85% Complete and Nine Days to Report
Halvard Rail's signalling upgrade is reported monthly to a programme board and quarterly to a lender under a facility agreement. For three consecutive months the report has been green. Completion is reported at 85% using a method that counts a work package as complete once its first activity has started. The lender's quarterly pack is due in nine working days. An internal auditor has asked for the decision log: it exists, is five weeks out of date, and three significant decisions - including a design deviation approved verbally on site - were never recorded. The lead engineer for the interlocking subsystem left last month; her handover notes cover the hardware but not the configuration baseline. The contract requires records to be retained for ten years, and the current document set sits partly on a personal network drive. The programme director has asked you to remove the cost pressure section from the lender pack so it does not distract from the good news.
1. Is the 85% completion figure defensible? What would you replace it with, and how would you implement the change without destroying trust in the data?
Read the worked answer
No. Counting a package as complete when its first activity starts measures mobilisation, not progress. It guarantees an inflated early curve and an end-loaded one, which is exactly why three greens preceded a problem. It also makes covenant reporting to the lender misleading, which is a contractual exposure rather than a reporting nicety.
Replace it with:
- Physical percent complete by weighted milestones, or 0/100 for short packages and 50/50 only where the second half is genuinely low-risk. Weight by planned cost or effort so the total is defensible.
- Earned value at programme level - EV/PV by period, SPI and CPI - with the basis of measurement documented.
- Remaining duration and forecast completion date per package; the lender cares more about date and cost at completion than about a percentage.
Implementing it without destroying trust:
- Run old and new methods in parallel for one cycle and show the reconciliation - the re-based figure will drop sharply, and the drop must be explained before it appears.
- Tell the board and the lender in the same message: the previous method overstated progress, here is the corrected position, here is the recovery plan. Presenting it as a sudden deterioration rather than a basis change invites a credibility problem.
- Put the method into the PMO standard so it cannot drift back.
- Do not restate history you cannot evidence; state the basis change with an effective date.
2. What are your obligations regarding the programme director's request, and how do you handle it?
Read the worked answer
The obligation: report accurately and completely to those entitled to the information. The lender pack is governed by the facility agreement, and omitting a known cost pressure may breach a representation and will at minimum mislead a party relying on the report. The board has the same claim to a complete picture.
Response:
- Clarify what is actually being asked. 'Remove' may mean 'shorten and contextualise', which is a legitimate editorial request. Offer a one-page summary: position, cause, range, actions, decision needed. Directors often want less noise, not less truth.
- If the instruction is genuinely to omit a material cost pressure, decline in writing, briefly and without moralising: 'I can't sign a pack that omits a known cost pressure of this size; I can present it in this form.'
- Escalate - finance, internal audit, the board chair or the lender relationship owner. One of them owns the disclosure decision; the PM does not.
- Record the request and the response in the decision log.
- Ask for the instruction in writing if it is repeated. Directors rarely commit that to email; if they do, the record protects both you and the organisation.
Working test: would the report still be defensible if read aloud in a room containing the lender and the auditor? If not, it is not an editorial choice.
3. Set out a 10-working-day remediation plan for the documentation and records issues, prioritised by risk.
Read the worked answer
Prioritise by consequence and irreversibility, not tidiness.
Days 1–2 - stop the bleeding
- Freeze changes to the configuration baseline.
- Take a verified copy of the personal-drive documents onto controlled storage today. Retention and potential legal-hold obligations make this the most urgent item.
- List what is missing: the interlocking configuration baseline, the three verbal decisions, five weeks of log entries.
Days 3–5 - reconstruct the decision log
- Recover the three decisions from emails, site diaries, meeting notes and attendees; record each as 'reconstructed on [date] from [source]' with the decision owner confirming. Do not backdate.
- Interview the departed engineer's peers and the contractor; establish configuration as-built versus as-designed. Flag anything unknown as an open technical risk, not a documentation issue.
Days 6–8 - put controls in place
- Decision log with a named custodian and a 48-hour logging rule; a standing agenda item at every meeting.
- Document register with retention class (ten years), storage location and owner; move everything off personal drives and revoke access.
- Handover checklist as a mandatory gate for leavers.
Days 9–10 - assure and report
- Sample-audit ten documents for completeness and location.
- Report residual gaps and the open technical risk to the board and the auditor with dates and owners.
The objective is to convert an unknown into a managed, dated gap - and to fix the process that allowed it.