Managing a stage boundary II: the end stage report and the board's decision

Lesson 7 covered the forward-looking work of a stage boundary: the next stage plan, the updated project plan and the updated business case. This lesson covers the backward look and the decision. The project manager reports how the stage went in an end stage report and sends it, with the next stage plan and the other updated products, to the board with a request to authorize the next stage. The board then makes its decision in directing a project. At Ridgeline Water this happens at week 38, at the end of the pilot.

The end stage report

The end stage report gives the board a summary of the stage just ending and the information it needs to decide on the next one. It is produced by the project manager near the end of each stage except the final one, where the end project report takes its place. It is written for busy people: it should say plainly how the stage performed against its plan and tolerances, what was delivered and to what quality, what changed, what was learned, and what the project manager recommends.

Reporting performance against tolerances

The heart of the end stage report is the comparison of what happened with what was approved. Because the board approved an exception plan during the pilot, the comparison is with the revised baseline, and the report says so clearly, with a short note on the original plan for context. The code below produces the comparison for all seven performance targets.

targets = [  # target, baseline, limit, actual, direction
    ("cost (thousand)", 1568, 1568 * 1.02, 1581, "max"),
    ("end week", 38, 39, 38, "max"),
    ("connections", 2800, 2750, 2812, "min"),
    ("daily read rate %", 98.0, 97.0, 98.4, "min"),
    ("largest single risk exposure (thousand)", 0, 100, 38, "max"),
    ("recycled meters by weight %", 95, 95, 97, "min"),
    ("leak savings vs forecast %", 100, 70, 85, "min"),
]
for name, base, limit, actual, d in targets:
    ok = actual <= limit if d == "max" else actual >= limit
    span = abs(limit - base) or 1
    used = (actual - base) / span if d == "max" else (base - actual) / span
    used = max(0.0, used)
    print(f"{name:40} baseline {base:>7.1f}  actual {actual:>7.1f}  limit {limit:>7.1f}  {'within' if ok else 'BREACH':7} tolerance used {used:4.0%}")

cost (thousand)                          baseline  1568.0  actual  1581.0  limit  1599.4  within  tolerance used  41%
end week                                 baseline    38.0  actual    38.0  limit    39.0  within  tolerance used   0%
connections                              baseline  2800.0  actual  2812.0  limit  2750.0  within  tolerance used   0%
daily read rate %                        baseline    98.0  actual    98.4  limit    97.0  within  tolerance used   0%
largest single risk exposure (thousand)  baseline     0.0  actual    38.0  limit   100.0  within  tolerance used  38%
recycled meters by weight %              baseline    95.0  actual    97.0  limit    95.0  within  tolerance used   0%
leak savings vs forecast %               baseline   100.0  actual    85.0  limit    70.0  within  tolerance used  50%

Every target is within tolerance. Cost used about 41 % of the revised cost tolerance, time none, and the read rate and recycling targets were beaten. Leak savings are below the forecast, using half of the band between the forecast and the 70 % threshold the board set, but still well clear of that threshold. The report notes, for context, that against the original pilot plan the stage cost 131 thousand more and took one week longer, for the reasons explained in the exception report from week 24.

Products and handovers at the boundary

The end stage report confirms which products have been completed and approved, which were accepted with concessions, and which have been handed over. Some projects hand products to operations at stage boundaries rather than all at the end; this is phased handover. Ridgeline handed the pilot district to customer operations at week 38: from the next billing cycle, those 2,812 customers are billed from smart meter readings. Customer operations formally accepted the meters, the billing interface on daily totals and the operating procedures, and the acceptance record is filed with the end stage report. Benefits from that district can now be measured in normal operation.

Lessons at the boundary

The end stage report includes the lessons from the stage. Where the lessons are substantial, or useful beyond the project, the project manager may also produce a lessons report and send it to the people in the organization who can act on it. Ridgeline's pilot lessons on legacy data quality and component lead times were relevant to other utility projects, so the project manager sent a short lessons report to the corporate project office, which added a data-quality check to the utility's project initiation template.

The request to authorize the next stage

With the end stage report, the next stage plan, the updated project plan, the updated business case and any changes to the PID, the project manager sends the board a request to authorize the next stage. The board then performs the directing activity of authorizing a stage or exception plan. Its questions are simple to state and not always easy to answer. Is the project still desirable, viable and achievable? Is the next stage plan realistic? Are the stage tolerances right? Are the risks acceptable? Is the project still within project tolerances, and if not, must the business layer be involved?

Ridgeline Water's boundary decision

The board met at week 38. The senior user asked whether customer operations could handle 16,800 new smart meter accounts during the winter billing peak; the project manager showed the phased installation schedule and the extra temporary staff in the plan. The senior supplier, Northline's director, confirmed the 14 crews in writing. The executive focused on the business case: the pilot evidence came from one district with older mains, and most of rollout stage 2 is in newer subdivisions. The board authorized rollout stage 1 with two conditions. First, leak savings must be measured in at least one newer subdivision during the stage, and reported at the next boundary as a key input to the rollout stage 2 decision. Second, highlight reports were to be weekly for the first four weeks, then fortnightly.

Because the rollout stage commits capital from the city's capital improvement plan, uses state revolving fund loan money and awards a large contract variation, the general manager took a staff report to council at its next meeting after the board's decision. Council approved the contract variation and noted the end stage report summary. The project plan had aligned the boundary with the council calendar, so the crews could mobilize at week 40 without waiting another month.

What the board should not do at a boundary

Boards sometimes treat stage boundaries as a formality: the next stage was in the project plan, so they authorize it without reading the report. That wastes the most important control PRINCE2 gives them. Others go to the opposite extreme and reopen every decision made during the stage. The board's job at a boundary is to decide on the future, using the evidence of the past. It may ask questions about the stage, but its decision is about whether the next commitment is justified. If the board needs assurance about specific claims in the report, project assurance should have checked them before the meeting.

Project assurance at the boundary

Before the board meets, project assurance reviews the boundary products on the board's behalf. For Ridgeline's pilot boundary, business assurance checked the leak savings calculation against the network team's raw data, user assurance confirmed with customer operations that the pilot district's acceptance records were complete, and supplier assurance checked that Northline's crew commitment for rollout was confirmed in writing. Their short notes went to the board with the end stage report, so the meeting could spend its time on the decision rather than on checking figures.