3 ms·
I got to work for a few years with pivotal labs in the Bay Area. They had a policy called “just don’t have f&$@ing meetings.” At first I thought this was a bit
by trcollinson 8y ago
I got to work for a few years with pivotal labs in the Bay Area. They had a policy called “just don’t have f&$@ing meetings.” At first I thought this was a bit crazy. How could we work without meetings? Then I realized that product owners, actual 10 minute stand ups, and a 30 minute weekly planning meeting was all we really needed to make amazing products. I’ve keep that philosophy ever since and it has worked very well in big and small organizations.
- dasmoth 8y agoPivotal is clearly a style that works well for some people. But couldn’t you categorise the constant pairing as a form of ongoin meeting? It doesn’t seem like the place to go if you want to be left alone...
- trcollinson 8y agoI absolutely agree with you that it is not the sort of place you want to go if you'd like to be left alone. No doubt there. However, I would say that constantly pairing really is nothing like an ongoing meeting either. A really good paired programming environment, if you like that sort of thing and work well in that sort of environment, is very effective for writing and deploying fantastic products and features. I find meetings are very not effective for writing and deploying fantastic products and features. I'd really rather not say there is a "Pivotal" style as much as an Extreme Programming paradigm. Extreme Programming says that an environment is better when you work normal hours. It states that an environment is better when there are fewer meetings. It states that an environment is better when you try paired programming, test driven development, continuous processes, simple design, good communication, etc etc etc. I think all of these things are very good. I also believe a few of these principles don't work well for specific individuals. There are principles that effect the business (fewer meetings, sustainable pace, continuous practices). So far I have always seen those principles as good. There are also principles that effect the individuals (paired programming, test driven development, close proximity for communication). I have seen those not work well for some individuals and should be adjusted to your individual employees.
- wdr1 8y agoIs it fair to say no remote teams were involved & most people sat close together?
- trcollinson 8y agoAt Pivotal Labs, yes. However, I work in a completely remote environment right now. We have an office for the business but no one in the product engineering teams or tech teams work in the office (or even in the same state or country as the office). We have a written daily roundup (like a 10 minute standup but it's written and it's asynchronous), a 30 minute retrospective once a week, and a 30 minute planning meeting once per week. That is the grand total of all meetings for the group as a whole. Individuals may meet and work together as needed. It works extremely well for remote work.
- TechieKid 8y ago> We have a written daily roundup (like a 10 minute standup but it's written and it's asynchronous) How does this help with discovery of issues that other folks on the team may have experience/suggestions solving? I.e. do you require everyone to read the daily round up?
- trcollinson 8y agoTo be frank, I expect people to work as a team. I expect that people will read the daily round ups, yes and they do. We do have closer to synchronous communication channels (Slack) which people can reach out on for help. And they absolutely do all the time. Our teams members are quite close and are very willing to say when they have a problem and to help one another when they can with problems. We have an absolutely open environment. It's a side effect of our process.