Retrospectives are healthy. Retrospectives are vital for improving our work life. Without them, we wouldn't have a systematic way of examining what needs improving and what we might do about it.
Even so, too often I see teams becoming so excited with the 57 things they've just discovered could do with improvement that they want to try them all at once. Beware greed in retrospectives - do your best to pick just one. Search for the most valuable improvement opportunity, design some experiments for countermeasures, and focus on that one improvement. Create a measurement system such that the results can be analysed with as much clarity as possible. Let the data speak.
Only then focus on the next most pressing constraint and repeat the process of root cause analysis, countermeasure design, experimentation and adaptation. We owe it to ourselves to be very reserved with our investments - we only have a very limited capacity for improvement, we should spend it wisely, on the item that we expect will lead to the best improvement.