Chapter 03
Programme confidence from evidence
How an owner converts the captured base into a defensible level of programme confidence — neither false reassurance nor unwarranted alarm — and how to recognise the divergences between report and record early enough to act.
01
Orient: confidence is not a colour
Programme confidence is the level of belief an owner can rationally hold about whether the project will hit its milestones. A traffic-light status is a summary of confidence, not a substitute for it. Confidence has a shape. Trade progress matches the baseline at the agreed level of detail. Critical-path activities show movement consistent with the schedule rather than movement that has been re-scheduled to look consistent. The site capture shows the work the schedule says should be there. Resource on site reconciles to the resource the recovery plan assumed when it was written. Materials called off are arriving, and the gate record matches the delivery schedule. Where any of these are absent, confidence drops, and the owner should be able to say by how much and against which milestone. The temptation in board rooms is to compress all of this into a colour. The colour is fine as a header. It is dangerous as the answer, because a colour cannot be challenged, only debated. A confidence position that names the components, the evidence behind each, and the dates by which the unresolved components will be resolved can be challenged, and a position that can be challenged is one that earns trust over time. The owner´s discipline is to insist on the components, not to settle for the colour. The colour follows from the components; the components do not follow from the colour, and any system in which the colour is set first and the components are written to fit it has stopped being a reporting system and started being a public-relations exercise.
02
Owner-frame: the spot-check
The owner does not need to look at every clip every week, and the owner who tries will lose the discipline within a quarter. They need to know that a spot-check would survive review. A spot-check at owner level looks like this. Pick the milestone closest to the next lender draw or the next stage gate. Ask for the captured-base view that demonstrates the work front: a 360 walk dated within the last week, a drone capture if the milestone is a structural one, a fixed-camera view if the milestone is about an enclosure or an installation. Compare it against the programme statement made in the standing pack. Read the safety summary for the same area, because a work front that is moving fast and reporting nothing is itself a divergence. Walk the variations register for that package and check the open notices. If those four artefacts tell the same story, the confidence is defensible. If they tell different stories, the divergence is the lesson, and the owner has caught it weeks earlier than they would have without the spot-check. The spot-check is not surveillance. It is calibration. Run it on different milestones each cycle, and the team comes to expect it as part of the rhythm rather than experience it as an audit. The spot-check that finds nothing is still useful, because it is the calibration that earns trust in the spot-check that one day finds something. An owner who only runs spot-checks when they are already worried teaches the team to read the spot-check as suspicion; an owner who runs them as routine teaches the team to read them as the way the project is read.
03
Workflow: stage-gate and design approvals
At stage gates and design approvals, programme confidence has to be specific because the cost of approving against intention rather than record only shows up later, when the next phase has assumed something that turns out not to be true. A stage-gate review uses the captured base to confirm the conditions assumed in the next phase: foundations are at the level the structural design called for, the substation enclosure is in the position the M&E coordination assumed, the cladding is set out where the facade approval was issued against, the riser locations are where the services drawing committed them. A design-stage approval that defers verification to closeout is an approval against intention, not against record, and it is the most common origin of the disputes that arrive eighteen months later. The captured base brings the verification forward, where it is cheap. A 360 walk and a comparison against the issued drawing is the work of an afternoon. The closeout dispute that the deferred verification produces is the work of a quarter. The discipline at the gate is to insist on evidence for each condition the next phase assumes, and to be honest about the conditions that cannot yet be evidenced rather than assume them through. The conditions that fail this test should not stop the gate by default; they should be named, dated, and tracked into the next stage, so the next stage at least knows what it is sitting on.
04
Governance: when to escalate
Concern should escalate when the captured base and the reported position diverge. A schedule that says the work front is on track and a captured view that shows it empty is a divergence. A safety summary that shows nothing of concern and a recurring near-miss pattern in the record is a divergence. A commercial register that shows no notices and a captured base that shows several disrupted weeks is a divergence. The owner who notices these early can ask for a recovery plan when there is still time to recover. The owner who waits until the divergence is undeniable inherits a project in trouble, often weeks or months after the trouble started. The captured base makes early noticing possible without surveillance and without the contractor feeling watched, because the noticing is done by reading the record the contractor produced rather than by adding scrutiny on top. The first conversation when a divergence is noticed should be a question, not an accusation. The contractor who was about to surface the same issue is given the room to do so. The contractor who was hoping to manage it without surfacing it is given the chance to bring a recovery plan rather than a defence. Both outcomes are better than the conversation that happens at the next milestone, when the divergence has compounded and the recovery options have narrowed.
05
What good looks like
Good is when programme confidence is reported in the same words by the project director, the owner, and the lender´s monitor. The confidence is calibrated: where the project says high, audits do not find surprises. Where the project says low, the recovery plan is named, the responsible parties are named, and the date by which the position will be revisited is named. The captured base is referenced in the appendix of the standing pack, not waved at as a reassurance. The owner´s board minutes do not need an apology a quarter later, because the position taken at the time was the position the record actually supported. Across a portfolio, the owner who runs this discipline tends to find that their projects in trouble are the ones that surfaced their trouble earliest, named it most honestly, and were either recovered while recovery was cheap or contained at predictable cost. The projects that fail expensively are the ones where confidence drifted upward in the reporting while the record was already saying something different, and where nobody called the divergence in time. That failure mode is preventable. The captured base is the prevention. The owner´s job is not to find the prevention; it is to insist that the system uses it, and to read the result against the report on a cadence that the project comes to rely on rather than fear.
Practice
01. Take last quarter's project review. Identify three claims, exceptions, or progress questions where time-aligned evidence would have shortened the conversation. For each, name the captured-base artefact that would have closed it.
Look for: Typical answers: a structural pour where dates were disputed (closed by a 360 walk and gate record), a cladding sequence where setting-out was queried (closed by drone capture against the issued drawing), a closeout claim about temporary works that lingered (closed by a fixed-camera view of the working area). The exercise teaches the reader to expect the artefact rather than ask for it under pressure, and to design the next capture plan so the artefact exists before the question arrives.
02. For your next stage gate, list the conditions the next phase assumes about what is in place. For each, write the one-line evidence ask that confirms the condition.
Look for: A workable list has six to ten conditions. Each evidence ask should be specific enough to be answered yes or no from the captured base, and short enough to fit in a stage-gate appendix. If the asks read as 'we believe' rather than 'the record shows', the gate is being approved against intention. The exercise tends to surface two or three conditions that nobody had thought to verify and that turn out to be where the next phase quietly assumes the most.
Checkpoint
For one critical milestone in the next quarter, can you describe the spot-check you would run today to confirm the reported position, and would it currently survive an external review?
Recommended reading