Capstone: from stage authorization to closure, with exam strategy

This capstone runs the four delivery processes end to end on the second case used throughout this track: Harbour City Libraries, the invented public library system installing self-service kiosks, RFID tags and online room booking across its six branches. PRIN-140 took it from the library board's mandate to an authorized project: four stages of 30, 360, 370 and 340 thousand, a total of 1,100 thousand against a ceiling of 1.2 million. PRIN-131 controlled its pilot branch stage up to the moment, in week 15 of the stage, when a kiosk firmware fault pushed the forecast stage end to week 29.1 against a 28-week limit. Here we pick up the story at that exception and follow it through the stage boundary, the third stage and closure. The lesson ends with exam strategy for questions on the delivery processes.

Step 1: the exception and the exception plan

The project manager's exception report recommended a stage exception plan with a revised time tolerance, because measuring patron uptake over a representative period was the whole point of the pilot, and the kiosk vendor was bearing the cost of the firmware fix. The board met within the week. It confirmed that project tolerances could absorb a delay of just over a week, asked the vendor's account manager to confirm the recertification date in writing, and requested an exception plan.

The project manager produced the exception plan in managing a stage boundary. It covered weeks 15 to the end of the stage: kiosk recertification in weeks 15 to 16.5, installation and training shifted accordingly, and the 10-week measured pilot running to about week 29.1. The cost rose by 12 thousand, mainly for extra staff cover during the shifted training, giving a stage cost of 372 thousand, within the existing cost tolerance of +5 % (378 thousand). The time limit was revised to week 30. The board authorized the exception plan, and the project manager resumed controlling the stage against it.

Step 2: the pilot stage boundary

The measured pilot ended at week 29. Self-checkout reached 58 % of loans in the last four weeks against a forecast of 60 %, and room booking errors fell to almost zero. The project manager had started managing the stage boundary at week 25. The end stage report showed the stage at 372 thousand and week 29.1, within the exception plan's tolerances, and the kiosk accessibility off-specification from PRIN-131 resolved by a software update. The updated business case used the measured uptake: staff time freed was slightly below forecast, but the project remained clearly justified. The next stage plan covered three more branches in 20 weeks for 370 thousand, using the pilot's actual tagging rate, which had settled at about 10,300 items a week with helpers.

The board authorized stage 3 with one condition: because uptake was the key benefit driver, the project manager was to report self-checkout share by branch in every highlight report. The chief librarian, as executive, also asked the senior user to brief branch staff associations before the rollout, because staff concerns about job changes had surfaced in the pilot.

Harbour City Libraries' accessibility criteria for kiosks and the booking system follow provincial accessibility standards, such as Ontario's AODA standards where they apply, and the library's own policy. The end stage report confirmed that the software update had resolved the off-specification for screen reader support and kiosk height for wheelchair users, and the accessibility review was repeated in each new branch.

Step 3: controlling stage 3

Stage 3 had three work packages: kiosk installation by the vendor, RFID tagging by a temporary team under the library's circulation manager, and training and change support under the branch services manager. The project manager authorized them in the first week, using pilot data for estimates and tolerances. Checkpoint reports were weekly. In week 8 of the stage, branch 3's planned renovation of its entrance was delayed by the city's facilities department, so the kiosks could not be installed in their planned location on time. The project manager examined the issue: swapping the order of branches 3 and 4 kept the stage on schedule at no extra cost, which was within the project manager's authority. The work packages were adjusted, the change recorded and the board told in the next highlight report. No escalation was needed.

Step 4: the final stage and closure

Stage 4 covered the last two branches and closure. The project manager started closing the project when the final kiosks were commissioned. Preparing planned closure checked the acceptance criteria in the project product description: kiosks in all six branches, at least 95 % of the circulating collection tagged, room booking online for every branch, accessibility criteria met, and staff trained. All were met except tagging, which reached 96 % of the circulating collection but left a special collection of rare and fragile items untagged, by design, as agreed by the senior user. The remaining 4 % of the general collection, mostly items out on loan during tagging, became a follow-on action for circulation staff.

stages = {"Initiation": (30, 29), "Pilot branch": (360, 372), "Branches 2-4": (370, 366), "Branches 5-6 and closure": (340, 343)}
ceiling, base = 1200, sum(b for b, a in stages.values())
actual = sum(a for b, a in stages.values())
for k, (b, a) in stages.items():
    print(f"{k:26} baseline {b:4}  actual {a:4}  variance {a - b:+4}")
print(f"total {actual} against baseline {base} ({(actual - base) / base:+.1%}); headroom to the {ceiling} ceiling: {ceiling - actual}")
tagged, collection = 0.96, 1.00
print(f"circulating collection tagged: {tagged:.0%} (criterion 95 %) -> {'met' if tagged >= 0.95 else 'not met'}")

Initiation                 baseline   30  actual   29  variance   -1
Pilot branch               baseline  360  actual  372  variance  +12
Branches 2-4               baseline  370  actual  366  variance   -4
Branches 5-6 and closure   baseline  340  actual  343  variance   +3
total 1110 against baseline 1100 (+0.9%); headroom to the 1200 ceiling: 90
circulating collection tagged: 96% (criterion 95 %) -> met

The project cost 1,110 thousand, 0.9 % above its baseline and well inside the ceiling the library board set. Handover went to three groups: library IT took the kiosks, RFID system and booking platform, with the vendor support contract transferred to its budget; circulation took the procedures and the follow-on tagging; branch services took the training materials. The end project report recorded the uptake at 61 % across all branches in the final month, the staff time freed, the two exceptions and the lessons, and it updated the benefits management approach: staff time freed and patron satisfaction would be reviewed at six and twelve months by the chief librarian. The board authorized closure and the library board received the closure notification.

Exam strategy for the delivery processes

PRINCE2 exams test the processes heavily, and especially the questions of who does what, when and with which products. The Foundation exam has 60 multiple-choice questions in 60 minutes, closed book, with a pass mark of 36 (60 %). The Practitioner exam, as described in 2026 study guides, gives 150 minutes for scenario-based questions worth 70 marks, with a pass mark of 60 % (42 marks). It mixes one-mark multiple-choice questions with three-mark matching questions, and only the official PRINCE2 7 manual may be used, as a printed copy or the official digital edition, with handwritten notes and permanent tabs allowed in the printed book. Check the current rules with PeopleCert when you book.

In the Practitioner exam, time is the scarce resource. Matching questions are worth three marks each and usually take longer, so it pays to plan time by marks rather than by question count. The code below shows a simple time plan.

total_minutes, marks = 150, 70
reserve = 15                     # minutes kept for review at the end
per_mark = (total_minutes - reserve) / marks
for kind, m in [("one-mark question", 1), ("three-mark matching question", 3)]:
    print(f"{kind:29} about {per_mark * m:.1f} minutes")
print(f"pass mark {0.6 * marks:.0f} of {marks}")

one-mark question             about 1.9 minutes
three-mark matching question  about 5.8 minutes
pass mark 42 of 70

With 15 minutes kept for review, each mark is worth just under two minutes: about two minutes for a one-mark question and nearly six for a three-mark matching question. If a question is taking much longer, mark your best answer, flag it and move on. Read the scenario for the facts that decide the answer: who raised the problem, which tolerance is threatened, which stage the project is in and what has already been authorized.

A last word on the processes

The processes can look like a lot of procedure. In practice, they are a small number of ideas applied consistently. Work is delegated through agreements with tolerances. Each level reports regularly and escalates only forecast breaches of tolerances it does not control. Every stage ends with a decision based on evidence. Every project ends on purpose, with its products handed over, its lessons passed on and its remaining benefits owned. Ridgeline Water and Harbour City Libraries applied those ideas with different scales of paperwork, and both reached a controlled end.