Dalton Caldwell’s audacious proposal has created quite a stir in the industry by proposing App.net, a realtime feed API and service that you have to pay to use. As I understand it, it’s Twitter, but more focused on the API and protocol, and, obviously, it’s not free.

It’s refreshing to see such against the grain thinking, and based on the current sentiment against companies with no clear business model, I think this was only a matter of time. I would love to see this work, because I think it takes guts, and I think Caldwell is a smart and capable guy (see: Imeem, Picplz). But I don’t think it will, for the simple reason that App.net relies on network effects, which is fundamentally at odds with an “everyone pays” model.

Paid models work really well for some web apps, the 37signals family of SaaS products being a prime example. Their products, however, do not rely on network effects to be successful. If I’m the only user of Basecamp, it wouldn’t affect me one whit. Facebook, LinkedIn, and Twitter, on the other hand, rely almost exclusively network effects. Most people do not use Facebook or LinkedIn on the basis of superior features, they use them because all their friends/contacts/potential employees are there.

That’s why those companies operate on a free-to-use model: they derive network value from the non-paying customers. It’s part of why I think Craigslist has remained so successful: most of its listings are free, which keeps the network/marketplace intact. In the case of Dalton Caldwell’s App.net, he’s excluding from the network anyone who isn’t willing to pay for a realtime feed API service (read: almost everyone). That makes the network less valuable, even for those people who are paying for it, which results in fewer people willing to pay, and down the spiral it goes.

If Twitter (or App.net) could price-discriminate perfectly, charging nothing to the cheapskates, and a ton of money to those willing to pay it (in the real world, this often takes the form of freemium), then a paid model would work, but barring that, it collapses because the network, the real value, is missing.

Perhaps if App.net caters to a specific niche that requires highly available realtime messaging - something having to do with swing trading or something similar would be interesting - they can find success outside of the mainstream. But based on the network effects that Twitter (and by extension App.net) relies on, a mainstream play doesn’t seem viable.

I wish Dalton the best of luck, and I hope he proves me wrong.