This is a very raw thought that I’ve been exploring with my friend Panos during a call.

I had the thesis that optimising for readers is usually the way to go, let’s examine that

Time spent writing >=< readers * time spent reading

5 min read 5 readers

25 min spent writing max?

On the eyes of the company: Time spent writing + readers * time spent reading = 50min

But if you spend 5 min optimizing the read down to 2 min 30 min for you 10 for them 40 for the company!

This is all pretty simple nice math


Simple example: Technical onepager decision record 15 min write 4 readers 2 min read time 23min company time

Expected outcome: Inform, provide discussion platform for async feedback

Alternatively, that time would invested in:

  • 4 person 30m meeting
  • Directly building the thing
    • Decision is bad:
      • Build time of 3h is gone
      • Significant potential risks
      • Has to be rolled back, more time
      • New decision has to be made
    • Decision is good:
      • Stakeholders have not been informed, their systems are affected
        • How long will discovery take?
        • Fastest path:
          • 5min Slack+Confluence search
          • 5min writing to an eng
          • 10 mins of bouncing around folks, end up in the correct team
          • ? chance that Pan or the approver are currently available
          • Wall clock time is at least 8h probably due to async!
        • Slowest path:
          • Discovery is days or never

Optimal outcome: Someone shares info, saving a future project a lot of time


What about reading comprehension?