Peer Learning is...

Exploring how peer learning helps in our fast-changing world.
Sign up for our short course

Evaluating "CoachingOurselves" - a peer-learning format for leadership development

When Philip Lenir was ready for a turn in his career away from speech recognition programming, he got interested in management education. His eye fell on Henry Mintzberg’s IMPM (International Master’s Program for Managers) program at McGill University. It was an international program, designed for senior managers, and the education was based on reflective learning in management practice.

Phil appreciated the principle of reflection, but he didn’t have the kind of money that could pay for the full IMPM course. So he wrote Professor Mintzberg with a challenge:

If reflecting on experience with your peers works in a classroom setting, then why not just do it in my office?

That was the start of a new business, called “CoachingOurselves”, which Phil founded together with Henry.

Evaluating the Peer Learning characteristics of CoachingOurselves

The concept behind CoachingOurselves is simple. A team of middle managers is brought together every month for 90 minutes to reflect on the current challenges in their work. Sometimes they are from the same team, and sometimes from across company divisions.


The group formation process is one of the variables that Phil needs to toggle with in the program design.

There are an infinite ways of putting these groups together and I like to think of this like the "seating at a big wedding problem". When choosing who sits at what table during a big wedding the bride and groom have to carefully think about personalities and affinities between members of their respective families. They are trying to bond the their two families by helping people develop relationships.

Every CoachingOurselves participant forms new relationships throughout the organisation because they're put into cross-organisational groups. They meet people whom they would not have likely gotten to know on their own, so we note that it expands connectivity.

The people in each group tend to be chosen centrally, which gives direct control to program curation. At the same time, this doesn't systematically offer connections outside those groups, or distribute the decisions about who can join the groups.


A facilitator supports the conversation by helping the group choose one of the discussion modules provided by CoachingOurselves. These modules are written by established management & leadership thinkers. They contain questions that have proven to be relevant issues from these thinkers’ research. The questions guide the conversations, and initiate participants to reflect.

Together, the modules cover the 5 management mindsets that make up IMPM’s management program: Reflection, Analysis, Worldliness, Collaboration, and Action. Topics range from Crafting Strategy, to Bullying, to High Performing Teams.

There are over 90 of such modules available, but a sub-selection is usually compiled that fits to what the company’s leadership and management framework deems relevant. Within these bounds, the program can adapt to current issues that participants face in their work.

CoachingOurselves doesn't apply any diagnostics on participants themselves for the design of the program.


During a session, participants each read out one of the pages in the modules, which contains a question. For instance in a module on Dealing with the Pressures of Managing, a question would be read out like: “I am a slave to the action, so I can’t fully concentrate on the big issues. Do you feel like that sometimes?”

At the end of the reflection session, participants are encouraged to note down their take-aways, and implement the best ones over the coming month.
Lenir explains how finding people with similar challenges enables them to take action:

In the early sessions participants take aways are in the line of: “Wow, I thought I was the only one going crazy around here, now I know we are all going crazy together."

After a few sessions you get responses like “I feel like I now 'know' the other participants in my group. Now, when I have problem with the shipping department getting my product out to the retail outlets, I just call 'Tom' in shipping and we talk. We talk like regular people trying to get stuff done -- not like we did before when we would interact like the head of shipping talking to the head of sales. It makes is sooo much easier to get stuff done!

Finally, usually by the fourth or more sessions we see groups decide to take on larger actions items. The groups learn to collaborate to make bigger changes happen.

For example, in one organization, after about the fourth session (it was on innovation), the group started discussing how the Finance Departments policy of getting project managers to go through a formal project review which included hard ROI analysis resulted in most innovative projects being cancelled before they even started.

So the group devised a strategy to get the finance department to change this policy. It involved multiple "informal" discussions key members of the executive committee, certain well timed and diplomatically "correct" emails etc... and the policy got changed. And the result is that the management team solved the problem that the organization needed solved, which was to increase the rate of innovation.

One of the key aspects that enables CoachingOurselves to work is building relationships between the participants in the group. At first facilitation is required to get things going. But once a safe space has been created, then people start sharing their experiences like they would with their spouse or family members, and the facilitation gets in the way.

Programs tend to run in a cycle of 6 meetings. After each cycle, the group evaluates whether they want to continue another cycle or not. In Lenir’s experience, once a company decides to stop covering for meeting time, the gatherings tend to discontinue.

The progress of relationship-building, and the control participants take over the meeting faciliation show that the program succeeds in initiating agency with participants. However, once people are activated, and inspired to progress the program doesn't support acceleration of further actions, other than a repeat of the program cycle.


CoachingOurselves holds characteristics that support the very beginning of learners to explore their leadership learning interests. The mix of participants, and the relative autonomy to select discussion topics creates a safe space to share, and gets people into their learning mode. The program raises awareness with people on their relationships, and what they can accomplish when they've made deeper relationships with people in the organisation. (see the ARC model characteristics below)

The peer learning characteristics of CoachingOurselves, modeled on the ARC framework

Next steps

We’ve been fortunate to work with a couple of companies for the long term, and we really see an impact. Fujitsu from Japan for instance has built a granular KPI data set over the course of the 12 years we’ve been working together.

We can measure the density of employees participating in our peer learning groups, and correlate that density to performance up to the department level, or to that of a store location. And the numbers pan out!

At department level, we see correlation with better profitability. Even on an individual level, we there is correlation with employee satisfaction. Our approach supports better performance.

Company's always raise this questions when they invest in their employees: "Will our investment in the individual aggregate to company-wide improvement?" CoachingOurselves does a great job in connecting learning on the level of the individual, and performance of the company.

The question for evolving CoachingOurselves is what happens when participants have completed the program’s cycle. There are already some hints that there's demand for more, with managers organising reunions once in while to keep sharing.

The ARC Model reveals the practical next steps are to focus on Responsiveness by introducing diagnostics, and Agency by accelerating participants in directions they feel enabled to pursue after the initial 6 meetings.

Deeper diagnostics

It would be interesting to perform deeper, systematic diagnostics on the progress participants have made during the program, and determine what their next step would be. There will be great insights that can be aggregated to company-wide policies, varying from practical improvements, to investment, and new business development.

The standard approach to diagnostics would be to introduce one-off coaching meetings, either with senior managers or peer groups.

A less costly approach to diagnostics could be to organise a lessons-learned event for all the participants of a coaching program cycle. By using an open space format, and recording the conversations that take place in the break-out sessions, you'll get a bird's-eye view of all the topics that are currently relevant, and various perspectives on them.

Accelerated progress

The insights from diagnosing the groups will contain clues on how to structure further support.

Building on this, it's worth experimenting with ways that allow alumni to suggest topics for modules that address their challenges as they see them. A conservative approach to this would involve new modules being created with a group of senior managers who were also coaching, and a freer form of this would involve open co-creation of modules with light support from the CoachingOurselves team.

As another example, some individuals will likely want to continue on a track to leadership, to shape their career in a specific direction. They could be supported by forming peer groups within the field they want to excel (like Shell does with its maintenance engineers in Fat Rat workshops), perhaps together with ongoing personal mentorship.

Measured results

CoachingOurselves is particularly extraordinary because it deploys small peer groups at such a large scale, with measureable results. Whatever the next step will look like, the program's foundation allows for testing ideas in the form a small experiment with selected cohorts of people. Even based on qualitative analysis, the program will be able to get signals of what initiative is contributing to company performance, and what is not.

Only once such measures bolstering diagnostics, and agency are in place will decentralising the group membership decisions become valuable. Then the direction of progress of participants, and that of the company are aligned, and they will be able to pick their own group members more effectively (perhaps even teaming up with peers outside of the company). It can only open more opportunities for company-wide learning.

© 2020 Source Labs Ltd.