Scrum and Kanban are team-level frameworks. They work well for one team delivering one product. The exam tests what happens when an organization needs multiple agile teams working on the same product — or when it needs agile delivery inside a predictive governance structure.
Four failures occur predictably without a scaling mechanism: unresolved cross-team dependencies, misaligned priorities across separate backlogs, high-risk integration events at the end of a release cycle, and redundant work caused by lack of cross-team visibility. Tonight’s three mechanisms — Scrum of Scrums, SAFe®, and LeSS — each address these failures with different levels of structure.
Scrum of Scrums is not a separate framework. It is a coordination meeting between representatives from multiple Scrum teams working toward a shared goal. One representative per team — typically the Scrum Master or a designated member — meets 2–3 times per week, or daily during high-coordination phases. The purpose is to surface cross-team dependencies, integration risks, and shared blockers before they become delivery failures.
Multiple Scrum teams + same product + coordination question → Scrum of Scrums is often the best exam answer — but context still matters. Who attends? Team representatives, not the PM or management directly. Output: dependencies and blockers resolved or escalated — not status reports.
Scrum of Scrums is NOT a status meeting, NOT a replacement for team-level Daily Scrums, and NOT run by the PM or sponsor. Management chairing a cross-team standup and asking for status = anti-pattern.
SAFe® (Scaled Agile Framework®) is the most widely adopted enterprise agile framework. PMI references it specifically in the Instructor Guide as a governance example for agile-based projects. The PMP exam does not test SAFe® certification-level detail — it tests reasoning about scaled coordination and governance. Know these five concepts.
| Concept | Definition | Exam Signal |
|---|---|---|
| Agile Release Train (ART) | A long-lived team of agile teams — typically 50–125 people — that plans, commits, and executes together in Program Increments. | Multiple teams planning quarterly together = ART. SAFe®’s primary delivery vehicle. |
| Program Increment (PI) | Fixed 8–12 week timebox. Structured as 4 development iterations + 1 Innovation & Planning (IP) iteration. | PI ≈ one quarter. IP iteration = planning + hardening — NOT for finishing deferred sprint work. |
| PI Planning | 2-day event where all ART members plan the next PI, identify dependencies, and commit to PI objectives. | ESSENTIAL — cannot be skipped. Remote PI Planning is acceptable. Async PI Planning is an anti-pattern. |
| Value Stream | The sequence of steps used to deliver value — solution or operational — from concept to customer. | SAFe® organizes around value streams, not functions. Key cultural shift from traditional org design. |
| Release Train Engineer (RTE) | Servant leader and chief Scrum Master for the ART. Facilitates ART events, removes impediments, coaches teams. | RTE = PM equivalent at the ART level. The PM may hold this role in a hybrid environment. |
Risks identified during PI Planning are categorized using ROAM: Resolved (eliminated before the PI starts) · Owned (assigned to a specific person to monitor and act) · Accepted (known but cannot be mitigated — team proceeds with awareness) · Mitigated (a plan reduces impact or probability). When asked what happens to risks in PI Planning: ROAM.
LeSS scales Scrum with minimal added structure. One Product Owner, one product backlog, one sprint cadence across all teams. All teams synchronize on sprint boundaries and participate in a cross-team Overall Retrospective each sprint.
| Dimension | LeSS | SAFe® |
|---|---|---|
| PO structure | One PO, one backlog | Multiple POs at team, program, and portfolio levels |
| Overhead added | Minimal — a few new meetings | Significant — ART, PI, new roles (RTE) |
| Planning cadence | Sprint planning | Quarterly PI Planning |
| Best for | 2–8 teams on one product | 50–125+ people, complex programs |
| Philosophy | Keep Scrum pure at scale | Enterprise agile transformation |
LeSS signal: simplicity, minimal overhead, one PO, one backlog, multiple teams. SAFe® signal: enterprise scale, quarterly planning, 50+ people, governance board. Simplicity at scale → think LeSS over SAFe®.
| Mechanism | Team Size | Structure Added | Exam Signal |
|---|---|---|---|
| Scrum of Scrums | 2–5 teams | One coordination meeting | Multiple Scrum teams + shared product + dependency question |
| LeSS | 2–8 teams | A few new meetings, one overall retro | Multiple teams + simplicity + minimal overhead emphasized |
| SAFe® | 8–20+ teams | ART, PI Planning, new roles (RTE), value streams | Enterprise, quarterly planning, 50+ people, governance board |
Most real-world projects are hybrid. The Stacey Complexity Model maps projects across two axes — how well requirements are understood and how well the technology or approach is understood — to inform the delivery decision.
| Zone | Requirements | Technology | Approach |
|---|---|---|---|
| Simple | Clear and agreed | Known | Predictive |
| Complicated | Mostly clear with analysis | Known with expertise | Predictive + expert judgment |
| Complex | Uncertain, evolving | Novel or experimental | Adaptive (Scrum / Kanban) |
| Chaotic | Unknown | Unknown | Act → Sense → Respond |
| Disorder (center) | Mixed | Mixed | Hybrid — tailor to context |
Regulatory compliance + evolving requirements → Hybrid. Fixed-price contract + sprint delivery → Hybrid (contract is a predictive constraint). Technical innovation + unclear scope → Complex → Adaptive. Well-defined infrastructure upgrade → Complicated → Predictive + experts. The exam describes characteristics, not quadrant names.
| Pattern | How It Works | When to Use |
|---|---|---|
| Predictive Shell, Adaptive Core | Project governed predictively (scope, budget, milestones). Dev teams work in sprints within each phase. | Regulated environments, fixed-price contracts, large enterprise programs. |
| Adaptive Planning, Predictive Delivery | Requirements managed adaptively (backlog, stories). Final product delivered as a single release. | Product dev where requirements evolve but the customer needs a single go-live. |
| Parallel Workstreams | Some workstreams run predictively (infrastructure, compliance), others adaptively (app, UX). | Technology programs where infrastructure must be planned but features can iterate. |
| Phase-Gate with Agile Development | Formal phase gates with exit criteria govern the program. Each phase uses sprints internally. | Capital-intensive programs requiring formal investment decisions at milestones. |
Predictive Shell, Adaptive Core. Fixed scope + fixed deadline + formal reporting (predictive) + sprints + backlog + PO (adaptive inside each phase). The PM manages predictive commitments while protecting the team’s ability to self-organize within sprints.
Each workstream is measured in its native language. Forcing one lens onto both creates a distorted picture.
| Predictive Metrics (EVM) | Adaptive Metrics |
|---|---|
| SPI — Schedule Performance Index (EV ÷ PV) | Sprint Velocity — story points completed per sprint |
| CPI — Cost Performance Index (EV ÷ AC) | Release Burn-Up — features delivered vs. planned |
| EAC — Estimate at Completion | Cycle Time — start to done per item |
| Milestone Status — on-track / at-risk / delayed | PI Objectives Status — committed vs. achieved |
EVM requires a fixed scope baseline. Scrum teams have a prioritized backlog — not a fixed baseline. Forcing SPI and CPI onto an adaptive workstream produces meaningless numbers.
| Predictive Environment | Agile / Hybrid Environment |
|---|---|
| Manages tasks and resources directly | Removes impediments; enables team self-management |
| Owns the project schedule | Manages release plan + PI objectives; team owns the sprint |
| Controls scope through CCB | Protects sprint scope; PO manages backlog changes |
| Reports status to stakeholders | Creates transparency via radiators (boards, burndowns, CFDs) |
| Manages risk via risk register | Uses retros, impediment logs, and spikes continuously |
PM assigns sprint tasks · PM runs the Daily Scrum as a status meeting · PM processes backlog additions through a CCB in an agile context. All three are wrong answers when they appear in exam scenarios.
Ang Fen is managing two parallel workstreams: (1) Oracle HCM implementation — predictive, fixed-scope, regulatory go-live deadline; and (2) digital wellness app — Scrum team, 2-week sprints. Eugene Lowe asks for a single integrated status report.
HCM workstream (predictive): Milestone status, SPI, CPI, top 3 active risks, issues requiring steering committee decision.
Wellness app (adaptive): Sprint velocity trend, release burn-up, PI objectives status, impediment log, upcoming sprint demo schedule.
PM insight: Each workstream in its native language. Ang Fen synthesizes the integrated picture for the steering committee — she does not force one lens onto both systems.
Tailoring answers: what delivery approach is right for this project? PMI mandates a reasoned, evidence-based decision — not a default. The suitability assessment considers three categories.
Read the scenario for six signals — two per category. Answers that say “always use Scrum” or “always use predictive” are almost always wrong. When all three categories signal adaptive, the answer is adaptive — not hybrid.
● HIGH PRIORITY
● MEDIUM PRIORITY
● OPTIONAL
Session 10 · PMP® Exam Prep Mastery · support@pmptrainingacademy.com