Escape Velocity with Agile – mastering the transition through Horizon 2

In Part 1, I proposed the alignment between the phases of an agile transformation and Mehrdad Baghai’s three investment horizons as follows:

• Early agile pilots equate to high risk horizon 3 investments
• Early strategic rollout equates to medium term horizon 2 investments
• Agile as the de-facto delivery method equates to horizon 1.

In this post, my focus will be on leveraging Moore’s thinking regarding effective transitions between horizons.  The final Escape Velocity chapter is titled  ‘Execution Power: Engineering the Escape’, and opens with the following statement: “The challenge is to deploy a next-generation initiative at scale, overcoming the inertial resistance of our current go-to-market system to do so.”

Moore describes an arc of execution where any initiative goes through 3 modes: Invention, Deployment and Optimization.  These by and large map onto the 3 investment horizons as follows.  A Horizon 3 initiative is purely invention.  Horizon 2 is a blend of invention and deployment with the balance shifting towards deployment as it nears Horizon 1, and Horizon 1 begins with some deployment but is primarily oriented towards optimization.

Two key themes resonate throughout the chapter.  The first is that a different style of management (and inherently a different set of managers) is required for each mode, and the second that effective transition between modes is the cornerstone to success – “each transition consists of a program to drive the organization to a tipping point so that what began under one set of managers is now fully taken up by another”.

The entire intent of a good agile adoption is to create a learning organisation with a culture of continuous improvement, which should make the transition from deployment to optimization trivial.   For me, the transition from invention to deployment is the focal point – beautifully summarized by Moore:

“… the handoff between inventors and deployers is unnatural because visionary inventors think their work is done as soon as the first instance of the new offer has been proved to work; pragmatist deployers are not willing to take responsibility for something until there is proven market momentum behind it.  This leaves the inventors fuming that no-one is taking up their offers while the deployers are rolling their eyes wondering who is ever going to clue these people in.  Absent any meaningful handoff, the deployers will cling to the low-growth legacy opportunities where their bread is buttered today”.

In driving this transition, Moore argues strongly the invention-mode requirement for a “fully integrated team in which all the mission critical functions – R&D, engineering, manufacturing, sales, services, and marketing report directly into a single entrepreneurial leader”.   He later emphasizes that the team orchestrating the transition must also not only be cross-functional but empowered … “Members of the team must be empowered to commit their functions to the new way of going – there can be no nonrowing observers on this boat.  Absent that capability to commit, cross-functional exercises degenerate into an endless series of meetings with no actionable outcomes other than to schedule the next meeting.”

In the context of an enterprise agile transformation, the functions represented may be different but are no less important to successful transition.  The functional lines must at minimum include governance, finance, release management, operations, enterprise architecture and legal.  In a heavily outsourced scenario, one must also argue for senior vendor representation in the group as part of the lean supplier transition from vendor to partner.

In a typical large enterprise, building high performance development teams is child’s play compared  with successfully re-aligning the corporate governance processes.   Agile transformation all too easily defines these functions as ‘outside the system-of-work’ and falls for the trap of localised optimisation.   The reality is true enterprise-scale success cannot exist without bringing the entire system of work into alignment.

My argument is that if you apply Moore’s ‘transition orchestration team’ concept to agile adoption, you must wind up with an agile working group formed along his lines.  Rather than a group of coaches and agile advocates attempting to drive change out into the governance line functions, you bring the line functions into the group. 

Inevitably, there will be a significant period where waterfall is still the horizon 1 delivery metaphor for the enterprise and agile is moving through horizon 2 maturity and the two compete for resource and influence.   If empowered representatives of the functional silos are embedded in the process throughout transition, the transition team not only understand the entire system of work but possess the influence required to successfully conceive and execute the change.  

The most powerful aspect of this approach is that you then seize upon the tacit knowledge creation cycle.  By working within the agile transition initiative, the line function representatives ‘learn by doing’.  They gain tacit experience of the underlying agile principals, and can drive their deployment into organisational process based on informed situational practice.  As the initiative moves from deployment mode to optimisation and the working group returns to its line function fold, the knowledge creation cycle is completed.