[csallen]

No, I'm saying that the value of Twitter isn't the product itself, it's the people using it. I don't get on Twitter because of the allure of a 140-character-limited textbox. I get on Twitter because it's the only place I can receive regular personal updates from my idols, role-models, and colleagues, and even engage them in conversation. The more people on Twitter, the more valuable it becomes to me.

You can build a rival service if you want, but nobody will use it until it provides value, it won't provide value until people use it.

[justinhj]

By that logic everybody would still be using Yahoo search

[csallen]

You're still missing the point. Search doesn't have lock-in via network effects. If all my friends use Yahoo search, it doesn't make my Yahoo search experience any better.

[justinhj]

Then how about myspace?

[csallen]

MySpace was a thriving site with powerful network effects. However, Facebook was started just one month later, not years and years later. By contrast, attempts to dethrone Twitter come 6+ years after Twitter's founding.

Also, MySpace was a technical fiasco. (Although it wasn't quite as bad as Friendster, which was so slow it was unusable.) Facebook and Twitter are never going to allow pages to become painfully slow, have unreadable backgrounds, or blare music automatically.

Thirdly, Facebook's network effects were more powerful than MySpace's. Their school-based distribution strategy guaranteed that your relevant friends would be on the service. Features like "events" gave everyone a reason to join. And, ultimately, Facebook appeals to everyone age 4 to 100+. MySpace's main appeal was to music-loving teenagers. The more people you appeal to, the stronger the network effects.

Long-story short, it's not impossible to beat Twitter at its own game. But it's really, really fucking hard. It's not simply a matter of building a slightly better Twitter. Likely, Twitter would have to fuck up their basic user experience in a major way, AND you would have to devise an extremely superior user acquisition model.