Virtual First Organizations

I listened to the Source Code Podcast titled "How Dropbox figured out the future of work". They now have people working remotely, people in the corporate office in San Francisco, and some on the hybrid work from home, sometimes in the office arrangement.

Dropbox blog has a quick introduction, and this blog post details how they are implementing it at Dropbox. I like how they emphasize four hours of collaboration a day with the remainder reserved for deep work.

I have found that asynchronous interaction is the ideal solution for any group decision. A "quick" half hour meeting of six people burns three person-hours. A quick two page synopsis of an idea may take a person an hour to refine and everyone else four to five minutes to read. We already saved fifty percent of the time. If you publish a paper in an environment that supports comments, this may be less.

Written communication tends to encourage additional thought, you do not have to toss out an idea before it has had time to mature. Your audience does not have a ten minute window to consider and question your idea. Potential downsides include tendencies to abuse people online, mitigated by real world identities. Online collaboration must be between colleagues, not users.

A written record of ideas, discussions, and conclusions allows people to return to the initial goal, alternatives and the final result. Assume that each discussion thread has a unique URI that can be referenced and provided in the final documentation so that readers can find additional details. Meetings, even well intentioned meeting notes do not provide this level of detail.

When time is treated as a lake rather than a river people can use their deep thought time to provide cogent criticism on papers. A system could also provide up-voting to surface popular topics to a wider audience. Unlike a platform such as Reddit, we can have expiration dates on our ideas, or request for comment (RFC) posts. Many collaboration platforms allow the author to close the issue’s RFC.

Traveling to the office is seldom an experience that myself (or anyone) would consider a useful endeavor. Let us discount that since it is only supposed to be forty minutes of your work day ^[1] everyone in attendance at the office is fresh and ready to collaborate on the day’s most important issues. Are they looking forward to discussing your most pressing issues or something that may affect their performance?

The agile method of dealing with uncertainty is the spike. Time box an exploratory phase to look into the magnitude to the problem. The Dropbox solution is to reserve four hours of your working day for collaborating, so maybe half of this is actually working with other developers, half with designers and other roles to get determine what need to be done. That leaves around two to four hours for development ^[2].

I find that if I reserve an hour for spelunking into the group conversations, then the other five hours that I can really focus can be spent on delivering customer value. I assume two hours of my eight hour billable day will be spent on other tasks such as time reporting, "agile" ceremonies, check-ins. Most weeks I average twelve hours on overhead work, twelve on collaborative work, and sixteen on heads down, value work. So about three quarters of my week could provide value.

Most of the twelve hours I spend in ceremonies (required meetings) could be spent with either stakeholders or their agents ^[3]. When the same co-location rigor is imposed on teams, you should expect output do decline due to the added overhead of simultaneous communication over synchronous mediums. "Zoom" meetings are observed to be "the same two people talking" for most of the Zoom population. Please respond via email in the link below if you have a different opinion.