When a project audit becomes necessary
A project audit is rarely requested light-heartedly. It comes when management has lost confidence in the information it receives. Committee meetings follow one another, the indicators have been amber for months, each postponement is presented as the last. In the corridors, people hear a different version from the one in the dashboards.
Internal monitoring is no longer enough, for reasons that have nothing malicious about them. The project manager has an interest in believing in the recovery. The supplier defends its reading of the contract. The teams have said what they knew, sometimes several times, to no effect. The sponsor has staked their credibility. Everyone holds part of the facts, and nobody is in a position to piece them together without being suspected of pleading their own case.
The other moment is less dramatic and just as useful: the eve of a commitment. Before voting a delivery budget, signing with an integrator or authorising a cutover, an outside view checks that the project is ready and that the risks are known to those who decide.
Health check, milestone audit, recovery audit
The word “audit” covers three exercises that do not answer the same question. Confusing them leads to commissioning work that is too heavy, or too light.
| Type of review | Trigger | Question asked | Expected result |
|---|---|---|---|
| Health check | Programme calendar, with no particular alert | Does the project present risks that routine monitoring does not show? | Points of attention and preventive actions |
| Milestone audit | Upcoming commitment decision | Is the project ready to pass this stage? | Reasoned opinion, reservations and prerequisites |
| Recovery audit | Project clearly off track | Can the project still achieve its objective, and under what conditions? | Decision options and recovery plan |
The rest of this page describes the most comprehensive case, that of recovery. The other two use the same method, on a reduced scope.
The scope: seven angles of examination
A project almost never goes off track for a single reason. We examine it from seven angles, scaling the effort to the questions in the terms of reference.
- The rationale. Is the objective still valid? Do the expected benefits justify the remaining cost, and not the cost already incurred?
- Governance. Who decides, and are the decisions requested actually made? Many delays are decisions on hold.
- Scope. Are the requirements stable, and who has the power to refuse a new request?
- The schedule. Is the remaining work estimated by those who will carry it out? Is the critical path identified?
- Budget and contracts. What remains to be spent, and what do the suppliers’ commitments actually say?
- The solution. Can the architecture handle the planned load, and what is the state of the quality delivered? When these points require in-depth examination, we draw on our skills in information system architecture and software quality.
- The people. Does the team have the necessary skills and availability? Are the future users waiting for the result?
The method: cross-check, verify, confront
The value of an audit lies in the soundness of its findings. Three principles guide our work.
Cross-checking. No finding rests on a single source. A statement heard in an interview is confirmed by a document, a measurement or a second independent account; otherwise it is presented as a perception.
Verifying against evidence. We do not stop at committee minutes. We look at the backlog and the defect log, we ask for a demonstration of what works, we redo the calculation of the remaining work based on the pace actually observed.
Confronting. The findings are presented to the project team before being reported to the sponsor. The team can correct an error of fact or provide a missing element. The conclusions, however, remain ours.
Interviews are individual and statements are not attributed. This rule is announced at the launch, by the sponsor themselves.
Reporting and the recovery plan
Reporting comes in three stages: what we observed, what explains it, what you can decide. The options are always presented with their consequences in terms of time, cost and risk.
- Continue with limited adjustments, when the difficulties are contained.
- Refocus the scope, the trajectory or the team, when the objective remains valid but the plan no longer is.
- Suspend for the time needed to remove a prerequisite: an architecture decision, a renegotiation, a business decision.
- Stop, when the remaining cost exceeds the expected value. This is a legitimate option, which the audit must make it possible to put forward.
If the project continues, the recovery plan sets the measures for the first weeks, a trajectory reset on verifiable milestones and tightened governance for the duration of the recovery. On a programme, its implementation falls within programme management.
The pitfalls we help you avoid
- The audit without a sponsor. If nobody has the power to act on the conclusions, the report joins the previous ones. We check this point before accepting the scoping.
- The trial. Looking for a culprit guarantees defensive answers and deprives the audit of its best sources.
- The paper audit. Checking that documents exist says nothing about the state of the project. A test plan can be complete and the tests can still fail.
- The encyclopaedic report. A long list of findings of equal weight allows no decision. A few major causes, ranked, are more useful.
- Reasoning on sunk costs. What has been spent does not justify continuing. Only the remaining cost and the value still achievable count.
What shows that an audit has been useful
The first sign is immediate: an explicit decision is taken at the end of the reporting, and it is dated. An audit followed by a request for further analysis has often missed its target.
The other signs can be observed in the months that follow. The milestones of the recovery plan are met, or any deviation is reported without delay. Decisions that were on hold have been made. The project’s forecasts stop slipping from one committee to the next. Finally, the team considers that the report describes its situation accurately, which conditions everything else. When the audit reveals a lasting weakness in monitoring, the natural next step is strengthening the project management office.