Coaching works - and what are we calling success?
"A good coach can change a game. A great coach can change a life"
John Wooden
When someone asks me whether coaching works, my first response is another question: What would “working” look like here?
Recently, I reviewed my own professional development using a reflective model from a mentoring course. I did not begin with the number of qualifications I had completed, there are plenty. I looked at subtler changes in me: how urgency feels in my body during a session; whether my questions strengthen the client’s agency; how I hold silence; and how I recognise the “just right” place at an ethical edge.
Those are meaningful changes that show up in my practice. They are also difficult to fit into a satisfaction score.
Coaching research gives us good reason for confidence and also for caution. A 2023 meta-analysis by Erik de Haan and Viktor Nilsson examined 37 randomised controlled studies, representing 39 samples and 2,528 participants. It estimated a moderate overall effect for workplace coaching across leadership and personal outcomes. Yet the authors also found indications of publication bias, and effects were larger when outcomes were self-reported rather than observed.
Another meta-analysis of 20 controlled studies found stronger effects on behaviour than on attitudes or relatively stable personal characteristics. Coaching was associated with gains in areas including goal attainment, self-efficacy, psychological capital and resilience. However, the researchers could not assess many specific organisational outcomes or the effects on people around the coachee, because how do we measure success?
For a senior executive, success might be a different strategic decision, more honest challenge in the room or a change in leadership behaviour noticed by others. For a young professional, it may be greater agency and a practical experiment. For a coach or mentor, it may include what the client does differently - not simply what the client says they enjoyed. For a HR or a L&D Professional, it may involve capability, mobility or retention, while respecting that coaching conversations are confidential and cannot be mined for organisational intelligence.
The buyer, sponsor, coach and client may all hold different pictures of success. If those pictures remain unspoken, evaluation becomes political: the most powerful stakeholder gets to decide whether the work mattered.
A four-view success I contract
Before a coaching engagement, here are some questions I place at the table:
Client: What change would be meaningful to you?
Behaviour: What might you do differently if progress occurs?
Context: What in the environment could support or obstruct that change?
Evidence: What signs can be noticed without breaching confidentiality?
Return to these questions during the engagement. Do not wait until the final survey to discover that everyone meant something different by “impact”.
Coaching is not therapy, and evaluation should not invite disclosure of private material. Nor should coaching become covert performance management. The client needs clarity about what is confidential, what may be reported and who owns each measure.
I believe coaching works. However, as a senior practitioner I am not hiding behind that sentence. I ask: for whom, towards what, in which context - and according to whose evidence?
Research that inspired this article
de Haan, E. and Nilsson, V. O. (2023), What Can We Know about the Effectiveness of Coaching? A Meta-Analysis Based Only on Randomized Controlled Trials, Academy of Management Learning & Education.
