9.2 Inspect and Adapt (I&A) Event & PI Performance Review

9.2 Inspect and Adapt (I&A) Event & PI Performance Review

Executive Summary: The Inspect & Adapt (I&A) event is a significant, cadence-based milestone held at the conclusion of every Planning Interval (PI) during the IP iteration. Attended by the entire Agile Release Train (ART) and executive stakeholders, the I&A event evaluates the actual business value delivered against commitments, examines systemic process health through quantitative and qualitative flow metrics, and drives relentless organizational improvement. The primary quantitative health indicator produced is the ART Predictability Measure, which benchmarks delivery reliability within an optimal target band of 80% to 100%.

---

Architecture and Execution of the Inspect & Adapt (I&A) Event

In accordance with the foundational Lean-Agile mindset of relentless improvement (one of the four SAFe Core Values), an Agile Release Train must pause at regular intervals to inspect its systemic performance and adapt its processes.

Event Profile and Logistics

• Cadence: Occurs once per Planning Interval, typically during the final week of the IP iteration immediately preceding the next PI Planning session.
• Duration: Typically half a day to a full day (3 to 8 hours), depending on the size and complexity of the ART.
• Participants: Mandatory attendance by the entire ART (50 to 125+ practitioners), including:
  • All Agile Team members (developers, testers, analysts)
  • Scrum Masters / Team Coaches (SM/TC)
  • Product Owners (POs)
  • Product Management (PM)
  • System Architects and Engineering Leads
  • Release Train Engineer (RTE)
  • Business Owners and key enterprise executives
  • Customer and user representatives (where feasible)

---

The Three Distinct Parts of the I&A Event

The Inspect & Adapt event follows a strict, sequential three-part agenda designed to move the ART from objective demonstration to empirical measurement and systemic problem solving:

+-----------------------------------------------------------------------------------+
|                        INSPECT & ADAPT (I&A) AGENDA FLOW                          |
+-----------------------------------------------------------------------------------+
|  PART 1: THE PI SYSTEM DEMO (1 to 2 hours)                                        |
|  • End-to-end demonstration of the integrated solution delivered across the PI   |
|  • Attended by Business Owners, PM, customers, and all ART teams                  |
|  • Business Owners assign Actual Business Value (1 to 10) to team PI Objectives   |
+------------------------------------------+----------------------------------------+
                                           |
                                           v
+-----------------------------------------------------------------------------------+
|  PART 2: QUANTITATIVE & QUALITATIVE MEASUREMENT (45 to 60 minutes)                |
|  • RTE presents ART-wide flow metrics (Velocity, Time, Load, Efficiency)          |
|  • Calculate the ART Predictability Measure against target (80%–100%)    |
|  • Review quality metrics (escaped defects, test coverage, technical debt)        |
+------------------------------------------+----------------------------------------+
                                           |
                                           v
+-----------------------------------------------------------------------------------+
|  PART 3: RETROSPECTIVE & PROBLEM-SOLVING WORKSHOP (1.5 to 3 hours)                |
|  • Systemic ART retrospective on organizational impediments                       |
|  • 6-step root-cause analysis (Ishikawa fishbone, 5 Whys, Pareto 80/20)           |
|  • Actionable improvement items committed to the upcoming PI Backlog              |
+-----------------------------------------------------------------------------------+

Part 1: The PI System Demo

• Scope: Unlike the bi-weekly System Demo that inspects the working increment of an individual two-week iteration, the PI System Demo demonstrates the cumulative, end-to-end integrated solution delivered by the entire ART across all iterations of the PI.
• Environment: The demonstration must occur in a production-equivalent or staging environment that validates true end-to-end integration, non-functional requirements (NFRs), and system security.
• Business Owner Valuation: During the PI System Demo, Business Owners evaluate the delivered solution against the original PI Objectives established on Day 2 of PI Planning. For each objective (both committed and uncommitted), Business Owners award an Actual Business Value (BV) score from 1 to 10, reflecting the real business utility realized.

Part 2: Quantitative and Qualitative Measurement

• Flow and Predictability Review: The Release Train Engineer presents quantitative ART metrics collected throughout the PI.
• Key Flow Metrics: The train reviews Flow Velocity (completed story points/work items), Flow Time (elapsed time from start to completion), Flow Efficiency (active work time vs. wait time), and Flow Load (work-in-process).
• ART Predictability Measure: The RTE consolidates the Actual Business Value scores awarded by Business Owners to compute the official ART predictability score.

Part 3: Retrospective & Problem-Solving Workshop

• Systemic Reflection: The ART conducts a brief, train-wide retrospective to surface broad systemic friction points that span beyond individual team boundaries.
• Root-Cause Analysis: The train transitions into the structured 6-step Problem-Solving Workshop (detailed in Section 9.3) to isolate root causes and produce concrete backlog improvement items.

---

The ART Predictability Measure In-Depth

The ART Predictability Measure (abbreviated PPM, a holdover from its pre-SAFe 6.0 name, Program Predictability Measure) is the flagship quantitative metric of an Agile Release Train. It provides business leaders with an empirical, objective evaluation of how reliably the ART delivers on its business commitments.

The Mathematical Formula

The ART Predictability Measure is calculated using the following ratio:

$\text{ART Predictability Measure} = \left( \frac{\sum \text{Total Actual Business Value Achieved}}{\sum \text{Total Planned Business Value Committed}} \right) \times 100\%$

The Target Performance Range: 80% to 100%

• Healthy Performance (80%–100%): A predictable ART consistently scores between 80% and 100%. This range demonstrates that the train plans realistically, manages risks effectively, negotiates dependencies proactively, and absorbs minor delivery shocks without sacrificing core commitments.
• Under-Delivery (< 80%): Predictability consistently below 80% indicates chronic delivery dysfunction: over-committing during PI Planning, unmanaged external dependencies, unstable technical architecture, high technical debt, or frequent scope thrashing by leadership.
• Over-Delivery (> 100%): While delivering more than planned may appear positive, an ART that consistently scores 120% or 130% is likely "sandbagging"—artificially under-committing during PI Planning to guarantee high scores. However, a score slightly above 100% (e.g., 105%) is entirely healthy when an ART delivers all committed objectives plus several uncommitted objectives.

---

The Role and Mechanics of Uncommitted Objectives in PPM

During PI Planning, teams encounter high-risk features, uncertain technical enablers, or external dependencies outside their control. Rather than overloading their plan, teams classify these items as Uncommitted Objectives (previously termed "Stretch Objectives" in older SAFe versions).

The mathematical treatment of Uncommitted Objectives in the PPM is a frequent, critical subject on the SAFe POPM exam:

1. Uncommitted Objectives Have Planned Value: During PI Planning, Business Owners assign a Planned Business Value score (1 to 10) to uncommitted objectives so teams understand their relative business priority.
2. Excluded from the Denominator (Committed Planned BV): When calculating the PPM, the planned business value of uncommitted objectives is strictly excluded from the denominator. The denominator consists only of the planned business value of committed objectives.
3. Included in the Numerator (Actual BV Achieved): If the team successfully delivers an uncommitted objective during the PI, the Actual Business Value awarded by Business Owners is added into the numerator.
4. The Strategic Rationale:
   • Protection Against Uncontrollable Variance: Teams are never penalized if an uncommitted objective is not completed. Because its planned value is excluded from the denominator, failing to deliver an uncommitted item does not reduce the predictability percentage.
   • Incentivizing Stretch and Exploration: Including achieved uncommitted value in the numerator provides upside credit. This mechanism allows a team to achieve over 100% predictability if they deliver all committed work and successfully execute uncommitted goals.

---

Worked Example: ART Predictability Measure Calculation

To understand how the PPM operates across an Agile Release Train, consider the following worked example for a train consisting of four Agile teams:

Team-by-Team Objective Scoring Table

Exam nuance — two ways to roll the number up. The SAFe glossary defines the ART Predictability Measure as an average of the achievement score for each team on the ART for the PI. Averaging the four team scores gives (116.0 + 60.0 + 113.0 + 83.3) / 4 = 93.1%. Pooling every team’s points first, as the table above does, gives 90 / 97 = 92.8%. Both land in the healthy 80–100% band and the two methods rarely diverge much, but if a question asks for the ART measure by definition, average the per-team scores.

Step-by-Step Calculation Walkthrough

1. Calculate the Denominator (Total Planned Committed BV):

   • Team Alpha: $10 + 8 + 7 = 25$ (The uncommitted score of 6 is excluded).
   • Team Beta: $9 + 8 + 8 = 25$ (The uncommitted score of 5 is excluded).
   • Team Gamma: $9 + 8 + 6 = 23$ (The uncommitted score of 4 is excluded).
   • Team Delta: $9 + 8 + 7 = 24$ (The uncommitted score of 5 is excluded).
   • Total Denominator: $25 + 25 + 23 + 24 = \mathbf{97}$ points.

2. Calculate the Numerator (Total Actual BV Achieved):

   • Team Alpha achieved all committed objectives (earned 23) plus delivered their uncommitted Biometric Authentication Spike (earned 6), giving an actual total of $\mathbf{29}$ points.
   • Team Beta delivered partial value on Obj 2 and failed Obj 3; uncommitted Obj 4 was not delivered (earned 0). Total actual points: $\mathbf{15}$.
   • Team Gamma delivered all committed objectives (earned 23) plus achieved partial value (3 points) on uncommitted Obj 4. Total actual points: $\mathbf{26}$.
   • Team Delta delivered partial value across committed objectives; uncommitted Obj 4 was deferred (earned 0). Total actual points: $\mathbf{20}$.
   • Total Numerator: $29 + 15 + 26 + 20 = \mathbf{90}$ points.

3. Compute the ART Predictability Measure: $\text{ART PPM} = \left( \frac{90}{97} \right) \times 100\% = \mathbf{92.8\%}$

4. Business Interpretation:

   • The ART's consolidated score of 92.8% sits squarely within the healthy target range of 80% to 100%.
   • Notice the impact of uncommitted objectives: Even though Team Beta suffered significant delivery issues (60.0% predictability) and Team Delta deferred their uncommitted objective, Team Alpha and Team Gamma successfully delivered uncommitted value. This pulled the overall ART predictability up into the target band without unfairly penalizing the train for the uncompleted stretch items.

---

Quantitative Flow Metrics Evaluated in the I&A

In addition to the PPM, the RTE and leadership inspect six core SAFe Flow Metrics during Part 2 of the Inspect & Adapt event:

---

POPM Key Responsibilities in the I&A Performance Review

• Product Management: Facilitates the presentation of integrated business value during the PI System Demo; validates whether delivered Features fulfilled their Benefit Hypotheses; partners with Business Owners to calibrate future strategic expectations based on empirical PPM trends.
• Product Owners: Actively represent team achievements during the demo; provide factual narrative on why certain objectives were deferred or altered; support the team in identifying flow bottlenecks during the metric review.