It's the Data, Stupid

In any programming environment, you have basically three pieces: the users, the programming, and the data. Google has figured out that the first two of those pieces don't matter.

The users will always be there, for one program or another. Thousands of startups each year go out searching for users, and a lot of them find them. The entire web-based business consists of one huge push for finding users. In terms of the larger web eco-structure, they're a commodity -- although a fickle one.

The programming is also a commodity. Programming languages and hardware costs are minimum and knowledge about programming is spread out across the world more than at any other time in history. And all of those driving factors are in their early stages. I live in a very rural area of the states, and I can't go to Walmart without bumping into somebody who knows how to code. From script kiddies to Excel macro wonks to homeschool bloggers, everbody and their brother knows how to make the computer do things.

The data, however, is a different story. Users may be fickle, but the more time they spend putting data into a certain program the more attached to that program they become. Is this just a function of how cool the program is? We used to think that. Most people still do -- you'll hear all kinds of arguments about how FaceBook is better than program X or how AIM kicks ICQs butt. People argue that the value in programs are about the problems they solve and the way they solve them. This is incorrect.

The initial value in programs is about the problem they solve and the way they solve them. Once the problem is solved, however, it's just ones and zeros, and any smart tenth-grader can write similar code during his free time at summer camp. What that means, to me, is that programmatic solutions are like rock-and-roll songs, not like building the space shuttle. Sure, there was a time where the chasm between the problem and the solution was so great that it was a work on engineering just to get it done. But heck, there was also a time when driving across the country was just such a similar task. Now you can get from New York to California in a few days for about the cost of a week's worth of wages. Times change.

But what truly changes, what has the most value in the world is data. Whatever your solution, whatever your rock-and-roll band, at the end of the day you've got some system for users creating data that is important to them -- a list of their friends, chat history, what their friends are doing, their financial records, stuff they've searched for on the internet, etc. Looked at it this way, the only purpose of creating a startup is to create new kinds of data that people are really interested about and serving it up to them in a catchy, viral format. The catchy viral format part will get old after a while, but the value of the data remains.

The more of this data you have, the more solutions you can create that have more immediate impact to the users. I think everybody gets this part -- that's why there's such a push for the OpenSocial format. But there's a darker side too: whoever owns this data can correlate and use it in ways that may be non-obvious to the outside world. Google can sell ads based on key terms in your emails, for instance. Social networking sites can sell lists of contacts (although most don't). We all seem to think we know when the line has been crossed, but it's my contention that simply having all the data in one place is, of itself, a counter-productive thing for society in the long run. Information is power, whether that power is currently being exercised or not.

It's the data, stupid.