Efficiency becomes useless when applied to work that never needed doing.

Doing an unnecessary thing quickly is still doing an unnecessary thing. Improvement to process is satisfying because progress is measurable, but no amount of speed rescues work that should not exist, and streamlining it usually entrenches it by making it cheaper to keep.

Before optimising a recurring task, ask what happens if it simply stops. Sometimes the honest answer is that someone would object, and that is worth knowing. Sometimes nothing happens, and you have found a much larger saving than efficiency would have offered.