The crew at the Software Process and Measurement Blog and Podcast is off on a bit of an adventure. We are sharing the conclusion of Made to Stick to provide some reading while we recharge. We will return on December 7th with the next installment of  How To Be A Stoic by Massimo Pigliucci. 

Today we tackle two chapters of Actionable Agile Metrics Volume II, Advanced Topics in Predictability. We begin with Chapter 9 – VoP, VoC, and Predictability which sums up Section I, Variability and Predictability. Chapter 10, titled Monte Carlo Simulation, Revisited, begins Section II, Advanced Monte Carlo Simulation and Predictability.

Chapter 9 – VoP, VoC, and Predictability

The concepts of the Voice of the Process (VoP) and Voice of the Customer (VoC) represent formative ideas in my view of work and quality. The time I spent studying Deming and Scherkenbach introduced me to these concepts (consider reading Scherkenbach’s Road to Continual Improvement which includes several deep dives on the topic). Chapter 9 links the Process Behavior Chart (PBC) which is at the heart of Section I and uses it as a quantitative visualization of the VoP.

The VoP represents the capability or capacity of the process; the PBC illustrates what the process can predictably deliver. The VoC represents the expectations and the promises given. Promises or policies to deliver that do not match capabilities destroy trust. The VoC will not be positive as the team will not be able to meet expectations or commitments. Work intake processes and techniques are integral to using tools like the PBC to match the VoP and VoC. If there is no control over what and when work enters a team or team of teams chaos will ensue.  

Generating a PBC provides the basis for understanding capabilities from a quantitative perspective. We would all agree that other perspectives are important; however, this is the single point of view that can be shared in perfect transparency.  If a team knows their capability they can make better decisions on where and when to commit to work. As Vacanti notes VoP and VoC need to be aligned. 

All human processes have some degree of variance. Software development in all versions includes many moving parts and people, and usually, some level of ambiguity yields variable performance. Combining variability and fixed timeboxes causes misalignment. Misalignment puts pressure on behavior when teams try to fit work into arbitrary timeboxes.  We will return to the impact on behavior in the future. In the interim, be aware that when adopting a timebox in Scrum the time box and team capabilities need to be synchronized.

To wrap up this section let’s quote Vacanti:

 “Whenever your VoC is misaligned with your VoP, you have three options for improvement: 1. Fundamentally re-engineer your process to attempt to meet targets. 2. Change your process targets or specifications; or, 3. Do Both.” The option you don’t have is to do nothing and prosper.

Chapter 10 – Monte Carlo Simulation Revisited

Chapter 10 tackles Monte Carlo Simulation (MCS). I am a huge fan of MCS and use it all of the time. All work requires forecasting the future and answering questions like when, how many, and how much. The concept of yesterday’s weather is used as a forecasting technique in Agile. It is easily implemented and is not always sufficient. For complexity or risky work, there is no Oracle of Delphi that provides a single, unambiguous answer to any of those questions. The answers are always a range of probabilities. In most circumstances, each possible outcome is more or less probable than the next. Uncertainty makes it difficult to talk about when, what, and how much. Monte Carlo Simulation provides a way to handle answering questions with significant uncertainty in the inputs that influence the outcome of the work so you can have “when, what, and how much” conversations.

The one reminder in this chapter that stuck with me is that the outcome of Monte Carlo Simulations looks like the past. As Vacanti notes all probabilistic forecasting methods are based on historical data which means they are a reflection of the past. When using any technique to forecast the sample and context needs to be considered. Since the past is such a major input into the simulation, shocks to the system can create scenarios where the past doesn’t predict the future. Shocks are a problem when using yesterday’s weather also. The discussion of context and sampling leads us to Chapter 11 – next week.

Buy a copy and get reading – Actionable Agile Metrics Volume II, Advanced Topics in Predictability.  

Week 1: Re-read Logistics and Preface – https://bit.ly/4adgxsC

Week 2: Wilt The Stilt and Definition of Variation – https://bit.ly/4aldwGN

Week 3: Variation and Predictability  – https://bit.ly/3tAVWhq 

Week 4: Process Behavior Charts Part 1 – https://bit.ly/3Huainr

Week 5: Process Behavior Charts Part 2 – https://bit.ly/424O5Wc 

Week 6: How Much Data? – https://bit.ly/47GVP24 

Week 7: Detecting Signals – https://bit.ly/3SjwfdO 

Week 8: XmR Charts and the Four Basic Metrics of Flow – https://bit.ly/48j5AU9 

Week 9: Myths and Other Considerations – https://bit.ly/3SvfgVU