High-yield PMI-ACP review for key rules, traps, decision cues, formulas, and final-week reminders.
Use this as your last-mile PMI-ACP® review. Pair it with the Syllabus for coverage and Practice for speed.
For exam format and official policy details, see Overview.
flowchart TD
A["vision and outcomes"] --> B["ordered backlog"]
B --> C["slice small and define acceptance"]
C --> D["deliver an increment"]
D --> E["review, measure, and learn"]
E --> F["adapt backlog and process"]
F --> C
If you can state these three items from a question stem, you are usually close to the best answer:
| If the scenario is really about… | Better answer pattern | Weaker answer pattern |
|---|---|---|
| uncertainty | use a small experiment or increment to learn fast | plan longer before learning |
| overload or delay | visualize work, reduce WIP, and remove bottlenecks | push more work into the system |
| conflict or low trust | coach, facilitate, and surface the issue | command, assign blame, or escalate too early |
| customer doubt | get earlier feedback from the real user or stakeholder | wait until final acceptance |
| changing needs | adapt backlog, scope, or plan while preserving quality | protect the original plan from learning |
| Domain | Nature of work | Better response pattern |
|---|---|---|
| Clear | known knowns | sense -> categorize -> respond |
| Complicated | known unknowns | sense -> analyze -> respond |
| Complex | unknown unknowns | probe -> sense -> respond |
| Chaotic | crisis | act -> sense -> respond |
If requirements are changing and learning is still needed, PMI-ACP usually favors actions that reduce uncertainty quickly.
| Mode | Use when | Primary output |
|---|---|---|
| training | there is a knowledge gap | shared skill baseline |
| coaching | behavior or performance needs to improve | better team outcomes |
| mentoring | longer-term role growth matters | broader capability development |
| Situation | Better agile response | Weak response |
|---|---|---|
| team is capable but blocked | coach and remove impediments | assign tasks from above |
| disagreement is surfacing | facilitate the conversation and shared understanding | shut debate down quickly |
| recurring behavior issue | make expectations and feedback visible | rely on frustration and informal blame |
| Element | Purpose | Common exam cue |
|---|---|---|
| Sprint Goal | focus on outcome | supports scope trade-offs |
| Definition of Done | shared quality gate | avoids “done-ish” increments |
| Review | feedback and value check | validates assumptions |
| Retrospective | system improvement | fixes root causes, not just symptoms |
Anti-patterns: skipping refinement, using daily sync as status theater, treating the Sprint plan as a contract.
| Metric | What it measures | Better use |
|---|---|---|
| cycle time | start to finish for an item | improving flow |
| lead time | request to delivery | customer experience |
| throughput | items finished per time | capacity and forecasting |
| WIP | work in progress | spotting overload |
| Signal | Better move | Weak move |
|---|---|---|
| cycle time is rising | inspect bottlenecks and reduce overload | add more concurrent work |
| throughput varies heavily | forecast using ranges and recent evidence | promise precise delivery dates |
| WIP keeps climbing | finish before starting more | reward everyone for staying busy |
| Better slicing patterns | Weaker slicing patterns |
|---|---|
| thin vertical slices | layer-only slices such as UI-only or DB-only |
| workflow step or user scenario slices | large component bundles |
| happy path first, then edges | giant all-features batch |
Spikes are for learning. Timebox them and turn the result into a backlog decision.
| Need | Technique | Why |
|---|---|---|
| fast sizing | T-shirt or relative sizing | low overhead |
| shared assumptions | planning poker | reveals disagreement |
| large set estimate | affinity estimation | scales well |
Estimates are for decisions. PMI-ACP usually favors useful ranges and shared understanding over false precision.
Fix pattern: visualize flow -> identify bottleneck -> reduce WIP -> remove constraint -> repeat.