The Coaching Black Box: Risk Mitigation during Change Management

: A case study of strategic renewal in the Chinese education market, this paper explores a non-directive coaching model and its impact on risk mitigation, knowledge exchange and innovation in strategic renewal through the application of multi-tiered coaching and manager coaches. Through an ethnographic action research methodology, we ask “Can coaching mitigate organisational risk and increase the likelihood of positive outcomes in change management?” and “Can managers, acting as internal coaches, increase knowledge socialisation and mitigate risk in the change management process?” The paper ﬁnds that there is no inherent failure rate in the change management process and that a strategic management approach can mitigate risk liberating managers and organisations to seek to create the collaborative environments that support organisational learning and strategic renewal, thus moving beyond a narrative of failure to one of strategic empowerment and a strategic management approach to risk mitigation. We conclude that a data-driven approach to organisational learning and Professional Learning Communities helps teams to ask the right questions and to mitigate risk through better aligning the organisation to its strategic reality, exploiting organisational learning to achieve competitive advantage and ensuring that systems and processes continue to match the emerging strategic reality.


Introduction
In 2019, the Chinese Ministry of Education launched its 2035 Education Modernisation Plan (Ministry of Education of the People's Republic of China 2019), with General Secretary Xi Jin Ping noting that this will be a fundamental pillar for the rejuvenation of the Chinese nation, as China pushes towards becoming a more knowledge-based economy, requiring a sustained and significant focus on training of teachers (Xi 2018).
At the heart of this policy, the initiative is a recognition that learning may be one of the few sustainable strategic advantages for an organisation (DeGeus 1988), as in an increasingly Volatile, Uncertain, Complex, and Ambiguous (VUCA) business environment that is being increasingly transformed and disrupted by technology developments, generally and particularly by Artificial Intelligence or Machine Learning, companies need to "change or die". We must also recognise that to reduce organisational risk by leveraging organisational learning into innovative strategy, and it must be approached from a systemic perspective (Senge 2006).
Approaching organisational learning and the inherent risk mitigation in this manner rejects the premise from the financial risk perspective that systematic risk cannot be reduced; by adopting a strategic management perspective towards risk, we accept that systematic risk can be mitigated by implementing a superior strategy that better aligns the organisation to its emerging strategic environment (Otim et al. 2012). A strategic management approach to risk mitigation enables organisations to also move beyond the financial risk perspective; unsystematic (organisation-specific risk) should be diversified to adopt a more proactive, and therefore potentially more impactful, strategic approach as we recognise that the organisation can learn and, in doing so, adapt strategy, systems, and processes to mitigate unsystematic risk better.
Given the significant changes to Chinese education, the challenges that individual schools will face as organisations implement this strategic renewal, and how closely these changes and business environment mirror both the wider Chinese business environment and challenges faced by global businesses, this research uses a case study of a school in Southwestern China to consider how coaching can be used to both mitigate risk in the change management process and foster innovative organisational learning.
The case study explores how a multitiered approach to coaching at the individual, team and organisational levels can be leveraged to increase the socialisation of knowledge and by doing so increase the opportunities for a collaborative approach to strategic innovation. Whilst, for team members and managers, this may create a more engaging and productive workplace, the potential for positive impact at on organisational level may also be considerable, as collaborative practices, facilitated through effective coaching, enable a shift from a traditional approach of risk diversification to a strategic management approach to risk mitigation. Through empowering teams, this shift in approach may have a significant impact on overall risk mitigation as better and more rapid alignment to the market may reduce "positioning risk", whilst, innovation through effective organisational learning is likely to reduce "competitive risk" and as teams are able to more efficiently react to changes in the organisation, systems and processes can be more readily adjusted, thus, it is likely that "operational risk" is minimised.

Change Management Failure
Since the announcement of the 2035 plan, many schools have sought to launch projects to "internationalise" education by either importing and deploying foreign resources or engaging international partners to train teachers and manage curriculum projects. Anecdotal evidence suggests that many of these initial attempts "failed". Amongst the project team designing the teacher training and coaching programme for the case study school, this became a hot topic along with how to avoid the 70% failure rate of all change management projects, something they were insistent was a real danger.
A 70% failure rate had not been discussed concerning local projects, and it was difficult to understand where this idea originated. However, a Google search of 70% of change efforts fail returns 108 million results centred around the idea that 70% of change management efforts do indeed fail, perhaps explaining the fixation on this figure. However, as Hughes (2011) notes, there is no empirical evidence that supports this claim, and as such, it is an urban myth.
Between 50 and 70 percent of the organisations that undertake reengineering efforts do not achieve the dramatic results they intended (Hammer and Champy 1993). Over time, this figure used to imply that 70% of change initiatives fail, resulting in Hammer issuing a clarification stating that there is no inherent failure rate in reengineering projects (Hammer 1996). McKinsey noted that executives reported that between 20 and 38% of change initiatives were completely or mostly successful (Keller and Aiken 2009). Therefore, it is tempting to suggest that this would mean that 62-80% of change initiatives fail. However, about a third of executives reported that their change management initiatives were mostly successful. This introduces the interesting concept that change management is not an all-or-nothing game and that partial successes are possible. This is also supported by an IBM study wherein executives reported that 41% of projects meet their objectives and that 59% miss an objective or fail (IBM 2014). Therefore, missing an objective does not necessarily constitute project failure, while partial or indeed significant project success is possible without achieving all objectives.
Although theory suggests a positive linear relationship between risk and return, an increasing number of studies suggest that strategy and organisational structure may be associated with risk mitigation, as lowered risk does not lead to lower return (Jemison 1987;Lau 1987). As we deepen our understanding of the relationship between risk and return, we must acknowledge that the actions of managers may indeed mitigate risk and increase returns (Tanriverdi and Ruefli 2004). As Otim et al. (2012) note, strategic management leads to risk mitigation through the creation of sustainable strategic advantage. Given this relationship, the strategic risk may be quantified through examination of relative performance to other organisations as expressed in the following equation (Adapted from Otim et al. 2012): Total risk = Positioning risk + Competitive risk + Operation risk = Performance relative to the market. Therefore, risk can be mitigated as organisational learning aligns strategic direction to the emerging strategic reality and empowers teams and managers to develop a competitive and operational advantage.
Recognising that there is no inherent project failure rate and that risk can be mitigated through effective organisational learning gives greater hope for project success. However, we must recognise that when we consider the increasingly VUCA environment, the complexity of the project being undertaken and the human side of team dynamics, often with a highly emotional stake in the change, a positive project outcome is far from guaranteed. This leads to our first research question: Can coaching mitigate organisational risk and increase the likelihood of positive outcomes in change management?

The Role of Coaching and Knowledge Management in Risk Mitigation
Having recognised the need to take a strategic approach to organisational learning (Senge 2006), we must recognise that perhaps the most significant challenge preventing organisations from achieving sustained strategic advantage through organisational learning is the very fact that the drive for decision-making in the strategic process often does not capture the learning from professional learning communities (Brown and Duguid 2001) or the often complex, interpretive system in play within an organisation (Weick 2001).
Traditionally, many organisations have focused on developing strategic innovation over several years rather than in a single event (Stopford and Baden-Fuller 1994). However, the rise in rapid and disruptive technology-driven changes that many organisations face in an increasingly VUCA environment means that a strategic renewal or step change in strategy is increasingly appropriate for the strategic environment (Crossan and Berdrow 2003). This strategic renewal process relies on the organisation being able to take an agile approach to strategic development as they read the strategic environment and align systems and strategies to meet new challenges (Barr and Stimpfert 1992). The process of strategic renewal, and institutionalisation of organisational learning inherent in this process, highlights a significant tension as the organisation must explore potential solutions and exploit them in a timely manner to achieve strategic advantage (Nonaka and Takeuchi 1995). Crossan and Berdrow's (2003) 4Is model gives some insight into how strategic renewal can be achieved by describing how the microprocesses of intuiting, interpreting, integrating and institutionalising organisational learning through a virtuous cycle of feedback and feedforward enable organisations to achieve strategic renewal. They argue that through cycles of continuous reflection and action that occur at the level of the individual intuiting and interpreting, the individual makes sense of the change and finds ways to implement, whilst also as part of a group they can engage in interpreting and intuiting, and at an organisational level learning is integrated and institutionalised. Central to this approach is recognising that for effective organisational learning and knowledge management to occur, there is a need to move beyond Tayloristic-style thinking about how knowledge workers perform their tasks to focus on knowledge management and organisational learning itself (Schütt 2003). The need for this shift is further intensified as we recognise that increasingly complex, data-driven technology solutions mean that in an information-rich society, information itself is easily searchable and therefore becomes external to the individual, thereby counting increasingly as public property. Accordingly, we recognise that knowledge is the property of the individual, created as they apply their understandings and professional experience to data and information (Swan 2011).
Such a shift to post-industrialist knowledge management approaches means that organisations increasingly need to identify and hire team members with the right knowledge and a mindset to collaborate with others to engage in organisational learning and develop an innovative strategy. This approach recognises that most organisational knowledge rests with the individual (Rasmus 2002). However, through the organisation creating an effective collaborative environment, knowledge can be shared and evolve as new learning becomes part of the organisational DNA (Schütt 2003). This duality between individual and organisational knowledge means that, given the difficulty in monitoring and controlling knowledge creation, a unique management skillset is required for effective organisational learning and strategic renewal to occur (Bhatt 2001).
Increasingly, coaching has been adopted to support collaboration and individual development (Ellinger et al. 2003) as it provides a mechanism through which team and individual development can be facilitated whilst ensuring that productivity is enhanced as both teams and individual goals may be better aligned to organisational objectives. In seeking to achieve a better alignment between goals, individual and team development, managers are increasingly required to coach at the team and individual levels, whilst their coaching effectiveness is becoming recognised as a precursor to change management or project success (Boyatzis et al. 2013). From a knowledge management perspective, deploying coaching at the team and individual levels may have a positive impact on both collaboration and socialisation of knowledge (Percy and Dow 2021); however, to achieve this, we must address the complexity of social processes, power structures and the challenges inherent in using managers as internal coaches, given that their behaviours, leadership style and coaching capability will have an impact on individual and team commitment towards change and achieving project goals (Chen et al. 2011).
From a strategic risk management perspective, the importance of "unlocking" individual knowledge and creating collaborative learning environments cannot be understated, since positioning risk may be mitigated as strategic direction is aligned to and overcomes phenomena created in the market (Lubatkin and Chatterjee 1991;Tanriverdi and Ruefli 2004); the competitive risk may be mitigated as organisations innovate and compete more effectively (Miles and Snow 1978), whilst operational risk may be mitigated as teams ensure that systems and processes are aligned to the emerging strategic reality (Cowell et al. 2007).
The above considerations lead to our second research question: Can managers, acting as internal coaches, increase knowledge socialisation and mitigate risk in the change management process?

The Coaching Approach to Risk Mitigation
In seeking to leverage the positive potential of a multi-tiered coaching approach, the team adopted a coaching approach defined as follows: Coaching is a relationship of rapport and trust in which the manager coach, working at individual and team levels, uses their ability to listen, to ask questions and to playback what individuals or teams have communicated in order to help them, as the manager coach effectively engages in the coaching dance, to clarify what matters to them and to work out what to do to achieve their goals and aspirations. (Percy and Dow 2021) Predominantly non-directive, this coaching approach recognises a significant tension for the manager as an internal coach. They must both dance between directive roles as managers and more non-directive roles to coach effectively (Thomson 2009). At the heart of this approach is a common coaching theme of raising awareness to help the individual/team take responsibility for their development and take appropriate action. The approach can be further understood through the underpinning SHARP: In using stories, the coaching manager can leverage them to help individuals and teams make sense of emerging, complex, strategic realities and change (Huzzard 2004), whilst at the same time challenging and breaking away from practices that were previously taken for granted (Pratt 2000) as teams can construct new strategic realities meaningful in their current situation (Cunliffe et al. 2004). As they construct these new realities and overcome challenges inherent in the change process, the coaching manager is mindful of the need to maintain "harmony" between the organisational, team and individual goals whilst ensuring that systems and processes within the organisation change, as needed, to support an emerging strategic reality. Mindful of the VUCA environment and the emotional strains on individual and team coaching, managers seek to support teams to "accept" the VUCA environment and to move forward, knowing that they may not always have complete information and solutions; yet, they must not be frozen by pressure-induced inaction as they take positive action.
In deploying this approach, the coaching manager structures team and individual coaching conversations using the GROWTH model (International Growth Coaching 2009) and a data-driven approach to examining "reality" and creating individual development plans or strategic direction. Acting on a data-driven narrative is more likely to lead to positive outcomes as critical evaluation helps the manager and their team assess the validity of new practices and the ability to generalise these practices (CEBMA 2018). Such datadriven analysis is not only confined to evaluating new technical approaches, as people analytics can also give a strong insight into how people add value (McCracken et al. 2017). Organisations that can harness data and people analytics effectively may be twice as likely to significantly outperform their competitors (Smith et al. 2019).

Impact of the Manager on Risk Mitigation
Much of the coaching literature focuses on dyadic coaching deployed for executive development (Baron and Morin 2011;Gessnitzer and Kauffeld 2015); yet the need for continuous incremental improvement alongside more innovative strategic development suggests that the use of internal manager coaches may be more effective in supporting knowledge socialisation and strategic innovation (Joo et al. 2012;Kim et al. 2013;Percy and Dow 2021). The shift to the manager as a coach can increasingly be seen in the coaching literature (Boyatzis et al. 2013;Heslin et al. 2006;Liu and Batt 2010), suggesting that effective coaching skills are likewise becoming more linked to managerial executive success (Boyatzis et al. 2013;Heslin et al. 2006). However, there is little research about the skills that managers will need to be effective in these areas or the potential negative impact of managers as internal coaches (Ellinger et al. 2003;Liu and Batt 2010).
Although the driving intentions behind coaching are positive, there is a tension that exists in the coach manager-coachee dyadic relationship in that the manager relies on the success of team members and the team for their career progression, whilst the coachee is reliant upon the manager for reward and development opportunities (Anderson 2013). Regulator focus theory (Higgins et al. 1997) suggests that the coachee, in this situation, is likely to have one of two dispositions, either being promotion or prevention-focused as they either try to display positive career and in-role behaviours whilst trying to achieve goals, or adopt avoidance behaviours in trying to avoid negative consequences. A manager who is overly goal-driven, pressure-based and focused on the likely negative outcomes of a situation is likely to elicit a prevention-focused response from the coachee as they try to avoid adverse outcomes (Heslin et al. 2006;Hui et al. 2013). On the other hand, when a coachee feels supported by their manager and confident that the coaching process is likely to have positive outcomes, they are more likely to be promotion-focused and actively seek to support the team in meeting their objectives (Agarwal et al. 2009;Ellinger et al. 2003).
Understanding this dynamic highlights the need to train managers and provide them with coaching skills to meet their new expectations better. However, there is an increased significance from a change management leadership perspective as Beer and Nohria (2000) note, leaders are likely to adopt two leadership styles when dealing with change. "Type E" leaders are likely to focus on economic incentives to drive change, layoffs, downsising and restructuring. On the other hand, "Type O" leaders are likely to focus on corporate culture, developing organisational and individual learning through Plan-Do-Review cycles. Although the US corporate environment has a bias towards Type E leadership styles focusing on quick returns for shareholders, it should be noted that neither style, deployed alone, is likely to lead to long-term success. Much like overly task-focused manager coaches, Type E leaders are likely to deliver quick returns. However, as trust and rapport are broken, they are unlikely to deliver sustained strategic advantage. Type O leaders may become too attached to employees, put off hard decisions, and fail to achieve financial and more structural business goals.
Beer and Nohira also note that the best chance of change management success is likely to come from deploying both leadership types together. An initial intuitive leap may conclude that change may be managed through sequential deployment, first restructuring, and building culture. However, as trust broken during restructuring is almost impossible to rebuild, this approach is likely to fail. They further suggest that the tensions between the leadership types and their inherent actions should be openly acknowledged as leaders set direction from the top, focusing on both the "hard" and "soft" sides of the organisation whilst planning for spontaneity.
There is a significant congruence with the organisational learning and coaching literature reviewed thus far, suggesting that the coaching dance may be important in supporting managers to strike a balance between Type E and O leadership styles. This entails maintaining an effective collaborative yet goal-driven environment whilst deploying professional learning communities to leverage cyclical learning at individual and team levels and, in doing so, seeking to achieve the benefits possible from effective use of the microprocesses in the 4Is model and its more agile approach.

Materials and Methods
From a research perspective, the project took place over three semesters (approximately 18 months during the 2020-2021 and 2021-2022 academic years), wherein three key phases of action took place as first coaching managers were provided with 12 h of training in non-directive coaching skills over three half-day workshops between which they worked with practice clients. Following the coaching skills training, coaching managers were supported to work with their teams to set up Professional Learning Communities (PLCs) and through data-driven, continuous Plan-Do-Review cycles to identify pedagogical goals and put these into action through a process of experiential learning and coaching support. Finally, based on lesson observations and PLCs data, training was provided to address leadership and teaching skill development needs.
From a "timeline" perspective, it is important to note that the original research project was intended to last for only one semester, covering the training and embedding of the coaching process. However, a significant event happened in the school just after this study finished as, due to a lack of market growth caused by the COVID-19 pandemic and some less-than-ideal hiring decisions during school setup, the parent company of the school decided that significant restructuring was required, hiring a new CEO, who with decidedly E Type leadership style restructured and laid off approximately one-third of the teachers and support staff. Although this was a painful time for the organisation, it offered a somewhat unique opportunity to see how the already successful coaching programme would hold up under the pressure of such a clear manifestation of the VUCA environment and to observe the dynamics of E Type and O Type leadership styles at play in a culture of coaching. Resultantly, the decision was made to begin a consecutive study and extend the original research period by 12 months.
In the initial coaching skills training phase of the project, 14 middle managers each received 12 h of coaching training, and additional support was provided, between sessions, for the managers to practice their coaching skills with practice clients. To provide a baseline of their coaching skills, at the start of the first training session, the managers participated in a coaching exercise where, as part of a coaching trinomen (observer-coach-coachee), they undertook coaching role-play. Each manager took a turn in each of the roles, with the coaches' coaching skills being scored by the observer and the workshop leader, an accredited master executive coach. This exercise was repeated at the end of the training to measure growth in coaching skills.
In the interest of fairness and achieving an objective measurement of coaching skills, a skills rubric was used to score the sessions, measuring the following areas: • Key coaching skills of listening, questioning and playback; • Ability to support the coachee in setting appropriate goals and action points; • Ability to use the coaching dance (moving between directive and non-directive approaches) to help the coachee arrive at their solution; • The use of a strengths-based approach to coaching.
Given the need to support coaching role-play and simulate the coaching manager's need to engage in the coaching dance, the participants were provided with a coaching brief to provide an after-lesson observation coaching session to a teacher who had not observed significant poor student behaviour during the lesson.
As the coaching programme moved beyond training to implementation in the school, the workshop leader took the role of coaching supervisor and supported the coaching managers in establishing PLCs with their teams and setting goals. In parallel, the school's management team modified the appraisal process to include regular teaching observation, followed by coaching debriefing sessions, and ensuring that bonus payments rewarded the adoption of new behaviours to support the development of a coaching culture. At the end of the first semester, semi-structured interviews and a modified Fitzpatrick scale were used to measure the coachees' satisfaction with coaching and the impact on the wider environment. As the study progressed into the second academic year, following the restructuring, more focus was placed on semi-structured interviews (due to the richness of data that they provided) using evidence from multiple sources to triangulate the stories being told in the interviews. This shift in focus occurred as the project emphasis moved from examining the rollout of the coaching programme to examining the complex shifts in individual and team relationships caused by the restructuring and how coaching supported teams and individuals in their continuing work and development, thus mitigating risk in the change management process.
Interviews were undertaken with each of the managers and team members separately as well, as part of the PLC process, teams were also debriefed about their projects and encouraged to tell their stories to support the embedding of organisational learning. For the individual interviews, semi-structured interviews were based around a modified version of the Fitzpatrick scale allowing for the areas of the effectiveness of coaching; development of job-related skills; effective project completion and achievement of Key Performance Indicators to be explored. Further exploration was facilitated through a series of questions around psychological safety, organisational learning and knowledge exchange. Interviewing managers and team members separately more effectively enabled tensions in the manager coach-coachee relationship to be explored by allowing differing perspectives to emerge and for contradictions or tensions to be explored. Throughout the process individuals and teams were encouraged to tell their stories to provide a rich narrative that could be used to explore how the coaching process either supported or hampered the micro-processes of Intuiting, Interpreting, Integrating and Institutionalising during strategic renewal.
In acknowledgment of the inherently social nature of coaching, knowledge management, and change management, an ontological approach to the study was adopted. This also recognised that multiple complex realities would form in a VUCA environment with multiple key stakeholders (Whitehead 2005). Given the social realities and complexities that arise from many perspectives, and recognising that knowledge is derived from both our experiences and how we understand them (Bryman 2008), we recognise that the methodologies from the "natural sciences" lack appropriateness for this study (Cohen et al. 2007).
An Ethnographic Action Research methodology (Tacchi et al. 2003) was chosen, which allowed for cyclical reflection on both strategic and cultural issues to shape the coaching interventions and take appropriate action to address emerging needs. This allowed a phenomenologically oriented paradigm to be used, and the resultant inductive analysis allowed significant themes to emerge from the data collected.
For credible description to emerge and for meaningful observation to occur, the project was conducted over eighteen months allowing for multiple action research cycles and strategies to be formed and focused according to the emerging reality. This process particularly supported deploying the multi-tiered coaching approach as individual coaching took place alongside team coaching that supported the goal-setting process and Plan-Do-Review approach as department PLCs trialled new initiatives. These PLCs were also supported by developing and using a series of coaching protocols that helped them structure a data-driven approach to their activities.
Adopting this approach allowed for the use of a Natural Cultural Learning process (Whitehead 2005) through which the "child" is taken as an ethnographic model and researchers can both learn and guide research as they both participate in and interpret events as they unfold, much as a child learns for the first time. This approach allows for multiple data gathering strategies, necessary to create a rich narrative that captures the complex realities that emerge that include: The author deciding how to weave the narrative and tell stories that arise from the research is an inherent weakness to this methodology, leading rise to concerns about reliability and validity as neutrality is almost impossible (Bath 2009). This weakness is compounded by the absence of control groups and what some may identify as the lack of rigour associated with quantitative methodologies (Cohen et al. 2007). However, it should also be noted that the complexities of social interaction necessitate such an approach as this allows researchers to be more focused on the contradictions that arise, and through this awareness to monitor and guide both data gathering and reflection processes in a way that ensures rigour of the research process (Ball 1990).

Results
Throughout the study, it became clear that the middle manager's role in a multi-tiered coaching approach to change management is pivotal in achieving positive results, as they both lead and help develop their teams. Therefore, it is important to understand managers' perceptions of their coaching skills and examine the impact of coaching skills training on their development. During the training exercise, the coaches and the workshop leader rated their coaching performance (using rubrics) against a 5-point Likert scale. After the first round of coaching role-plays and scoring, there was a significant gap of two points between the managers' perceptions of themselves as coaches and the workshop leader's rating. Although this suggests that managers have an unrealistic perception of their coaching capability, we note that across the course of the 12 h of training, this narrowed to an average gap of 0.5 points.
In terms of skills development, listening was the strongest skill moving from an average of 3 to 4.5, whilst engaging in the coaching dance was the least developed skill, moving from 1.5 to 3. Although protocols and prompts were provided, we note that both more directive and non-directive managers struggled with this; when switching to a more directive approach, the directive (often more E Type) managers found it difficult to "dance" back to a non-directive approach, and the non-directive (often more O Type) managers found it difficult to "dance" to a more directive approach or to challenge the coachee. Having noted this, the average increase in coaching skills for all coaches was two points.
In providing coaching skills training for the managers, we recognise that the traditional HR roles of informal training and team members' development are now commonly passed to managers (Latham et al. 2005). Therefore, it is increasingly important to understand the skills and capabilities required for a manager to be successful in these areas. We note that those managers who were more successful ensured that their teams understood their goals and purpose (Bennett and Bush 2013;Klein et al. 2006); they gave appropriate guidance around tasks that needed to be accomplished, gave ongoing feedback and took personal responsibility for helping the team to meet their goals (Heslin et al. 2006;Lorinkova et al. 2013). These observations were underpinned by the findings from the first semester Fitzpatrick scale results, as the average team member rating for effectiveness of coaching was 4.3, with the team members noting that they preferred a manager who gave structure through the coaching process and who was willing to share professional experience and advice when necessary. In terms of how coaching supported their development, team members gave an average rating of 4.7, whilst the managers noted that it was initially difficult to get team members to take responsibility (Redding and Hsiao 1990) and that there was a strong need for a goal-setting culture that was reliant on resilience and adaptability (Chang et al. 2005). However, the first semester saw a significant improvement in meeting goals and alignment with team and organisational goals (Scharma 2007). Therefore, it was not surprising to the team that, based on evidence from appraisal and goal tracking, project and goal completion was rated at 4.8.
In terms of teaching skills development, on a 5-point scale, the average growth from a baseline evaluation over semester one was 1.5 levels and over the full 18-month project was an average of 2.8 levels. These figures are supported by real-world achievement over the 18-month project, and the school achieved an average of 3.8 awards per teacher (at the district level or above). Indeed, several teachers represented the city in teaching competitions, and case studies of the innovative practices being developed in the school were shared in regional and national press stories.
Although it was expected that such drastic restructuring would have a significantly negative effect on both morale and outcomes across the teams, this was not the case. Through observation and discussion with team members, it seems that the teams acted as a buffer, providing social cohesion and a collaborative environment that was more immediate to team members so that when they got over the initial shock of the restructuring, their more immediate team environment and relationship with managers had a stronger impact than that of more senior managers. This was supported by a subtle yet significant change in messaging from the school leaders, who, recognising the danger the sudden restructuring represented, emphasised that innovation is tough, goals must be met for the good of the students and wider community, and that teachers working in the school were "safe" as long as they continued to strive for progress. This came with a shift to leadership training for middle managers, focusing on servant leadership, highlighting that good servant leaders are responsible for delivering outcomes for the wider community and earnestly serving their teams to ensure they are the best possible. This change in messaging and approach was significant as it recognised the tensions in behaviours necessary to streamline the organisation against a backdrop of needing to innovate, support each other, and collaborate to achieve goals. It was also a refocusing of earlier messaging that had focused almost solely on the benefits of coaching. However, in the context of building a coaching culture and mitigating risk in change management, it was a savvy move, as although "servant leadership" may be seen by some as being an oxymoronic statement (Russell and Stone 2002), its adoption is linked to many positive outcomes at the individual level as team members are more likely to assume leadership roles as required by the project (Lacroix and Verdorfer 2017) and show fewer subversive behaviours that may derail projects (Sendjaya et al. 2017). Furthermore, as they feel better-perceived levels of organisational fit, team members are likely to display more positive career and role behaviours (Irving and Berndt 2017), becoming more committed to change (Kool and van Dierendonck 2012). This has knock-on effects at both team and organisational levels as, through increased feelings of psychological safety, teams are likely to collaborate better and display greater innovation and creativity (Irving and Berndt 2017;Liu and Batt 2010;Schaubroeck et al. 2011). At an organisational level, trust is improved (Schaubroeck et al. 2011), and leaders are seen as being more effective (Taylor et al. 2007).
From the interview responses of the team members and managers, we could see some of the "insulation effect" at the team and individual levels; through a servant leadership approach, managers were applying the core principles of coaching (Passmore and The Association for Coaching 2010) and, through a non-directive approach, engaging in the coaching dance, by doing so, communicating well with their teams in a manner that helped them to navigate the VUCA environment effectively, thus ensuring good team outcomes by delivering effective team and individual development (McLean et al. 2005;Park et al. 2008). Whilst this narrative begins to capture some of the impacts of the middle managers, it does not capture the more holistic picture that emerged from the interviews and data. To see this, we need to look at an Amended 4Is Model (Figure 1 The amended 4Is model, built upon Crossan and Berdrow's original model, adds several key elements that may help to clarify the role the middle manager has taken in mitigating risk in the change management process, how coaching facilitates a collaborative environment, how storytelling can help sense make and sense break, and how data helps to drive innovation. Storytelling was used throughout the project and is increasingly recognised as a strong facilitator of the coaching process (Reissner 2009). It supports transformational learning and change by facilitating cognitive and behavioural change as the coachee imagines and acts on a positive brighter future (Du Toit 2007). When used in this way, storytelling helps teams and individuals to ask the right questions and to collaborate to "socially construct" new strategic realities that are more fitting to their current situation as they sensemake and sensebreak by questioning and challenging the past in the process of creating new narratives for success (Cunliffe 2008;Reissner 2008).
During the project, there were some cases of transformational learning, as coachees experienced a shift in their frame of reference for their experiences (Mezirow 1997); however, we found that experiential learning was more likely as storytelling helped The amended 4Is model, built upon Crossan and Berdrow's original model, adds several key elements that may help to clarify the role the middle manager has taken in mitigating risk in the change management process, how coaching facilitates a collaborative environment, how storytelling can help sense make and sense break, and how data helps to drive innovation.
Storytelling was used throughout the project and is increasingly recognised as a strong facilitator of the coaching process (Reissner 2009). It supports transformational learning and change by facilitating cognitive and behavioural change as the coachee imagines and acts on a positive brighter future (Du Toit 2007). When used in this way, storytelling helps teams and individuals to ask the right questions and to collaborate to "socially construct" new strategic realities that are more fitting to their current situation as they sensemake and sensebreak by questioning and challenging the past in the process of creating new narratives for success (Cunliffe 2008;Reissner 2008).
During the project, there were some cases of transformational learning, as coachees experienced a shift in their frame of reference for their experiences (Mezirow 1997); however, we found that experiential learning was more likely as storytelling helped individuals and teams to make sense of their new experiences and through making the learning explicit, they were able to apply the learning to new challenges (Kolb 1984). Through facilitating a cognitive shift, both forms of learning helped individuals and teams foster emotional and behavioural learning that we observed had a lasting impact on their perceptions and behaviours (Reissner 2008). Critical to mitigating risk in the change management process, this learning forced teams and individuals to seek to ask the right questions and engage in critical dialogue, and through a process of considered reflection, increased performance is achieved (Brockbank and McGill 2006) and, as stories act as a carrier of meaning and new understanding, and improved strategic direction emerges (Brown and Duguid 2001;Weick 2001). As we understand how stories support learning and the emergence of strategic direction through the coaching process, we can understand that coaching is not just about creating a psychologically safe environment, fostering collaboration, or developing personal impact and leadership skills. Through raising awareness and helping individuals and teams take responsibility for performance, coaching drives them to ask the "right questions" to collaborate in developing new strategic direction and take action; in doing so, they can take a strategic management approach to risk mitigation.
PLCs were a big part of both experiential learning, allowing the teams and individuals to take action. The data-driven approach that was taken in PLCs had a significant double impact, allowing the manager to use the data as a "mirror" against which the team could evaluate its progress, thereby providing an opportunity to dance into a more directive role if needed, and, as they decided how to measure success, the team were able to define their questions and strategic approach more sharply. Concerning the development of individual teaching skills, the development of people analytics proved to be particularly useful as clarifying outcomes, measuring progress, making intervention as required, and rewarding the adoption of new behaviours through the appraisal process all brought strong returns.
Once a clear direction is provided at an organisational level, the manager's role cannot be underestimated. As can be seen in the amended 4Is model, the manager coach is both a leader of a team and part of a middle management team, acting, therefore, both as a critical interface with senior managers and a sense maker and breaker for the team. Consequently, devolving training and development to managers has a significant impact as they can intervene when they need to provide training and support to individual team members whilst bridging the gap between an individual and the team. Being able to act immediately and support organisational learning means that the manager can take a strategic management approach to risk mitigation, ensuring that strategic direction remains appropriately aligned to changes in the VUCA environment and developing competitive and process-driven responses to this new emerging reality.
Regarding the first question: Can coaching mitigate risk and thereby increase the likelihood of positive outcomes in change management? The research findings suggest that there are several key ways how the use of a multi-tiered approach to coaching may mitigate risk in the change management process, as it supports the ability of teams to both explore and exploit new ideas in a timely manner through providing the psychological safety and opportunities for cognitive development that enable them to sensemake, collaborate and act strategically. Through this process, they are also more able to align the team and individual goals to the change management initiative and, by doing so, are more likely to achieve positive outcomes.
Regarding our second research question: Can managers, acting as internal coaches, increase knowledge socialisation and mitigate risk in the change management process? Although the managers initially overrated their coaching skills, a relatively short training course could significantly impact and improve coaching skills and coaching outcomes. This is critical, as developing better relationships with their team managers could form effective, data-driven PLCs that enable the teams to unlock individual knowledge, innovate, and quickly break away from old practices. A critical part of this process was that through effectively interrogating data, the teams could ask the right questions that enabled them to take action that was closely aligned to organisational objectives, mission and vision.
In extending the study, we had expected to observe that the coaching methodology that had worked so well under "fair weather" conditions might come under pressure in the "storm" created by the VUCA environment and sudden restructuring. However, it is clear that the middle managers created a safe environment, at a team level, that insulated the teams from the worst effects of the storm. We believe that this was in no small part due to the coaching dance enabling managers to more effectively move between leadership styles, creating an environment of accountability yet creating psychologically safe and collaborative team environments. It is also clear that the team managers were able to act as an interface between individuals, the team and the organisation as they aligned goals and resources to achieve the desired change and facilitated strategic organisational learning.
Given the above, we can see that the adopted coaching approach provides a solid foundation for risk mitigation, whilst the Amended 4Is model highlights the importance of the data-driven PLCs in helping to foster a strategic approach to organisational learning and mitigating risk in the change management process. However, as we reflect on the models, their ability to support risk mitigation and the narrative that has emerged, we are mindful that some key actions required to support coaching efforts or reinforce risk mitigation efforts are not well represented. To address this gap in the narrative, the Coaching Black Box (Figure 2) is provided to support discussion about what happened when PLCs and coaching efforts were needed to support interventions when teams were not functioning at the optimal intersection of Team-Task-Individual-Coaching. Given the above, we can see that the adopted coaching approach provides a solid foundation for risk mitigation, whilst the Amended 4Is model highlights the importance of the data-driven PLCs in helping to foster a strategic approach to organisational learning and mitigating risk in the change management process. However, as we reflect on the models, their ability to support risk mitigation and the narrative that has emerged, we are mindful that some key actions required to support coaching efforts or reinforce risk mitigation efforts are not well represented. To address this gap in the narrative, the Coaching Black Box (Figure 2) is provided to support discussion about what happened when PLCs and coaching efforts were needed to support interventions when teams were not functioning at the optimal intersection of Team-Task-Individual-Coaching. In recognising this gap in the narrative, we note the similarity to "survivor bias", wherein during the second world war, the US military analysed returning bombers to consider where armour should be placed to protect them better. Examining the hit marks, they initially decided to reinforce the areas most hit by the enemy (Mangel and Samaniego 1984). However, an insightful analyst (Abraham Wald) noted that it was the areas that had no hit marks that required reinforcement, as it was most likely that those were the areas that had been hit on the non-returning aircraft and therefore had caused critical damage. In the spirit of avoiding survivor bias and reflecting on the coaching process, we ask ourselves, "which areas of coaching practice needed more "armour" to mitigate risk better?" In examining the Coaching Black Box model, we can see that it is centred around the In recognising this gap in the narrative, we note the similarity to "survivor bias", wherein during the second world war, the US military analysed returning bombers to consider where armour should be placed to protect them better. Examining the hit marks, they initially decided to reinforce the areas most hit by the enemy (Mangel and Samaniego 1984). However, an insightful analyst (Abraham Wald) noted that it was the areas that had no hit marks that required reinforcement, as it was most likely that those were the areas that had been hit on the non-returning aircraft and therefore had caused critical damage. In the spirit of avoiding survivor bias and reflecting on the coaching process, we ask ourselves, "which areas of coaching practice needed more "armour" to mitigate risk better?" In examining the Coaching Black Box model, we can see that it is centred around the PLC cycle. We recognise that PLCs were the "vehicle" that drove organisational learning and the change management process, whilst coaching was like the "oil" that kept the engine running smoothly. We also recognise that managers could not become too relaxed in their approach to the PLC process as each new challenge needed to be approached as if for the first time, with small changes in team dynamic or external environment with the potential to undermine psychological safety collaborative learning environment. This was particularly the case when members left or joined the team. With this in mind, each time managers began a new PLC cycle, they were encouraged to reconnect with their team and revisit the relationship between team goals, organisational goals, and expectations to determine how they would work within the current PLC cycle. Given the data-driven nature of the PLCs, this was also the perfect opportunity to revisit data and set goals in the current context, whilst managers were also encouraged to be mindful of their management style and coaching skills during the process. This fresh start mindset was particularly useful as with engagement in each project, individual team members evaluate the usefulness of the coaching intervention and approach, alongside their relationship with the manager, and subconsciously decide whether to engage with positive career and in-role behaviours. Finally, managers needed to be aware that change in a VUCA environment requires resilience, and therefore monitoring that teams or individuals are not being stretched too far, as even though they may have a Growth Mindset disposition, this may change to a particular challenge. These elements combine to either reinforce or break the rapport and trust needed for effective coaching, providing a positive or negative synergy to the coaching process and, therefore, moderate team and individual engagement.
As we move beyond data-driven goal-setting and take a deeper look at the multitiered coaching process, we note that we concur with Bhatt (2001) as although middle managers are now able to respond to the training and coaching needs of team members and teams at the moment of need, the knowledge management and coaching process become increasingly difficult to monitor and support as it becomes less formal and less visible than traditional training and coaching approaches. Therefore, we note the existence of the "Coaching Black Box", wherein the coaching manager provides a goal-focused, multi-tiered, coaching approach and where, although it is comparatively more difficult to observe the process, the nature of the impact can be seen through team dynamics and both individual or team outcomes. With this in mind, an essential element of "armour" in the process was to ensure that effective coaching supervision was provided to help managers reflect on the actions they were taking and help design appropriate interventions at the team and individual levels to help achieve PLC goals. In providing coaching supervision, we recognise that support is required as although managers were provided coaching skills training, they were not accredited coaches and did not yet have the rich depth of experience that comes from coaching practice and insights developed over time. We also recognise that significant reflection and coaching intervention were required to help maintain the optimal coaching environment, where Team-Task-Individual-Coaching intersect to create an optimal organisational coaching environment that supports knowledge management and innovation alongside a strategic management approach to risk mitigation.
In recognising the tensions of maintaining the optimal coaching position, the following reflections on what happened when managers or teams could not maintain this effective collaborative environment are provided to give a deeper insight into how risk was mitigated. As we begin this reflection, we note that we concur with Beer and Nohria (2000) as we also observed that E Type managers (overly task-focused) broke trust and often could not rebuild this to provide the psychological safety required to create collaborative environments. This was observed in the case study as the newly appointed CEO was dismissed after six months due to pushback from faculty and local officials concerned about the culture and lack of trust emerging within the school. On the other hand, O Type (in this case, highly coaching-focused) approaches, whilst effectively building culture, failed to deliver on business goals. In this case, a young leader (at the school level) was a "victim" of the restructuring, as despite building a psychologically safe environment and being well-liked by their colleagues, they were unable to deliver goal-driven outcomes effectively. Therefore, we note the importance of the coaching dance in helping coaching managers more effectively reach an appropriate balance. However, we also note that this alone is not sufficient as they must also effectively engage at a team and individual level.
The need to reach leadership style balance and engage with individual and team levels was seen throughout the case study, as we note that following a significant success or when lacking direction, teams often operated in a Team-Coaching-Individual space, either "coasting" after the success or suffering from the lack of direction. In both cases, we found that re-engaging with the goal-setting process and taking data-driven action through PLCs helped to redress this balance. However, in doing so, it was important not to apply inappropriate amounts of stress, often seen when teams were given unrealistic expectations or deadlines to meet; becoming stuck in the Team-Task-Individual space meant that, in the short term, teams were unable to create the psychological safety required for collaboration or innovation. Suppose this imbalance had been maintained over time. In that case, we note that rapport and trust would be eroded, and the teams would be unable to maintain effective working relationships with each other or their manager. While most teams were able to avoid this, we note that for the team that became caught in this space, there was a comparatively high rate of faculty turnover. As news of the working environment spread, it was increasingly difficult to hire new faculty.
For teams stuck in the Team-Task-Individual space, or when new members joined the team, a coaching intervention was undertaken to help teams build a foundation of trust and clearly define purpose and goals. Although this intervention initially focused on the Coaching-Team-Task space, its aim was to enable managers to align individual goals with those of the team and achieve the optimal coaching environment. For those teams that managed this process well, there was a significant gain in performance over time, particularly if the emphasis was placed on this during the teams' "forming" phase as we noted that there was significantly less "storming", or that storming was more productive as teams were not afraid of appropriate conflict and the creative energy that is provided. However, for those teams that failed to align individual goals, there were some unexpected outcomes.
Placing emphasis on a multi-tiered coaching process and moving coaching to occur below the C suite may suggest a lack of emphasis on coaching in the Task-Individual-Coaching space. However, we note that the traditional coaching outcomes of increased executive development, professional growth and increased engagement were achieved. However, we also note that two significant outcomes were observed when there was no alignment with team goals. Firstly, one team manager and several team members (from multiple teams) leveraged the coaching process to develop their skills and then leveraged them to move to other organisations, leading to a loss of organisational knowledge and destabilising the team. Secondly, when there was a lack of team alignment and when coaching led to significant performance gains or rewarded an individual, this often led to jealousy and breakdown in team relationships. However, perhaps most damaging for a team caught in the Coaching-Team-Task space was when a frustrated individual acted in a goal-focused manner operating in the Task-Individual space to achieve what they thought the team should be doing; this was observed to quickly lead to conflict that was often difficult to move beyond as a team member was in outright defiance of the coaching manager. This was evident in a team that had been successful for most of the study; however, in six weeks, they degenerated to dysfunctionality as the team leader failed to provide leadership, and a team member sought to fill this vacuum.
Overall, the appropriate and timely intervention worked to keep teams functioning in the optimal coaching space, and we concur with Crossan and Berdrow (2003) as they highlight the 4Is microprocesses as they support organisational learning and strategic change. However, we also note that for this to be most effective, middle managers needed to act effectively, engaging in the coaching dance and providing timely intervention at individual and team levels to ensure that these processes worked and were aligned to achieving organisational objectives. As we move beyond the Coaching Black Box to the end of the PLC cycle, we also note the importance of the middle manager in facilitating reflection so that the team can engage in cognitive processes necessary to embed learning at a team level and institutionalise new organisational learning.

Conclusions
The case study effectively demonstrates how a multi-tiered coaching model helps make an agile approach to the strategic management of risk mitigation a reality as individual and team goals are aligned to achieve organisational change. Teams are empowered to use a data-driven approach to ask the right questions to meet their needs and the emerging strategic reality, thereby, achieve a sustainable strategic advantage as they achieve strategic renewal through innovation. Such an approach to the strategic management of risk mitigation is likely to have a significant impact on overall risk mitigation as better and more rapid alignment to the market may reduce "positioning risk", whilst, innovation through effective organisational learning is likely to reduce "competitive risk" and as teams are able to more efficiently react to changes in the organisation, systems and processes can be more readily adjusted, thus, it is likely that "operational risk" is minimised. However, in transferring the functions of coaching and training to the manager, we must acknowledge the existence of the Coaching Black Box and, in doing so, continue to ask what training and support are required to increase the personal impact and leadership skills of middle managers.
The case study provides a novel approach and differentiates from existing literature in the field as recognising that there is no inherent failure rate in the change management process and that a strategic management approach can mitigate risk liberates managers and organisations from creating collaborative environments that support organisational learning and strategic renewal, thus moving beyond a narrative of failure to one of strategic empowerment. Whilst we recognise the ability to coach to support the microprocesses that make this possible, given the VUCA environment and highly emotional nature of change, change management is a "messy" process and, as such, greatness is unlikely to be achieved by following processes alone, as a little "luck" and an agile approach to organisational learning are required to achieve strategic renewal.
The Coaching Black Box and its place within the PLC cycle also differentiates from existing literature as it moves beyond a dyadic approach to coaching and examines areas in which the coaching process may fail and therefore inhibit the change management process as managers take on a more coaching focused role. This reminds us that achieving an optimal coaching or organisational learning environment that provides a strategic management approach to risk mitigation is not a static state, as the middle manager must continuously evaluate the needs of their team and provide appropriate coaching intervention or training at the time and point of need for both the team and individuals as needed. In recognising this shift in leadership approach and the need for the manager to engage in the coaching dance effectively, we acknowledge that further research is required to establish the nature of coaching techniques and interventions that are best suited to support managers and their teams as they seek to move from each segment of the Coaching Black Box to the optimal coaching and risk mitigation position.
In acknowledging the significant positive potential of maintaining an optimal coaching and risk mitigation position we recognise that there is considerable need for continued research to create a data-driven organisation that can use predictive analytics and machine learning not only to minimise the downside risk in their organisation, but also to become a truly learning organisation where upside potential is also maximised. Having taken place in an education context, in China, with a relatively small sample size the learning from this case study should be applied with some caution in new environments. However, even though the case study may represent somewhat of a "flying pig" scenario it does provide significant signposts for further research, highlighting the significant need for research into how to better support middle managers in the coaching process alongside a deeper investigation into circumstances that may cause the coaching process to fail.
Author Contributions: Conceptualization, W.P. and K.D.; methodology, W.P. and K.D.; formal analysis, W.P.; writing-original draft preparation, W.P.; writing-review and editing, K.D.; supervision, K.D.; project administration, W.P. Both authors have read and agreed to the published version of the manuscript.