Scaled Agile Framework Agilist (SAFe)

My experience as a Scrum Master has been enlightening. Being able to execute a project while adapting to the current client needs is exciting and beneficial… but it for sure is different from a lot of project thinking.

As I understood the concepts of SCRUM – I learned how to serve a team on a team level. But I was intrigued to understand how this concept could work in unison on an Enterprise level.  I had the opportunity to pursue a certification in SAFe. That stands for Scaled Agile Framework. It is amazing how all the parts work together.

Annotation 2020-08-29 143735My training was under Chad Manley at BCForward. Fascinating class followed by some extensive study and the requirement to pass an exam. The class was over 2 days and the prep for successfully passing the same took about 30 days for me.

There’s too much to cover here, but just as a sample, here are my notes on SAFe Principles.

SAFe Principles

#1   Take an Economic View

An understanding of economics drives decisions. Economic variables such as development cost, production cost, delivery lead time, and value – directly inform decision making.

Deliver Early and Often

Apply an Economic Framework

Lean Budgets and Guardrails (defined by LPM) (updated periodically)

Understanding solution economic trade offs

Leveraging Suppliers

Sequencing Jobs for the maximum Benefit (WSJF)


#2 Apply Systems Thinking

Everyone understands and commits to the common goals of the larger system. The whole is optimized instead of the parts.

Solution = System

Enterprise = System

Optimize Full Value Stream

Only management can change the system.


#3 Assume variability; preserve options

Decisions are delayed until the last responsible moment; alternatives are constantly and aggressively explored.

Set Based Design over Point Based

Coupled with Cadence based learning milestones


#4 Build incrementally with fast, integrated learning cycles

Cadence based learning cycles are used to gain knowledge, evaluate alternatives, and inform decision making.

Integration Points (uncertainty into knowledge – technical feasibility – sustainability)

  • Create knowledge from uncertainty
  • Occur by intent

Faster Learning by Faster Cycles

  • PDCA
  • True measure of system progress

#5 Base milestones on objective evaluation of working systems

Progress is measured by objective measures, rather than traditional phase gates.

Variability will find a way to express itself – pain point for phase gate development

Instead build system in increments to create value increment

Promotes frequent measurement by stakeholders


#6 Visualize and limit WIP, reduce batch sizes, and manage queue lengths

Small batches of work, controlled Work in Progress (WIP), and small queues ensures fast flow of value and learning.

Visualize and limit WIP

  • Overload confuses and slows work, like rush hour
  • Make WIP visible – kanban
    • Visualizing work can cause system problems to be addressed
    • Let limits – when met, no new work is taken on

Reduce batch sizes (u curve optimization)

      • Small batches go through system faster with reduced variability
      • Focus on reducing transactional costs
      • Primary purpose of DevOps

Manage Queue lengths

  • The longer the queue the longer the wait
  • Visualizing the work helps
  • Also
    • Team and Program backlogs short
    • WIP limits for each process step
    • Beware long term commitments


#7 Apply cadence; synchronize with cross-domain planning

Regular synchronization continually aligns all system builders and ensures all perspectives are understood and resolved.

Agile needs safety zone with freedom to innovate

Business needs confidence to operate

Overcomes Uncertainty of development

  • Cadence – pattern of events providing steady heartbeat of development
      • Align Development Cadence
  • Synchronization – multiple perspectives understood, resolved, integrated at same time
      • With Cross Domain Planning – limits plan deviation


#8 Unlock intrinsic motivation of knowledge workers

Knowledge workers exhibit curiosity and have fundamentally different motivations. Leaders are responsible for creating an environment in which these workers can thrive.

Leverage System View – collaboration is a system

Compensation – pay enough for people to focus on the work

  • Incentive based is limited either positive or negative

Connect aim of Enterprise with Worker’s Daily Work

  • Autonomy with Purpose
  • Mission
  • Minimum Possible Constraints

Create Mutual Influence

  • Workers heard and respected
  • Workers smarter and have more context than ever


#9 Decentralize decision making

Autonomy empowers individuals and enhances motivation. Leaders support decentralized decision-making by equipping teams and individuals with the knowledge and judgment needed to make good decisions.

Creating a reliable decision-making framework is critical to ensure fast flow of value.

Centralize Strategic (leadership makes choices, supported by stakeholders affected)

  • Infrequent – not urgent – need deeper consideration
  • Long-lasting – those difficult to change once made
  • Significant Economies of Scale – deliver large/broad benefits

Decentralize Everything Else (by workers w/local context and detailed knowledge)

  • Frequent – recurrent and common
  • Time-critical – with causes high cost of delay
  • Require local info – need specific context, local affect