A controls function has four parts. They only produce something useful when they run off the same set of numbers.
Separated, each one still generates a report. A schedule that nobody has reconciled to the cost model. A forecast that does not reflect the risk register. A risk analysis run against a network that has since been re-baselined. Every one of those is defensible in isolation and misleading in combination, and the project only discovers the contradiction when a decision depends on it.
What follows is how I approach each discipline, and what each one should actually deliver.
1.0 Planning & scheduling
Critical path built to be used, not filed. Logic that survives contact with the site, durations traceable to a productivity rate, and detail matched to the phase the project is actually in. Read more →
2.0 Cost engineering
Estimate through to forecast with the trail intact — every movement between the two attributed, and the estimate class stated wherever the number appears. Read more →
3.0 Risk analysis
Quantified rather than assumed, so contingency has a number behind it and a ranked list of what is driving it. Read more →
4.0 Delivery assurance
Reporting built around trend rather than position, with thresholds agreed before anyone had an interest in where they sat. Read more →
How engagements typically start
Most begin one of three ways: a baseline that needs building before sanction, a programme that has lost confidence in its own forecast, or an independent review ahead of a funding decision. All three start the same way — with an honest assessment of what the current data can and cannot support.