The Checklist Manifesto
Human memory is unreliable and costly to fill and alter. Pilots diligently use checklists in both normal and exceptional circumstances to ensure consistency and quality. Atul Gawande’s The Checklist Manifesto – How to get things right (recommended to me by MB and JS) is the engaging story about their recent adoption in medicine. It discusses one aspect of creating reliable and effective processes (something The E-Myth Revisited explores brilliantly more broadly).
- Develop checklists for routine tasks, in particular those performed under pressure
- Choose between Do-Confirm and Read-Do form
- Checklists should usually have 5-9 steps
- Assume people know what they are doing
- Use several checklists for complex processes if necessary
- Always test checklists in practice and tweak as necessary
- Checklists seem to provide a process that teams bond around
- Unaquianted teams benefit significantly from name/role introductions
Checklists “automate” processes performed by humans (in contrast to full machine/computer automation). The benefits – consistency, reliability, and quality – are the same however. If something on a checklist could be completely automated it should be, but often the information communicated to a human is the key benefit.
The book left me curious. The case for checklists in companies and institutions is clear (and exciting – again, see The E-Myth Revisited), but how effective could personal checklists be? Perhaps for the start and end of the day? Leaving the flat? Packing? Would checklists for design (for example, requirement lists to ensure a practical kitchen design) work? (I assume these exist but have not gained wide adoption.) How do declarative (do-confirm) and procedural (read-do) checklists differ in essence? How much does development and adoption cost in effort and overhead? Questions I hope to explore over the coming months.