Hilary J. Gettman ✉
Suzanne K. Edinger
(Antwerp Management School)
While the criticality of a strong coach-client relationship has received significant attention, this study represents one of the few investigations of coach behaviours that impact the relationship. Using qualitative and quantitative methods, we explore "contracting", defined as the collaborative determination of logistics, parameters and framework of the coaching engagement, as an important foundation for an effective relationship. We create a preliminary measure, the Contracting Inventory Scale, and investigate contracting’s connection to the coach-client relationship. Additionally, we explore executives’ perspectives on contracting as "infrastructure", a behavior that is necessary, but by itself does not lead to great outcomes, and discuss implications, noting that this study provides a platform for future empirical work and useful information for coaching practice.
executive coaching, leadership development, coach behaviour, contracting, coaching relationship,
Accepted for publication: 16 January 2019
01 February 2019
© the Author(s)
Published by Oxford Brookes University
The practice of executive coaching has been expanding dramatically for well over a decade (McKenna & Davis, 2009a) and comprises an increasing percentage of the multi-billion-dollar leadership development industry (Bono, Purvanova, Towler & Peterson, 2009). Organizations increasingly use executive coaching as a key component of their leadership development programs, with the number of organizations and the amount of money spent on coaching continuing to grow (Blackman, Moscardo & Gray, 2016; de Haan, Duckworth, Birch & Jones, 2013; ICF Global Coaching Study, 2016).
While research has not kept pace with the growth of coaching in practice, there has been an accelerated interest in more empirical and theoretically driven research (Blackman et al., 2016; de Haan, Bertie, Day & Sills, 2010). While research on coaching is still in its early stages, it has provided some valuable insights (Blackman et al., 2016; Stern & Stout-Rostron, 2013). Overall the consensus has been that coaching is largely an effective developmental practice, with positive outcomes ranging from increased self-awareness and improved relationships with others to improved leadership effectiveness and performance (Blackman et al., 2016; de Haan et al., 2010; Feldman & Lankau, 2005). Additionally, there has been consensus regarding effective coaching tools, including 360-degree feedback, qualitative interview assessment, use of assessment instruments such as personality and interest and leadership inventories, though there have been differences in approach based on the training of the coach (Bono et al., 2009; Gettman, 2008).
Most notably, the quality of the relationship between the coach and client has been almost universally believed to be a critical driver of coaching effectiveness (Baron & Morin, 2009; Bluckert, 2005; de Haan et al., 2013; Ely, Boyce, Nelson, Zaccaro, Hernez-Broome, & Whyman, 2010; Frisch, 2001; McKenna & Davis, 2009a). Many experts even assert that the relationship between the coach and client is the principal tool coaches have to affect change (Bluckert, 2005; Gentry, Manning, Wolf, Hernez-Broome & Allen, 2013).
Since coaching is an emerging field of inquiry, it seems logical to turn to the related and more established field of counselling, where there is considerable empirical and theoretical support for this assertion. We are not contending that coaching and counselling are interchangeable, rather that there are sufficient parallels for the well-developed theoretical and empirical work in counselling to provide useful insights and tools to coaching practice and research (McKenna & Davis, 2009a, 2009b; Passmore & Gibbes, 2007). In fact, counselling work has been used extensively in coaching research (e.g., Baron & Morin, 2009; de Haan et al., 2013; Greif, Schmidt & Thamm, 2010; Machin, 2010; Sonesh et al., 2015). However, we are mindful of the differences and use counselling constructs judiciously.
Counselling researchers have shown that a positive counsellor-client relationship is a critical component for successful counselling (Flückiger, Del Re, Wampold, Symonds & Horvath, 2012; Gelso & Samstag, 2008). They purport that the relationship provides the context that allows the client to accept and follow the counselor’s treatment and determines whether the counsellor can engage in social or interpersonal influence with the client, which helps bring about change (Flückiger et al., 2012; Horvath & Greenberg, 1989). This relationship is a strong parallel to the aim of coaching engagements (Joo, 2005) indicating its potential usefulness in the coaching context.
We are interested in examining the coach contribution to the coaching relationship. To be sure, no one factor or party can explain the relationship; it has been almost universally viewed as a dynamic collaboration. Nevertheless, for both academic and practical purposes, understanding the contribution of the coach is valuable information for the profession. Prior research investigated antecedents to a positive coach-client relationship. Researchers have examined factors such as the "fit" between coach and client regarding demographics and personality (de Haan et al., 2013; Scoular & Linley, 2006), and characteristics of the coach and/or client, such as readiness, motivation, personality, values, experiences, and credentials (Baron & Morin, 2009; Gentry, Hernez-Broome, Allen, Prochnow & O’Dea, 2010). Thus far, however, work on the coach contribution has focused predominantly on who the coach is (e.g., demographic fit, personality), while few have focused on what the coach does (e.g., techniques, behavior). Many coaches view their impact on the development of the relationship as rising at least in part from their behaviours, rather than solely their personal attributes, such as personality or demographics (Blackman et al., 2016; Bluckert, 2005; Sonesh et al., 2015).
There has been recent work investigating the role of coach behaviours in the coaching process. For example, Greif et al. (2010) developed behavioural "success factors", based on psychotherapy research, which the authors argue contribute to successful outcomes. Newsom and Dent (2011) examined the frequency of various coach work behaviours (e.g., "goal setting and attainment activities") and looked at differences in frequency based on education, and other demographic variables. Similarly, Bono et al. (2009) assessed the differences in coach behaviours between those with psychological and non-psychological training. This work indicated that the examination of coach behavior is a worthwhile endeavor. However, these studies did not tie behaviours to the quality of the coaching relationship.
The counselling literature more directly supports the idea that coach behavior might affect the relationship’s development. It has been long understood that "what is said [by the counsellor], how it is said, and their broader strategies (including how they are used) surely impinge on, color, and alter the relationship" (Gelso & Hayes, 1998, p.147; Duff & Bedi, 2010). In fact, research has established a connection between counsellor behaviours and the relationship (Martin, Garske & Davis, 2000; Najavits & Strupp, 1994) and there has been movement towards improving research and training by determining "evidence-based alliance-fostering counsellor behaviors" (Duff & Bedi, 2010, p. 92).
To be clear, we are not asserting that coaches unilaterally drive the relationship through their actions. As Ely et al. (2010) aptly noted, "Even the most skilled coach cannot create an effective coaching relationship on his or her own" (p. 587). Everything a coach does occurs within the context of the collaborative coaching process, which in some cases may include more parties than the client and coach, although these are the two parties we focus on in this study. However, we do contend that coach behavior plays an important role in the development of effective relationships, and importantly, it is the ingredient over which coaches have the most control. We attempt to address the role of behavior here.
The purpose of this study is to begin to explore behaviours of coaches that might contribute to a successful coach-client relationship. Specifically, we set out to examine coach behavior involved in the process known as "contracting". Based on our review of the coaching literature and models of coaching practice, we define contracting as the collaborative determination of logistics, parameters and framework of the coaching engagement, including the inclusion of others, and the goals, roles and responsibilities of each party. In some models, these types of activities comprise the first stage or step in coaching (Feldman & Lankau, 2005; International Coaching Federation (ICF) Professional Coaching Core Competencies, 2012 & 2017; Lee, 2013). While these activities may appear to be mere logistics or housekeeping, and therefore not particularly noteworthy or impactful, we argue that contracting is indeed important for several reasons.
The coaching literature suggests that contracting is essential in that it gives direction and purpose to the engagement, aligns coach and executive perceptions, creates buy-in, reduces ambiguity for the executive, clarifies the role of third parties, and eases concerns regarding confidentiality that could impede the relationship (Frisch, 2001; Hollenbeck, 2002; Lee, 2013; McKenna & Davis, 2009a, 2009b; O’Broin & Palmer, 2010; Turner & Hawkins, 2016; Witherspoon & White, 1996). In fact, Kilburg’s (1996) work suggested that these activities were an important factor in creating a successful coaching relationship. Indeed, contracting has been called "a best practice in itself" (Sherman & Freas, 2004, p. 88).
In 2008 the International Coaching Research Forum created a list of 100 research proposals intended to spur research in areas important to coaching practice (Kauffman, Russell & Bush, 2008), four of which were directly related to contracting. In 2013 Stern and Stout-Rostron assessed the progress made on these proposals. Contracting was the only area in which the authors did not find progress. There is a clear need for more work in this area.
Before discussing why we expect contracting to impact the coaching relationship, it is important to delineate how we conceptualize the relationship. First, we note that in many instances there are additional parties directly or indirectly involved in the engagement (organizations, supervisors, etc.), but for the purposes of the relationship, our focus is specifically on the relationship between coach and client. Other parties may impact that relationship, but we measure the relationship at the dyadic level. The coaching literature broadly discusses the criticality of a strong partnership and relationship between coach and client as a necessary condition for effective coaching (Lee, 2013; O’Broin & Palmer, 2010). As discussed above, the counselling literature has clearly shown that a positive working relationship (“working alliance”) is an important and necessary component for successful counselling (Flückiger et al., 2012; Gelso & Samstag, 2008). The working alliance is critical since it represents the personal rapport or bond between the client and counsellor, as well as their agreement on the goals and process of counselling. This alliance provides the context that allows the client to accept and follow treatment, and there has been significant empirical support that bond and agreement were strong predictors of outcomes (Ardito & Rabellino, 2011; Gelso & Hayes, 1998).
This conceptualization of the relationship has been thought to be transtheoretical and applicable across many domains. In fact, the importance of the alliance has previously been discussed and assessed in the coaching literature (de Haan et al., 2013; McKenna & Davis, 2009a). Therefore, we posit that these two factors – (1) bond and (2) agreement on goals and processes – are important aspects of the coaching relationship. Further, we posit that (3) personal trust and (4) perceptions of coach expertise/credibility are also important aspects of the coaching relationship. The coaching literature stressed personal trust of the client in their coach (de Haan, 2017; de Haan et al., 2013; Ely et al., 2010; Hollenbeck, 2002; ICF Core Competencies, 2017; Machin, 2010; McKenna & Davis, 2009b) and clients’ confidence in their coach’s capabilities (Boyce, Jackson & Neal 2010; Natale & Diamante, 2005) were critical to the ability of coaches to successfully influence and prompt change in their clients.
Accordingly, we consider four parts of the coach-client relationship: (1) "Bond" - the quality of the personal attachment and feelings of mutual respect between the coach and executive, (2) "Agreement" – perceived agreement about which goals executives should try to reach in coaching and what they should be doing to attain them, (3) "Trust" - the personal trust executives have in their coach, and (4) "Expertise" – the executive’s perceptions of the coach’s expertise or capability.
We suggest that contracting affects the coaching relationship in three specific ways. First, we predict that contracting will impact executive perceptions of the coach’s Expertise. There have been indications in the coaching literature that coach experience and knowledge are antecedents of a positive coaching relationship (Bacon & Spear, 2003; Joo, 2005). Strong contracting behaviours is one way of demonstrating knowledge and experience as a coach, and therefore should affect executive views of coach Expertise.
Second, we expect contracting to be related to perceived agreement about which goals and methods for goal attainment are best for the executive (Agreement). Those who engage in contracting behaviours (e.g., discuss objectives, parameters, expectations around level of commitment, responsibilities) are more likely to have come to a mutual understanding and agreement on the goals and methods for development through these discussions. Lee (2013) argues that contracting leads to a "unity of process and purpose" (p. 43) and a greater sense of partnership between the client and the coach. Indeed, a failure to discuss these issues may lead to a rupture in the coach-client relationship (McKenna & Davis, 2009a).
Finally, we expect contracting to be related to the personal attachment, feelings of mutual respect (Bond), and personal trust (Trust) between the coach and client. The process of coming to mutual agreement on goals, roles, and expectations should engender positive feelings of shared purpose, while the clear explication of parameters should increase trust in the coach’s motivation and intentions. Indeed, coaching researchers have argued that this replacing ambiguity with predictability and clarity serves to increase trust in coaching engagements (Lee, 2013; O’Broin & Palmer, 2010). Further, in counselling, client impressions of their counsellor and their relationship established early in the engagement have been found to be robust predictors of outcomes (Ardito & Rabellino, 2011; Duff & Bedi, 2010). The research suggested that if a positive relationship was not developed in the first few sessions it may derail the counselling engagement (Ardito & Rabellino, 2011). These findings lend support to the idea that what occurs early in a coaching engagement (when contracting typically occurs), and the impression a coach makes early on, may be especially impactful in the development of the coach-client relationship. Thus, contracting should be especially crucial in short-term engagements.
In this study we investigate contracting as part of the coaching process and its potential importance to the coach-client relationship. To this end, our study objectives are to explore the parameters and content of contracting as well as the role it plays in coaching from coach and client perspectives. Also, we develop a measure to assess contracting, provide preliminary evidence of its validity, and begin to examine the proposition that contracting provides a foundation for a quality coach-client relationship. We hope to raise questions regarding contracting and its importance, as well as to provide a useful tool for further study of the phenomenon.
In order to address these objectives, we first created and piloted a preliminary measure of contracting called the Contracting Inventory Scale (CIS). Measure items were created by drawing on multiple sources. First, items were based on a review of the coaching literature, including articles that directly discuss contracting, as well as many that discuss contracting related behaviours/interactions without explicitly identifying them as such (see Table 1).
Source/Support for Item(see key below)
Source key1. Bluckert (2005)2. Frisch (2001)3. Hollenbeck (2002)4. ICF (2012, 2017)5. Kilburg (2010)6. Lee (2013)7. Natale & Diamante (2005)8. Newsom & Dent (2011)9. O’Broin & Palmer (2010)10. Sherman & Freas (2004)11. Turner & Hawkins (2016)12. Witherspoon & White (1996)13. Subject matter experts
We also drew on ICF’s coaching competencies (ICF Core Competencies, 2012; 2017), and the researchers’ professional experience as coaches. Finally, we solicited the opinions of two other subject matter experts. These expert coaches had over 25 years of coaching/leadership development experience between them and have different educational and professional backgrounds which provided a broader perspective. The experts reviewed the draft CIS items and made suggestions regarding their content and wording. The resulting measure included nine items that asked executives to what extent they agree that their coach engaged in the contracting behaviours in Table 1 on a Likert-type scale of 1 (strongly disagree) to 7 (strongly agree). As depicted in Table 1, which includes the bases and support for each item.
In determining the content of the preliminary contracting scale, our information came from coaches in various ways (e.g., reading their literature, consulting coach-created guidelines, and talking with coaches directly). To help improve content validity of the scale, we sought the perspective of the other half of the relationship, the executives.
To this end, and to explore the connection between contracting and the coach-client relationship, we conducted semi-structured interviews with executives who were currently participating in a coaching program. We had several goals during the interviews. First, we wanted to examine the content of the preliminary CIS from the executives’ points of view and to modify measure items if necessary. We also wanted the executives’ perspective on what contracting activities mean for them (e.g., how they viewed them, what they perceived the impact to be on the coaching relationship, if they notice them occurring). Finally, we wanted to explore more generally the executives’ perceptions of their relationship with their coach, as well as their view of coach contracting behaviours that they believe impact the relationship and their coach’s effectiveness.
The respondents were 19 executives who participated in coaching as part of a cohort in an executive MBA program at a large public university. Thirty executives were solicited, yielding a response rate of 63%. During the 18-month program, each executive received approximately 10 hours of coaching. The process for contracting was unique to the individual coach and provided useful variability for exploring the construct. When the interviews were conducted the participants were in the 15th month of the program and had participated in four of five two-hour coaching sessions. The participants were 68% men and 32% women, and the ethnicity demographics were 68% white, 21% South Asian (e.g., Indian, Pakistani), 5% Latino, and 5% African-American.
The interviews were approximately 30 minutes long and were conducted over the phone and recorded. Executives were asked both open-ended and specific questions about the coaching relationship and contracting: (1) open-ended questions about their relationship with their coach (is it good and what do they base that on) (2) any behaviours/attributes of their coach they thought influenced the relationship, (3) a general question about setting expectations and specific questions about the occurrence of each contracting behaviour item (if not already mentioned by the executive) and (4) the perceived impact of contracting behaviours on their relationship with their coach. Further, executives were asked general open-ended questions about what they thought worked in coaching, what their coach could do differently, and anything else they wanted to talk about. These questions were to encourage them to generally discuss aspects of coaching and coach behaviour that they believed impacted their relationship with their coach and the effectiveness of coaching. These discussions were important for our understanding of the role of contracting as well as for informing future research.
The interview data were analyzed through deductive content analysis, which involves analyzing qualitative data according to an existing theoretical framework (Patton, 2002). The primary framework used in this study was the content and nature of the newly developed CIS, that contracting consisted of nine specific behaviours and that it has an impact on the coach-client relationship. Deductive content analysis is an efficient analytical approach that allows researchers to test existing theory with newly collected data. The first and third authors independently coded each interview transcript, looking for text related to the following categories: (1) evidence supporting the nine newly developed contracting items (whether coaches actually engage in these behaviours), (2) the existence of contracting behaviours not covered in our measure, and (3) a perceived link between contracting behaviours and the quality of the coach-client relationship.
Additionally, an inductive approach was used by the authors to code comments that were not easily categorized as part of our pre-established framework. Inductive analysis allows themes to emerge organically from data and does not attempt to match these data to a pre-existing framework or theory (Patton, 2002). We chose to use this additional approach for exploratory purposes to help us uncover concepts related to contracting that we had not thought of a priori, and to serve to point us towards areas for future research.
The purpose of the first two coding categories was to investigate the content validity of the CIS. There was 100% agreement between the coders that there were no new contracting behaviours mentioned, and that all the existing behaviours had been experienced by at least two executives, thus supporting the content validity of the scale. Therefore, no changes were made to the scale items.
More interestingly, and contrary to our expectations, the third coding category yielded very little data as contracting did not make much of an impression on the executives. When the executives were asked an open-ended question about contracting, they had little reaction, and when prompted with follow-up questions about specific contracting behaviours they often replied "yes" or "I think so." This lack of response was in contrast to their reactions to other areas they believed impacted the relationship with their coach, such as the coach being flexible or being prepared, where they generally were more engaged, spoke in greater detail and showed more enthusiasm. Thus, contracting behaviours, while they occurred, did not appear particularly noteworthy to the executives.
Given this information, it is not surprising that when asked directly about the effect of contracting on the coach-client relationship, the executives generally did not see it as having a major impact. The executives’ answers to the contracting behaviour question were coded as "yes", the executives saw a connection between contracting and the relationship, or "no", they did not. Only three of the 19 participants believed that contracting affected the relationship with their coach, although over a third stated that they thought contracting impacted the effectiveness of coaching in some way (e.g., the executives felt more accountable, or contracting caused them to be more prepared mentally for the coaching experience). Thus, we found a disconnect between coaches’ views of contracting (as evidenced in the literature) and these executives’ perceptions. In sum, the interviews provided support for the content of the contracting measure, while indicating that the executives did not place much weight on the importance of contracting itself.
Our exploration of coach and executive points of view about contracting left us with diverging opinions and an open question: Is the coaching community incorrect in their belief that contracting is important, or were these executives simply not aware of its impact? We followed up by surveying 111 executives via a web-based survey presented in Appendix A. We asked questions intended to indirectly examine the association between contracting behaviours and the coaching relationship, as well test the structure and reliability of the preliminary CIS.
Seventy-seven of the 111 participants were from three cohorts of an executive MBA program, and 34 were from a variety of organizations around the United States (attained by requesting that coaches involved in another study ask their clients to participate). There was no set process for contracting in the MBA program, and the other executives had no known connection to each other, so no set contracting process was followed. These participants were 64% male, 36% female; and 74% white, 9% African American, 10% South Asian (e.g., Indian, Pakistani), 3% East Asian (e.g., Chinese, Japanese), 3% Hispanic and 1% Middle Eastern (e.g., Egyptian, Saudi). Fifteen of the MBA program respondents had participated in the initial interviews, giving us an opportunity to complete a multi-method analysis.
As discussed above, we looked at four aspects of the coaching relationship and each was assessed using modifications of established measures. All relationship scales and items are located in Appendix A. First, personal attachment and respect between the coach and client ("Bond") was measured using the Bond subscale (α = .91) of the Working Alliance Inventory (WAI) (Gettman’s (2008) modification of WAI from Horvath & Greenburg, 1989). Second, the Task-Goal subscale of the WAI (α = .92) was used to measure perceived agreement about the most appropriate goals for the executive to pursue in coaching and which methods would be best to reach those goals ("Agreement"). Due to the general, non-clinical language used in the WAI, the items needed only slight modification, largely substituting "coach/coaching" for "therapist/therapy" (Gettman, 2008).
The third and fourth dimensions of the relationship were measured using subscales of the Coach Rating Form (CRF) (Gettman’s (2008) modification of Counselor Rating Form from Barak & LaCrosse, 1975) which were developed to capture the dynamics that allow a counsellor to exert interpersonal influence over a client. The "Trust" subscale (α = .93) was used to measure the trust executives have in their coach’s sincerity, openness and absence of motives for personal gain. "Expertise," or executive perceptions that their coach had expertise, experience and skills to assist with their development, was measured using the Expertise subscale of the CRF (α = .89). Similar to the WAI, the original items in the CRF were very non-clinical in nature, only necessitating a referent change from "therapist" to "coach".
Contracting was measured using our nine-item CIS. In addition, for the 15 respondents who participated in the interviews and survey, we created a non-self-report rating of contracting. Based on executive interview responses, the first and third authors independently coded their reports of coaches’ contracting behaviours as strong, moderate or weak. These ratings were based on: how much the executive talked about contracting when responding to the general prompt, how many of the contracting item behaviours they reported that their coach engaged in, and how enthusiastically they discussed those behaviours (e.g., "I think so" versus "Oh s/he definitely did that, several times"). There were no disagreements regarding these ratings.
In order to determine the underlying factor structure of the CIS, we conducted exploratory factor analysis. Our results indicated that the scale had a single factor as only one factor had an eigenvalue greater than one, and it accounted for 65% of the variance. Further, the loadings of all items onto that factor were between .62 and .89, well above the .4 threshold for determining if items are strongly related to the factor and should be retained as part of the scale (Costello & Osborne, 2005). The resulting nine-item scale had excellent reliability (α = .93).
The relationship between contracting behaviours and the various components of the coaching relationship was examined using two different assessments of contracting behaviours. First, hierarchical linear regression analyses were conducted analyzing contracting and the relationship measures as measured via survey. Since our data came from more than one source, we included source in the first step of the regression to account and control for systematic differences that might exist amongst the groups.
Our predictions that contracting would be positively related to the coach-client relationship were supported for all relationship variables (see Table 3). Contracting was positively related to clients’ beliefs about their coaches’ Expertise and Agreement on goals and methods for coaching. Additionally, personal Bond and Trust were also significantly related to contracting behaviours.
Data source only(Model 1)
Data source & Contracting(Model 2)
Note: Data source was represented as dummy variables with one data source serving as the reference group. The coefficients are not shown here, just the model summary statistics.*p < .05. **p < .01.
Second, we used contracting as measured by the coded interview responses of the 15 executives in the first part of the study. We examined potential significant differences in the quality of the relationship between executives whose interviews indicated their coach exhibited strong versus weak contracting behaviours using a one-way ANOVA. While parsing the data in this way and concentrating on extremes reduced sample size to six, it provided us the opportunity to explore contracting using a multiple method approach to triangulate our results. Despite the small cell sizes, we found a significant positive relationship, or strong trend, between contracting behaviours and three of the four relationship variables (see Table 4).
Contracting from interviewsstrong – weak (n=6)
Overall, our results provided support for the proposed connection between contracting behaviours and the quality of the coach-client relationship.
An unexpected discovery in this study was the apparent disconnect between what coaches and executives thought about the importance of contracting behaviour. As discussed above, we found that contracting appears to be relatively low on most executives’ radar, and that most did not view contracting as critical, noteworthy, or even interesting. By contrast, the coaching literature is clear about the importance of contracting, and our data suggested that contracting may have been more important than the executives believed. A review of the executives’ comments revealed that contracting was akin to a "hygiene" factor, something that is necessary to avoid negative outcomes, but does not in and of itself lead to positive outcomes (Miner, 2005). A classic example of a hygiene factor is money: not having enough money makes people unhappy but having additional money beyond a certain threshold does not lead to greater happiness (Kahneman & Deaton, 2010). In the executives’ own words:
"…I didn’t find it terribly necessary but I think that if you don’t cover those things then the ambiguity could be a derailer for the coaching engagement."
"If the process hadn’t been laid out… [the coaching engagement] could be hit or miss."
"…it didn’t add value to the coaching experience. It simply was – I’ll say infrastructure. I mean we didn’t talk about the room we’re in or the chairs we’re sitting on – we assumed that we would sit in chairs, and that would be a problem if you didn’t have them."
The hygiene factor link is a possible explanation as to why contracting did not appear to make an impression on the executives: it was just expected "infrastructure." It may be that contracting does not distinguish good coaching from great coaching, but instead distinguishes ineffective coaching from the whole range of effective coaching – from adequate to exceptional. In fact, Bluckert (2005) briefly referred to some aspects of contracting (e.g., confidentiality, expectations regarding homework, time, place and fees) as "hygiene’ factors" (p. 337). The lack of recognition on the part of executives does not diminish the criticality of contracting: if it is necessary infrastructure upon which to build successful coaching relationships, then it is critical that it be done and done well. However, placing additional emphasis on contracting beyond a certain point may not add value. Determining where that point may be is an important question for future work.
Our inductive approach in the interviews allowed another interesting issue to emerge. Two executives (one South Asian/Indian and one Caucasian American) volunteered their concern that some of their South-Asian colleagues were not particularly open to the idea of coaching based on cultural background, were not fully engaging in the coaching process, and therefore not obtaining the full benefits. The executives believed that individuals from this culture might require a more intensive introduction to the purposes of coaching and a deliberate "selling" of its benefits, suggesting that contracting might be of importance for this population, an idea that has been raised in the coaching literature (Peterson, 2007). We found preliminary support for these concerns in our data. Post-hoc, one-way ANOVA testing suggested that four Indian executives (not including the executive who had raised the issue) had lower ratings on three of four relationship measures than the rest of the sample (for details, see Appendix B).
Due to the small sample size, this assessment was clearly preliminary and definitive conclusions cannot be drawn from it. Even so, it highlights some important issues. First, the findings were consistent with general cultural characteristics – in a high-power distance culture like India (Hofsted Insights, n.d.), there is a culturally ingrained fatalism, an acceptance of one’s place, which poses a challenge to full engagement in any process involving the pursuit of individual change or betterment (Elder, 1966). In fact, this precise issue has emerged in the counselling literature (Manickam, 2010).
Additionally, the suggestion that more extensive contracting would be warranted for this population mirrors some of the most thoughtful work on coaching across cultures. For example, Peterson (2007) argues the necessity of more intensive targeted processes of getting to know the client and coming to a mutual understanding of the purpose, parameters and roles in cross-cultural coaching engagements. Passmore’s (2009) book on coaching across cultures highlights the great variation in development of coaching models specific to particular cultures, with India lagging behind. There has been consensus that much more needs to be understood about how to coach across cultures (Nangalia & Nangalia, 2010; Peterson, 2007; Ramanathan, 2017), and our study provided a particular path for such inquiry – that there may be measurable cultural differences in the relative importance of contracting to effective coaching.
Using multiple sources to bolster content validity, we created a measure of contracting, and demonstrated the structure and reliability of the scale in a sample of 111 executives. Additionally, we found preliminary evidence that contracting was positively related to the coach-client relationship, supporting our proposition that contracting is important in establishing a successful coaching relationship.
The cross-sectional design of our study limits our ability to draw causal conclusions from our findings. Thus, for example, we cannot say definitively whether strong contracting causes a positive coach-client relationship, or whether the reverse is true. In future studies, it will be important to survey executives at multiple points in time so causality can be determined. The CIS identified nine contracting behaviours that were uni-dimensionally related to a positive coach-client relationship. Our intent was to develop a holistic measure of contracting behaviour so each behavior was represented by a single item (e.g., there was one item related to confidentiality). The CIS scale construction technique prevented us from determining the relative importance of the behaviours represented by each of the nine items.
Future work could create multiple item subscales for each of the contracting behaviours that may then assess the relative importance of each which might be of interest to coaches in informing their contracting practice. Further, such work could determine if there are circumstances where certain aspects of contracting become critical (e.g., clear understanding of the scope of confidentiality may be especially crucial in circumstances where the executive has a difficult relationship with her boss). Such additional research can help us to refine our understanding of contracting behaviours and their impact.
Our sample of participants were executives in an MBA program, a context which differs in important ways from many coaching engagements. This context can limit the role of the executive’s organization, which is so powerful in many engagements. Due to this result, the full impact of contracting around the inclusion and role of third parties likely was not fully captured in this study even though third-party inclusion was covered in a CIS item. Future research should include a larger, more representative sample of coaches and executives from different organizations to better assess the role of third parties, and to improve the generalizability of the findings. Further, it would be surprising if contracting in some form was not important in the mentoring relationship and other types of coaching. However, since this construct and measure were created and validated using only executive coaches and clients, more research is required to determine what contracting would look like and what its role would be in these other areas.
Finally, direct effects of contracting behaviours on coaching effectiveness (rather than just the coaching relationship) should be investigated. Several executives discussed the impact of contracting on coaching effectiveness, indicating a fruitful and useful avenue for further research.
This study contributes to both the research and the practice of executive coaching. We have taken a step in moving beyond examining who a coach is as it impacts the coach-client relationship by empirically examining what a coach does that might have an impact. While the work on coach characteristics has certainly been useful in selecting coaches or matching them to appropriate clients, it does not help coaches work more productively with the clients they have. Coaching researchers have taken initial steps into examining and understanding coach behaviours essential to effective coaching, but so far there has been little theoretical or empirical work regarding those that might contribute to an effective coach-client relationship, a key part of effective coaching engagements. Our study adds to the nascent empirical work by providing a measure of contracting and preliminary support for the role of contracting behaviour in the development of a strong coach-client relationship. The nine behaviours we identify in our measure focus around communication, setting expectations, and discussing various parameters of the coaching engagement. These findings provide a solid starting point for practitioners interested in establishing a high-quality relationship with their clients.
Additionally, we advance empirical work in this area by developing the Contracting Inventory Scale and providing evidence of its content validity, uni-dimensional structure and reliability. This measure allows for detailed investigation of the impact of contracting behaviour on the coaching relationship, processes and outcomes. Clearly, more work is needed before we completely understand the role coach behaviour plays in forming effective coaching relationships, and this study provides a step in that direction, as well as a platform for further exploration.
One expert was a PhD in psychology and a professor of management in a large public university; the other expert was an MBA with decades of executive experience, an ICF master certified coach, and an instructor in an ICF certifying program at a large, private university.
We drafted our items at a relatively general level of specificity in order to encompass multiple ways of engaging in that behavior, rather than risk excluding those we did not think of, causing problems with content validity.
The coders had an initial inter-rater agreement of .85, and all differences were discussed and reconciled.
Note that in the WAI for counselling, agreement about tasks and goals were measured with two separate subscales, but in Gettman (2008) the task and goal sub-factors collapsed into one factor. We tested these items in our data and found the single-factor result.
Note, due to an established surveying process in the additional executive MBA programs, the items were collected on a 5-point Likert scale (which had the same end and mid-point labels as the original scale). Data were scaled to a 5-point scale before the analyses were conducted. Further, the data were non-normal (executives tended to rate towards the high end of the scale), violating an assumption for use of parametric analyses. To reduce this non-normality, variables were transformed (using the Blom transformation) before running parametric analyses.
We used principle components analysis with an oblique rotation (direct oblim) since we expected the data to contain several inter-correlated quantitative variables.
Note, we only completed this analysis with the participants from the interview study since we had information about their cultural background. We did not have similar information for the larger sample.
Hilary Gettman is the Associate Professor of Business Administration and Director of Management Programs at the Leo J. Meehan School of Business, Stonehill College; Executive Fellow and Coach at the Smith School of Business at University of Maryland.
Suzanne Edinger is Assistant Professor of Management at the Leo J. Meehan School of Business, Stonehill College.
Karen Wouters is Professor of Leadership at Antwerp Management School.
All relationship items were rated using the scale below
Please indicate how much you agree or disagree that each of these words describes your coach.
Please indicate how much you agree with the following statements regarding your relationship with your coach.
non-Indian – Indian Executives (n=14)
RADAR: Research Archive and Digital Asset RepositoryAbout RADAR