Hey there! I'm Nat Bennett, and you're reading Simpler Machines, a newsletter about how to build better (and simpler!) software.
In my perfect world we'd have about 900% more designers. They'd be everywhere. Every team would have someone whose full time job was to facilitate user research, whether that team had a graphical user interface to draw or not.
In the real world I often have to do my own research, and my own design. This is especially true when I'm working in an operations or a platform capacity. One of the reasons the "platform" play is hard is that you have to approach design the same way you'd approach it with any product, but the teams developing an internal platform rarely have that skillset.
You end up seeing a lot of well-meaning disasters, where the platform team did a lot of research, but the research was all surveys and focus groups, and they tricked themselves in all the ways that surveys and focus groups let you trick yourself.
If you're also in this situation – you're some kind of backend or infrastructure team, and you're building something that has customers who will need to adopt it – here are some things you can do that will help your design process a lot. And hopefully help you avoid spending a year building something that no one at your company will use.
- Keep a list of "Jobs to Be Done"
- Use "dump and sorts" to generate initial ideas
- Watch people try to do those jobs and take notes on sticky notes
- Bring two people along to every interview
- When you interview people, ask them to tell stories
- Use affinity mapping to process notes
- Use bug reports and support requests to learn more about user needs
- "Slack debrief" weekly, with your whole team. Review every question or problem reported by someone using your tools.
- Try to fix everything. If one person reports a problem, there are probably a dozen more who are bothered by it but didn't tell you.
- If you can, fix problems by updating documentation
- Avoid surveys
- Read "Just Enough Research"
- Read "Humble Inquiry"