On Web 2.0, application uses YOU!

25 March 2006

Most of what separates Web 2.0 from pre-Web 2.0 is not really about the web. Paul Graham writes that Web 2.0 is about three things: AJAX, Democracy, and Don't Maltreat Users. Democracy here means using users to collect decision-making information, such as evaluating whether something is worth reading or buying, for other users. Yes, visitors come to your site to get value that other users brought to it. Ka-ching!

Tim O'Reilly writes that Web 2.0 is about constantly improving software that's available as a service; mixing data from multiple sources, including users; offering your own data and services in a flexible, mixable way; and going beyond the page metaphor (which sounds like AJAX to me.)

Anyway, the key part of being Web 2.0 is that you're building value from many small information contributions that users don't mind making. Every user whitewashes a little bit of the fence. Paul Graham points out that Google is a good example of this. When I say great burritos in San Francisco, Google uses my link-making work (and that of others) to amass awesome burrito (and other thing)-recommending power and rule the world. And I like it because I want my favorite San Francisco burrito place to succeed.

When people put geographical directions up using microformats, someone will crawl them and string the route decisions together to get a directions search engine with common sense (because it borrowed the common sense of millions of users) that doesn't tell people to make an illegal left into oncoming traffic, the way a certain map site used to tell me to leave my old house every day. (70mph combined speed motor vehicle slalom! Yaaaaahooooo!)

Where AJAX fits into all this is that you're snarfing one reputation information unit per click, quickly, instead of waiting for a whole page to render to suck the value out of the user's head into your MySQL cluster where it becomes valuable. And you have to let users pull data back out and mix it, since that creates attention incentives for other users to push data in.

So far this Web 2.0 stuff sounds like it's all about web sites. How can companies that aren't basically web sites or mail-order catalogs be Web 2.0? Some already are. Remixed FedEx lately? Download their sample code and try their API.

Hold on a second—you don't have to be a FedEx partner to do that? No, and that's the first concrete difference between Web 2.0 and non-Web-2.0 companies. From a pre-2.0 point of view, the partner program is what enables companies to interact with you. Start thinking 2.0, though, and the partner program looks more and more like pointless bureaucracy that keeps non-partner companies out. Just as you want Googlebot to crawl your product pages, (and some of you will go flame Matt Cutts if it doesn't) you want any company whose stuff can plug into yours to try your API.

You could probably do a pretty reliable Web-2.0-or-not-o-meter based on dates in the RSS feed for API announcements vs. dates in press releases matching /partner/i.

What next? Larry Augustin points out that sales and marketing accounts for 82 percent of new software license revenue. Ouch! Let's throw some Web 2.0 magic at that number. And I don't mean the sales part. The web, together with open source licensing, easy-to-demo ASP, and virtualization, is already taking a huge chunk out of the sales side.

But a huge, expensive part of software marketing is involved in information gathering, too. It's really expensive to hire Software Marketing people to gather requirements from users, write big word processor documents full of what the users want, and show each other Microsoft PowerPoint presentations, then tie up the developers showing them again.

Within organizations, we know all about using Extreme Programming and other agile methodologies that replace the obsolete-as-soon-as-finished overdocumented waterfall development process. But often, as soon as projects cross organizational lines, we're back to the kind of thing that with-it software people rightly make fun of.

Where web sites are concerned, Web 2.0 took the excellent idea of APIs from software developers and gave them to the webmasters. In off-web businesses, it's taking development methodologies from the developers and giving them to marketing people.

But how do you know what to build when the Big Dumb Word Processor Document of What To Build is gone? This is where it gets fun. The customers are already telling you what they want, if you know how to listen to them. Web 2.0 companies are concentrating on building the place for that conversation to take place, instead of writing the damn document themselves. You get better, faster, cheaper when you lose the waterfall.

After all, some users will sit still for Focus Groups and other 20th-century marketing, but even if they do, you're stuck dealing with the resulting data yourself. Canonical Ltd. takes the specification process where Google takes the search result ranking process—outside the company as much as possible. That doesn't mean that random users design Canonical's products for it, any more than search engine spammers define Google results. But in Web 2.0 you get the users to whitewash the fence.

There's some overlap between being a Web 2.0 company and being an open source one. Here's where I think Web 2.0 goes further than open source. If open source is trees, Web 2.0 is hemp. Instead of harvesting big particpation from a committed developer, tester, user, partner, or customer, you get a small quantity of fiber per transaction, fast, and you do a lot of them.

I think there's a limit to how far pure conversationality and social software can take this, and that we're going to have to get hairier information-sucking-a-little-bit-at-a-time-from-peoples-heads tools such as prediction markets. But Web 2.0 unplugged from the web can take us a lot further, faster, than the alternatives can, and, especially in the area of business software, companies are already using it.