Since encountering Dean Leffingwell’s Scaled Agile Framework(SAFe), I have presented the model to dozens of senior executives and middle
management from numerous enterprises grappling with the question of how to
effectively scale agile. Without
exception, the ‘Big Picture’ resonates.
It is easily incorporated in their mental model of the key ingredients
in delivering enterprise scale outcomes, and a 1 hour overview leads to the
response ‘I can believe in that’.
The challenge, of course, arises when you start to delve
into the follow-on question – “Now how do I make it work in my special
situation?” Like any good framework, it
is designed on a set of principles.
Implementation requires leveraging your understanding of the principles
to tailor the detail to your situation.
In this series of posts, I will endeavour to illustrate with
working examples the fashion in which it has been applied in the group I have
been guiding for the past 9 months.
Approach
The success of this implementation has attracted a great
deal of interest in both the enterprise the group belongs to and other
Australian companies grappling with scaling agility. On a weekly basis, groups visit to
understand ‘how it is done’.
Over time, we have established a ‘2-hour tour format’ that
walks the model. Tours always begin at
an A0 poster of the ‘big picture’. We
spend 5 or 10 minutes giving a brief overview, then proceed to work our way
through the portfolio, program, feature and team kanban walls illustrating the
way work flows through the layers of the model and the type of discussions they
facilitate. My intent in this series will
be as much as possible to provide a simulation of a physical tour. It will be structured as follows:
- Part 1 - Introduction and Context
- Part 2 - Demand management and the portfolio kanban
- Part 3 - Program level pipeline management and the program kanban
- Part 4 - In-play work and the program level feature wall
- Part 5 - Results to-date and Conclusion
Context
The setting is a large enterprise in the midst of an Agile
transformation. They experienced good
success with many smaller agile projects, but encountered significant
challenges in the early stages of applying it to major programs of work. Over the course of 2012, 4 major groups across
the enterprise successfully adopted SAFe.
Each applied it to different problems and tailored it to suit – the one
I will be writing about focuses on business intelligence and data warehousing.
In April, I heard Dean describe his typical approach with a
new client as “give me your worst program, I’ll fix that first and we’ll keep
adding a program at a time from there”.
It’s a fairly apt description of the case at hand. Agile project teams existed, but knitting
them together into a program with appropriate cohesion, visibility and
governance continued to be traumatic.
The model was applied first at the program level (the 2nd and
3rd layers of the framework), then extended up to the portfolio
level and additional programs are now being transitioned in.
One other piece worth noting is that we are not yet
describing a full enterprise-level SAFe implementation. It is being actively explored, and what is
known so far is that it will require 2 additional portfolio layers as the
various groups are brought together. I
will be describing an implementation keeping 3 programs or delivery groups and roughly
200 people busy, but the eventual target involves more than 5,000.
