Insight is satisfying and easily mistaken for progress. Someone leaves a session with a clear explanation for a twenty-year pattern and genuinely feels something has changed. Often nothing has.
The useful distinction is between intellectual insight — I know why I do this — and emotional insight, where the understanding is accompanied by feeling the corresponding affect. The first can even become a defence: intellectualisation lets you talk about pain without touching it, and a therapy that settles there can run for years without movement.
Common-factors research shows why this matters. What predicts outcome is the therapeutic alliance and practice between sessions rather than the number of realisations. Behaviour change often produces insight rather than the reverse — hence behavioural therapies' emphasis on experiment before explanation.
None of which makes it useless. In a recurring relational problem, understanding the pattern is what turns a series of accidents into something you can work on. It is simply the beginning rather than the end.