| Elastic works similarly, though not to the extent that was described in that Gumroad post. It also varies a bit by team, but I've worked on many teams over the last 5.5 years here, so I have a decent perspective on what's normal and how things evolved to this point. Engineering is distributed around the world, so it happens in a highly asynchronous way centered around GitHub issues, the vast majority of which are in public repos. Slack and Zoom are used, but if they're used to make decisions, the recording is saved for others to consume and the decision is documented on GitHub. Meetings are discouraged, but not non-existent. To give some context, I'm a manager of two teams and this week I had 4.5 hours of meetings (including 1:1s), which is pretty normal. When I was an independent contributor on a single team, I often had weeks where I had a single 30 minute meeting. In practice today, I suspect an engineer at Elastic will spend an average of ~2 hours a week in a meeting, with a few spending a great deal more than that and others spending less. This culture is demonstrated top-down and has been a common thread from the early days, through the IPO, and continues today. Edit: We also have a general philosophy of features being done when they're done rather than when we reach some arbitrary date. This doesn't mean we don't have timelines (we have ~2 month long release cycles), but if we can pair down scope to make a release, we will, and if we can't do it then we'll just move the feature to the next release instead. We've codified a lot of the philosophy that feeds into this workflow here: https://www.elastic.co/about/our-source-code |