Ethan Kurzweil has a presentation posted on his blog from 8 months ago that I just saw for the first time today. This is my favorite slide in it. I would love to see the fit curve on this slide proved out with empirical hosting/development costs data from startups over time. Considering how long Bessemer has been in the business, I’m sure they could do this if interested.
This graph feels intuitively right to me… but what does it mean when it costs pennies to build and host a high-scale web service? As costs approach zero, I’ll bet there will be an explosion in the number of vertical-specific services that serve small niche audiences. By explosion, I mean increasing by orders of magnitude. I can already feel the reverberations of that explosion today echoing backwards in time as the number of web services launching everyday is accelerating.
Very few of these services will be businesses, but that doesn’t matter because they’ll cost nothing. That’s the point… businesses are hard to create, web services will be trivially easy and cost zero.
Or think of it this way:
What every dotcom spent well into the seven-figures, hundreds of thousands of hours, and countless cycles for to get content up online, one person can do by himself with tens of hours, Drupal, and a shared hosting account.
That’s revolutionary. It means we can take for granted so much of the “infrastructure” part of the equation, but not lose the knowledge. And because the basic tools are becoming cheaper and more metatools keep coming out, it means there isn’t much comparable cost in rewriting pretty big parts of the stack.
