I was out with some friends over the summer, one of whom is a college soccer coach, and we were talking about what it is that makes great teams great. I love talking to to coaches and people who have played for great coaches (just ask Ryan about how I always bug him for Coach K stories) -- they always seem to have the best social hacks to get people to work well together.

College teams can be particularly difficult to manage because the tenure is short and there's a lot of player turnover -- so the team dynamic is constantly being reset. It is similar in startups, where teams reshape and reform as they grow.

My coach friend described the process as "Form, Storm, Norm, and Perform". At the time, I took it to be another one of those witty and handy coach-isms, but, alas, it turns out this is an established group performance framework developed by psychology professor Bruce Tuckman back in 1965.

Anyway, I have been thinking about it a lot recently, as I see so many teams going through the various stages. For example, Plaid, which I mentioned last week, was acquired by Visa today for $5.3B -- a great product and from what I hear a really positive and effective team culture. Clearly in the "Perform" phase :-)

What I especially like about the framework is that it acknowledges the importance of, or at least the temporal existence of, the "storm" phase. The storm phase can be hard when you've never been through it before, because it contains conflict and you're not sure if it will end. But it does, and in the best situations, working through that is what enables you, and your team, to norm and perform.

Above all, what the framework reminds me of is that teamwork and success are about chemistry. Chemistry is hard to define, but it has a lot to do with trust. Trust in each other, trust in the vision, and trust in process. It is a beautiful thing when it comes together.

Discuss on Twitter is a WordPress plugin that uses Twitter as the commenting system for your blog.

I've been developing it over the past few weeks along with Fred Wilson and Kirk Love as part of the launch of AVC 3.0 which went up yesterday. It's currently live on this blog, as well as AVC and also USV.com.

The idea is pretty simple: whenever you publish a new post, the post is auto-tweeted to your account. Then, you get a "Discuss on Twitter" button on your post, which will prompt a reply to the auto-tweet, as well as a "View Discussions" button which will link to the Twitter thread.

Twitter is a natural place to discuss long-form content on the web, and moving blog commenting to Twitter can both engage people in a natural way and help expand distribution and reach.

I should pause here and note that the core WP/Twitter functionality is provided by the amazing WP to Twitter plugin created by Joe Dolson. Joe's plugin does all the heavy lifting of authenticating with twitter and teeing up the tweets. It's really configurable and nice. Thank you Joe.

This is alpha software (Version 0.3 to be precise), so we are still testing and tuning. Totally open to feedback, so please send any ideas on github or in the comments.... er, on Twitter

Enjoy!

A few years ago I wrote about one of my favorite product sayings: "Half, Not Half-Assed", which comes from my favorite book on product development & teamwork, Getting Real (from the team behind Basecamp). I actually first got hooked into this thinking when I saw one of the Basecamp founders, Jason Fried, talk at a web design conference back in 2004 and it was a pivotal moment in my career (thanks Jason!).

I was reminded of that idea today, as I was working on a side project (an internal tool for USV). In this particular case, we had built a version of this tool a few years ago, but basically abandoned it because we could not build a good workflow around it. Looking back, the diagnosis on why was that it was overbuilt: too complicated, too much. And implemented in a way such that we just didn't bother to keep up with it. In other words, half-assed, rather than "half".

In version 2.0, I cut the scope and complexity down by 90%. Every new piece of detail went through the filter of "what's the trade-off between nice to have and this will be sustainable to maintain." But further, it was really easy to make design decisions because most of them had been made already before, and the key technical challenges had been worked through.

So it was very easy to build version 2 in a lighter, leaner, meaner way.

I am a big fan of iterating and taking things step by step. And in this particular case, it was that process that led the version 1 to be really interesting, but also a bit meandering, bloated and off target. Because the iterations were also an exploration.

This time, It felt good to do an iteration that started over, but from the place of wisdom of having built it all once before. Iterating from scratch. Very satisfying in its simplicity, and hopefully it will be on-target this time!