Do as I say, not as I did

lovely article, again thanks to Found|Read

Last September, Ev Williams gave a speech about some of the mistakes he made as CEO of Odeo.

Since then, a lot has happened. We turned out to totally wrong about one thing: "So what's he doing to fix these mistakes? Not refunding the VCs their investment, that's for sure." That's exactly what Williams did just a month later , refunding his VCs and angels their $5 million stake.

Williams' fortunes have changed radically since then, as Obvious Corp, which he formed to buy Odeo, has also developed the smash hit web product of the season, a casual blogging tool called Twitter.

Williams said he created Obvious to pioneer "a new model for building and running web products," one that uses cheap and rapid development to test an idea before turning it into a company. So far, it appears to be working.

This week, Williams indulged us by reviewing the list of Odeo-screwups we covered last fall and, importantly, shared with us what he's doing differently this time at Twitter.

Mistake #1: "Trying to build too much"
Retake: Where Odeo had a mess of products, Twitter is singularly focused on the short-form shout-out. Tell your friends what it is that you're doing in 140 characters or less. The thrift and simplicity of Twitter posts are comparable to the site, which simply takes the messages from SMS, IM, web form, or third-party application and sends them back out. Says Williams via email, " It does very little. (In a good way.)"

Mistake # 2: We weren't the target users of our product
Remake: The makers of Odeo weren't podcasters and didn't listen to many podcasts themselves, so they lacked intuition for their users' needs. Twitter is the opposite, according to Williams, because it's a product his team uses and loves. "[Obvious employee] Jack Dorsey introduced the idea of Twitter to us, because he'd been wanting it for a long time. We built a prototype and started using it internally and, based on that, decided to invest further."

Mistake # 3: "Not adjusting fast enough"
Remake: Odeo couldn't compete when Apple introduced a competitor, but Twitter has tried to be more agile. Rather than stay bound to long-term strategic visions, Twitter has made many adjustments to its product over the last several months, aiding its astronomical growth this March.

Says Williams, "We didn't have the formula right for Twitter at first. We liked the app, but for the first few months, it wasn't clicking with users. We changed the positioning, the relationship model, and other things until it started working. I think we could have been faster, but we got there. Now we're trying to adjust to the scaling requirements."

Mistake # 4: "Raising too much money too early"
Remake: Williams' new theory is "Some things are perfectly worthwhile but don't need to be a company" in the "hits-driven" consumer web business , where anything less than a 45-degree trend on the growth chart considered flat-lining. Due to the pressure of responsibility to its funders, Odeo had to be a company before it had proved it was a successful product. Twitter hasn't raised any outside funding yet, though Williams says "It's likely we'll need to before long, but we're past the point where I think it would be too early."

Mistake # 5: "Not listening to my gut"
Remake: Williams says, "This has a lot to do with who I'm working with, what we're working on, raising money, etc. Safe to say, we're doing better in all departments."