacOS,for example, get new fixes and patches every month. Antivirus programsare updated every day, but they aren’t Web 2.0, are they? The samething goes for ‘lightweight programming models’. Also, I think peoplemean more by the term than the way in which it’s programmed. Most userscouldn’t care less, they just want it to work well.

9. Using the web as it was meant to be used: Thisone is from Paul Graham’s essay on the subject. He’s referring to theincreases in usability that are achieved through very good design aswell as things like AJAX, but also by allowing users to develop theirown ways of organising the information they have, the way del.icio.usand flickr do. Again, I have a couple of problems here. Firstly, it’s abit loose: I’m sure that there were always some very well-designedsites that worked exactly as you wanted them to. The old (and nowdefunct) UK train timetable site was a perfect web app in many senses:it got you train times quickly and easily. But no-one would call it Web2.0. Second, it’s a little self-satisfied as a definition and implieswe’re reaching an end-point. A lot of the sites described as Web 2.0have quite clearly got it wrong.

10. Nothing: One of the more popular answers inJason Fried’s user poll. It’s a hard one for me to evade given that Ihave just come up with counter-examples or objections to all thedefinitions I’ve been able to find. Still, I resist the idea that thisis nothing. Here are two answers to the question I think are true. (a)A Web 2.0 application, site or service will have a combination of thefeatures given above. Just as black and white aren’t satisfactory fordescribing the colour of everything, neither is Web 1.0 vs. Web 2.0. Itisn’t a binary division of the web, or a revolution. Instead, we have aspectrum. Those sites and services which satisfy a number of thesecriteria or characteristics are more Web 2.0 than those which don’t.That is not a value judgement, of course. Sites with no Web 2.0 featurescan still be wonderful. Sites with a lot of them can be awful. Also (b)Web 2.0 is still too young as an expression to have reached the pointwhere we have consensus about what it means. It means different thingsto different people at the moment. It may only be with hindsight thatwe come to be able to narrow things down enough to be able to say whatit was in one sentence.