Feature Requests are merely a key value pair and you need to find the key
Aug 30, 2026
Thoughts come from https://lalitm.com/post/find-problems-staff-engineer/
(you may also see my Readwise annotations here if you can’t commit to reading the whole article but I recommend it)
This article reminds me of the finer arts of Product Management genjutsu that I was just starting to learn before I moved into my current career quest
Basically:
- I think of “feature requests” as merely a paired value with a “problem {stakeholder} meets in their day to day”
- Early in your career, yes you are rewarded for hopping to the “feature request” and doing it Fast + Cheap + Maintanable
- Importantly, all you see is the feature requests, so you can get in the bad habit of thinking that all that exists are feature requests
- But over time you should really start to “stack trace” the request up to the set of paired Problems and start picking and choosing from those
Yes. This will take longer, and you will be tempted to think (or other people may tell you) that your shipping velocity has slowed but in doing this, you actually find the real problems, and if you’re doing it right you find that you can fix/pre-empt MULTIPLE feature requests by solving one problem
- And lastly, if you are at like 2-3 years into your career and you’re still just implementing the feature request as asked it kind of becomes an anti-feature