Thinking concurrency is always faster
Mistake
Assuming that a concurrent solution is always faster than a sequential one. If the workload we are parallelizing is too small, the overhead of creating and scheduling goroutines quickly becomes the bottle-neck.
Fix
Create a hybrid approach that only leverages goroutines when the elements being processed exceeds some threshold.