09 May 2026 · 4 min read
The useful middle
Working between product, design, and engineering without turning the overlap into another job title
The most useful work often happens before a task belongs cleanly to one discipline. The problem is still changing, the interface is still teaching us what the product should be, and implementation details are exposing decisions that were previously invisible.
That middle can look inefficient from the outside because it does not produce tidy handoffs. It produces understanding instead.
Stay close to the material
Design improves when it understands the constraints of the code. Engineering improves when it can feel the intention behind an interaction. Product decisions improve when both can be tested against something real instead of discussed in the abstract.
The goal is not for everyone to do every job. It is to keep the distance between a decision and its consequences short enough that the team can still learn from it.
Make the handoff smaller
A shared prototype, a written decision, or ten minutes pairing on the difficult part can replace a large amount of process. The best collaboration often feels less like coordination and more like keeping the work in one room for a little longer.