Yes, moderately, and more reliably than most things sold alongside it. The honest version comes with a caveat the industry rarely repeats: the studies with proper control groups report smaller effects than the studies without them. Written by someone who sells coaching, which is exactly why the caveat is here.
Theeboom, Beersma & van Vianen, 2014
The Journal of Positive Psychology, 9(1)
The most-cited meta-analysis on this question screened 107 articles and pooled 18 studies of coaching in organizational settings. Its conclusion was that coaching is an effective tool for improving how individuals function at work.
Broken down by outcome, the effects were:
| Outcome | Effect size |
|---|---|
| Goal attainment | .74 |
| Performance and skills | .60 |
| Work and career attitudes | .54 |
| Wellbeing | .46 |
| Coping | .43 |
For context on what those numbers mean: an effect of around .5 is conventionally read as moderate. This is a real result. It is not the order of magnitude implied by most coaching marketing.
A second meta-analysis — Jones, Woods and Guillaume, published in the Journal of Occupational and Organizational Psychology in 2016 — examined learning and performance outcomes from workplace coaching along with moderators of those effects. It reaches broadly compatible conclusions. Its specific figures are behind a paywall and are therefore not quoted here rather than repeated second-hand.
This is stated plainly by the meta-analysis authors: studies that included a control group displayed smaller effects than studies that did not. Their reading is that control groups do a better job of accounting for sources of bias — including people simply improving over time, which they would have done anyway.
How much? Nobody can tell you precisely, and that is the point. The effect is real and it survives controls — it just gets smaller. Any coach quoting you the uncontrolled numbers without mentioning this is quoting the flattering half of a finding.
Almost everything is self-reported, and there is almost no longitudinal work. Both are named as limitations by the authors themselves. Someone who has just spent six months and a significant budget on a coach is not a neutral judge of whether the coaching worked, and effort justification is a well-documented human tendency rather than an accusation.
Goal attainment — the largest effect in the table — is the most vulnerable outcome of the set. In most coaching engagements, coach and client set the goals together and then jointly assess whether they were met. That is a reasonable way to run an engagement and a weak way to measure one. This is a reading of the design rather than a finding reported in the paper, and it is offered as such.
Because a moderate, repeatedly replicated effect is a good deal more than most items in the leadership-development catalogue can claim, and because the honest version is still a strong offer.
It also suggests how to run an engagement so that the result is less dependent on the coachee's own verdict:
Measure something the client does not score themselves. This practice baselines psychological capital among the leader's people and maps the advice network around them — both sourced from followers rather than from the leader. The leader's impression of their own progress is interesting. It is not the evidence.
Re-measure at the end, and report it even when nothing moved. A null result is information you paid for. It usually means the constraint was structural rather than individual, which is worth knowing in month six and worth a great deal more in month one.
Say in advance what would count as failure. If no outcome could ever disconfirm the engagement, the engagement is not being measured.
Tell you coaching will definitely work for you. The literature does not support that sentence for anyone, and a coach who says it about themselves is telling you something about their relationship with evidence rather than about their results.
Theeboom, T., Beersma, B., & van Vianen, A. E. M. (2014). Does coaching work? A meta-analysis on the effects of coaching on individual level outcomes in an organizational context. The Journal of Positive Psychology, 9(1), 1–18. Source of the effect sizes, the control-group finding, and the stated limitations.
Jones, R. J., Woods, S. A., & Guillaume, Y. R. F. (2016). The effectiveness of workplace coaching: A meta-analysis of learning and performance outcomes from coaching. Journal of Occupational and Organizational Psychology, 89(2).
Avey, J. B., Reichard, R. J., Luthans, F., & Mhatre, K. H. (2011). Meta-analysis of the impact of positive psychological capital on employee attitudes, behaviors, and performance. Human Resource Development Quarterly, 22(2).
Reviewed September 2026. Where a paper was not obtainable in full, its findings are named but not quantified here. If something on this page is wrong, write and say so — corrections are published.
The instrument most coaching engagements open with, and the one with the weakest numbers.
Strong · licence requiredWhat the coaching is trying to move, and the one validated measure in the toolkit.
Real construct · contested measurementWhat the coaching develops toward, and where the practice’s own research sits.
Every construct this practice uses gets the same treatment, including the ones that make it look good and the ones that do not.