Workplace coaching – what’s the verdict?

Key points

  • Coaching can be an effective way to increase individual performance.
  • Online coaching can be just as effective as face to face coaching.
  • More coaching does not mean better coaching.
  • Internal coaches are more effective than external coaches.


Workplace coaching is now a popular way of delivering ‘one to one’ learning and development. Coaching is a helping relationship with an emphasis on active learning. At work, its focus on the immediate needs of the learner. This allows the learner the chance to apply new knowledge and improve performance almost instantly.

In 2012 the International Coaching Federation estimated that the coaching industry, which includes workplace coaching as well as personal and life coaching, was worth $2bn annually. Just over half of coaches, around 28,000 people, were workplace coaches, specialising in executive, business or leadership coaching. These are big numbers. But does workplace coaching work, and if so, how can businesses get maximum value from their coaching spend?

A 2015 meta-analysis of seventeen rigorous studies of workplace coaching by Rebecca Jones, Stephen Woods and Yves Guillaume provides some, but by no means all, of the answers. They found that workplace coaching has different effects on different aspects of individual learning and development. In terms of the way this type of coaching is delivered, we may need to revise some some cherished assumptions. What’s more, some types of feedback can undo the positive effects of workplace coaching.


Coaching outcomes – Good, Better, Best

Perhaps the most important finding from this research is that workplace coaching works. In order to assess how effective workplace coaching is, Jones and colleagues grouped learning outcomes into

  • skill-based: acquiring and automating new behaviours such as technical or leadership skills
  • affective: changing attitude or motivation through improved well-being, self belief or work satisfaction
  • results: improving individual performance

Looking in detail at these three outcomes of workplace coaching – the researchers found a small to moderate effect of coaching on skill development and a slightly bigger effect on mood and motivation , but the biggest effect was on performance improvement.



The message is clear. To get maximum value from coaching, focus on improving individual performance.


A little light mythbusting

The researchers also compared the relative effectiveness of how coaching was delivered; what they called practice moderators. Surprisingly, they found:

  • The coach does not need to be in the room for the client to benefit. There is no significant difference between face to face coaching and blended (face to face, online and phone based) coaching.
  • More coaching does not mean better coaching. Neither the number of sessions nor the length of the coaching programme appeared to make a significant difference to the outcomes.
  • Internal coaches beat external coaches. Coaches who work in the business, but who are not in a line management relationship with the client, are more effective than coaches from outside the business. The researchers noted that this might be due to internal coaches having a better cultural and organisational understanding than people outside the business. This understanding might give internal coaches more knowledge about how the individual can be more productive and effective.
  • NOT providing 360 degree feedback has a large positive effect on coaching outcomes.  This could be due to the distracting and potentially conflicting nature of feedback from different people, especially if they are using different criteria to evaluate success.

These findings may be unpopular with large coaching organisations, but they send out a clear message to the businesses that already have, or are considering, large and expensive face to face external coaching programmes.



Workplace coaching – good practice guidelines

The verdict from this research is clear. When it comes to workplace coaching, best results are likely to come from focusing on improving individual performance.

In terms of how workplace coaching is delivered, these guidelines may not apply to all businesses in all situations. But if you are involved in designing or delivering workplace coaching, then bear in mind that:

  1. Coaching doesn’t need to happen face to face. You can save time and money using online coaching.
  2. Internal coaches are more effective than external coaches. You may want, for example, to consider developing a coaching culture rather than outsourcing coaching to external organisations or individuals.
  3. Neither the number of coaching sessions nor the length of coaching programme has an appreciable effect on outcomes. Setting clear coaching objectives will help both coach and client to know when they are done.
  4. 360 degree feedback can undo some of the benefits of coaching. Coaching programmes that use methods such as 360 degree feedback could improve their effectiveness by dropping this time-, energy- and cost consuming process.

Unanswered questions

While these findings provide much needed clarity in relation to workplace coaching, there are still a number of important issues that this study does not cover, such as:

  • What are the preconditions for successful coaching?
  • What are the most effective coaching approaches? For example, is a short face to face session with an internal coach better than a longer online session with an external coach?
  • Do some of the widely used coaching models, such as GROW, work better than others?

Grover and Furnham’s (2016) systematic review of workplace coaching contains a partial answer to the preconditions for successful coaching. They suggest that the client’s readiness to be coached and the quality of the coaching relationship are likely to have a more significant impact than client expectations and the specific theories and techniques used. However, we cannot draw any firm conclusions until further high quality research is available.



Trustworthiness score

This study only included research with control groups or before-and-after measures. Due to the strict selection criteria and the efforts of the researchers we are able to give this meta-analysis a trustworthiness score of 90%. This means there is only a 10% chance that the findings are due to other factors, or random chance. Given that it may well be several years before there are sufficient high quality studies to update this analysis, we should recognise that this is the best available evidence on the effects of workplace coaching.

Learn how we critically appraise studies to assign them a Trustworthiness Score.

We aim to provide you only the best available scientific evidence to inform your decisions.


Did you like this evidence summary? Share it with your network by clicking on the buttons below!

Follow us on LinkedIn, Twitter and subscribe to our newsletter to receive the most trustworthy scientific research summarized in less than 1000 words!


References:

Rebecca J. Jones, Stephen A. Woods and Yves R. F. Guillaume (2015). The effectiveness of workplace coaching: A meta-analysis of learning and performance outcomes from coaching. Journal of Occupational and Organizational Psychology, ISSN Online: 2044-8325.

Grover S. and Furnham A. (2016). Coaching as a Developmental Intervention in Organisations: A Systematic Review of Its Effectiveness and the Mechanisms Underlying It. PLoSONE 11(7): e0159137.doi:10.1371/journal.pone.0159137.

Leave a Reply

5 comments

  1. I read the original papers cited and I am astonished by the conclusions you and Jones et. al; reach. When I read about ‘learning styles’ and the ‘Kolb cycle’ in the paper, that is enough proof for me that the authors are too uncritical and uninformed because they have been debunked as urban myths. Jones et al. consider it sufficient as a ‘methodological’ comparison to compare face-to-face coaching with telephone, videophone or internet (they call it e-coaching). Moreover, of the 17 studies, they accept that 12 don’t even mention the methodology used. It could be bullshit methodologies like NLP, psychoanalysis or Transactional Analysis (the ICF accepts almost everything). O, and don’t talk to me about ‘working alliance’. The researchers even fail to see that if there is no difference between face-to-face and ‘e-coaching, the working alliance is unlikely to be important. Indeed, research on online CBT therapy has demonstrated that using online homeworkassigments combined only with e-mail exchange has the same effect as face-to-face therapy. Working alliance is overrated and probably no more than a placebo effect. On the other hand, hundreds of meta-analyses have demonstrated the superiority of CBT (theory, protocol and techniques) for most anxiety problems (prevalent at work), anger problems (often leading to leader dominant or aggressive behavior), depression etc. It works with kids, adolescents and adults. So CBT-informed coaching is likely to work, as research by Adam Grant has indicated. Whether other coaching approaches work for behavioral issues is an unanswered question. The Grover and Furnham study is much more critical and prudent. They support my idea to look more at CBT informed coaching, and they warn for confounded effects in the Jones study. They point to problems such as lack of control groups, cross-sectional study design, underpowered studies, lack of longitudinal assessment etc. So not so fast please, coaching is still in the context of “exploration” and not yet in the “context of proof” and far from the context of replication.
    Patrick Vermeren
    president of EvidenceBasedHRM in Belgium (not for profit)

    • Mark Seabright

      Dear Patrick,
      Thank you for your comments. You raise a number of useful questions, which I hope to answer here.
      1. Just to be clear, the key points and good practice guidelines in my summary were drawn from the research findings. The conclusions were the work of the researchers. The purpose of the summary is to report the findings accurately in a form that is digestible for decision makers, and then to assess the reliability of the findings by looking at the quality of the study as a whole.
      2. If I understand you correctly, your references to learning styles and the ‘Kolb cycle’ occur in the literature review of Workplace Coaching on page three of the Jones et al article. Whether or not these have been ‘debunked as urban myths’, it would seem a little harsh to dismiss the findings of the article and the research methodology which underpins it purely for a small element of the discussion of the literature on the subject.
      3. Regarding the ‘methodological’ comparison, I fear that there may be some confusion between the ‘Coaching Technique’ and ‘Research Design’ categories in Table 2 (page 12) of the Jones et al article.
      The Research Design section on page seven clearly states that the researchers only used high quality within subjects (before and after) and between subjects (control group) studies. This helps to explain why the summary provides a high trustworthiness rating.
      The Coaching Technique, as you rightly say, is not specified in most cases, according to the researchers. However, Coaching Technique was not assessed in this study, so this is a somewhat moot point.
      4. Following on from point 3 above, I am sure that there is evidence that CBT based approaches work in a number of contexts, but it is important not to confuse the Coaching Technique, whether CBT, GROW model, NLP etc. with the Format of Coaching, i.e. face to face, video, phone or email. I am pleased to hear you say that CBT therapy has been effective with a combination of online homework assignments and email exchanges. I agree that CBT informed coaching is likely to work. What we need is a number of high quality studies to examine the parameters, conditions and impact of both CBT and other approaches. This was the point I was attempting to address in the third of my Unanswered Questions.
      5. The confounded effects referred to by Grover and Furnham in relation to the Jones et al study related to the fact that ‘all of the studies that utilised multi-touch feedback also used an external coach’, thus opening up the possibility that it may have been an external coach, rather than the feedback which is at fault. It is impossible to say whether or not this might be the case, but it does nonetheless suggest caution in the use of feedback in coaching programmes. My scope to explore these issues in an article of a thousand words is necessarily limited. That is why I mentioned in point four of my good practice guidelines that feedback can cause a problem. I would also point of that the overall trustworthiness score of 90% makes it clear that there is room for other explanations of the findings.
      6. The Grover and Furnham comments regarding ‘lack of control groups, cross-sectional study design’ and so on are made about coaching research in general, and not the Jones et al study in particular (see Grover and Furnham P7).
      I agree that coaching as a discipline is still at an early stage of development. Much more research is required into all aspects of coaching. However the challenges of controlling for extraneous and confounding variables may deter all but the bravest researchers. In the meantime, the findings from this meta analysis remain the best available, but perhaps not yet the best possible, evidence.
      I hope this addresses your issues. Thank you again for taking the time to express your views.

  2. Pingback: Managers... How Much Justice Is Enough? • ScienceForWork

  3. Mark, your assertion of good practice guidelines that “Internal coaches are more effective than external coaches” is troubling me… For the avoidance of doubt I’m a strong advocate of internal coaching.

    The meta-analysis (Jones/Woods/Guillaume) found stronger effect for internal than external coaches but very clearly stated that interpretations needed to be viewed appropriately cautiously given the relatively low number of studies using internal coaches. In fact the Grover/Furnham paper states it is difficult to make conclusive judgments in this regard. It’s one thing to rightly question effectiveness in context and quite another thing to assert that good practice is using internal coaches.

    Would love to hear your further thinking on this.

    -1
    • David,
      Thank you for your comment. As far as the studies are concerned, the Jones et al meta analysis looked at 17 studies in total, involving 2,261 people. Of these, 3 studies involving 182 people concerned internal coaches. As you say, Jones et al state that their interpretations should be viewed appropriately cautiously, and Grover and Furnham confirm this. The question for me concerns the use of the word ‘appropriately’.
      As I say in the good practice guidelines section, ‘these guidelines may not apply to all businesses in all situations’. Context, as we know, is critical. As a practitioner myself, I agree with Jones et al when they say that ‘internal coaches inevitably have a better understanding of the organisation’s culture and climate’ and this is why they are likely to be more effective. If we define an internal coach as someone who has acquired that understanding and an external coach as someone who hasn’t, then the finding makes sense. My own view is that ‘internal’ versus ‘external’ is about more than whether or not coaches are on the payroll.
      Businesses aren’t going to stop coaching because the research says it’s difficult to make conclusive judgements about what works best. The purpose of the good practice guidelines in the evidence summary is to give people a starting point for designing and evaluating workplace coaching interventions. In practice, if internal coaches are the cheaper option, it would make sense to start there and if they don’t work, to then use appropriately briefed external coaches.
      However, what’s most important to me are the preconditions for effective coaching, namely the client’s readiness and the quality of the coaching relationship, whether that’s with an internal or an external coach. If businesses really want to maximise their ROI from coaching, then in my view their effort is best directed at pairing clients with the right coach when the client is ready, and not before.
      Obviously, more research is required in this area. In the meantime, I hope this response has addressed your issue, and I’d be interested to hear what you think.