A programme can look finished and still be full of holes — logic that hides the real sequence, contingency spent before the job starts, risk that shifts without the end date moving. Every finding below is real, from programmes run through Ordo — client and project names removed. Three very different files, three very different stories.
A ~1,060-activity infrastructure tender programme. Looked complete. Wasn't.
Nearly a third of the network ran on Start-to-Start and Finish-to-Finish links instead of Finish-to-Start. Heavy non-FS logic makes a programme hard to follow, easy to game, and its critical path hard to trust — you can't see what really drives what. The review flags every non-FS link so it can be justified or corrected.
Over half the activities carried no resources or cost at all. A programme that isn't resourced can't be levelled, can't be cost-loaded, and can't support a credible delay or prolongation claim. The review counts exactly how much of the scope is unresourced.
A third of the programme was scheduled to start as late as possible. ALAP pushes work to the last responsible moment and removes the float that would absorb a delay — so the moment anything slips, it's critical. Standard DCMA checks miss this entirely; our extended checks surface it as its own flag.
A ~110-activity building programme that scored respectably — and still had holes worth catching.
This one passed most checks — but ten activities still had open ends, nine couldn't push their successors if they slipped, and three lags hid weeks of unmodelled time inside links rather than as trackable work. None of it was obvious on the bar chart. That's the point: a programme can look sound and still not hold up when you rely on it.
A 402-line Excel Gantt tracker — the kind most small firms actually use.
It listed every task, date and duration — and passed the duration and date checks. But with no links between activities, there's no critical path, so nothing can tell you what happens to the end date when one job slips. Ordo says so plainly, checks everything that is assessable, and shows exactly what a real linked programme would add.
Every programme is scored against both the DCMA 14-point assessment and the UK's CIOB Planning Protocol 2021 — 15 pass/fail stress tests. The P6 tender above failed six of them (open logic, leads, too many non-FS links, excessive float and over-long activities), returning a clear NON-COMPLIANT verdict a contract administrator can act on.
Between two accepted submissions five months apart, the completion date barely changed — it actually pulled four days earlier. But underneath: 316 activities added, 694 logic links re-wired, 26 critical-path durations compressed, 476 activities lost float, and the critical path spread onto 122 new activities. The end date held — but only because the programme was being propped up by compression and consumed float. A stable date can hide a rising risk; the gap analysis makes it visible.
Eighty-four milestones shifted between the two submissions. The gap analysis lists every key-date movement — was vs now vs delta — so a stakeholder can see exactly which commitments changed, rather than taking "on programme" on trust.
On a programme forecasting late, recovery modelling showed 60 working days was reachable three ways: crashing critical activities (a cost, but low risk), fast-tracking by overlapping 57 links (free, but real rework risk), or a combined approach. It names the exact activities to shorten and links to overlap — and it refuses to "crash" fixed work like concrete curing. Recovery is a choice between cost and risk; see both before you decide.
A 90-day fit-out activity with a single successor and no breakdown passes a duration check but isn't credible for its scope. The AI forensic-planner layer reads the flagged activities and says which are genuinely light on detail, which logic couldn't be built in that order, and which durations were compressed without a method behind them — with a recommendation for each.
Every review produces a branded, evidenced report — emailed to you and print-ready as a PDF. Two examples:
| Relationship Types (FS%) | 70.9% | FAIL |
| Resources | 51.1% | FAIL |
| High Float | 32.4% | FAIL |
| ALAP Density | 30.2% | FAIL |
| Unique Activity Names | 32.1% | FAIL |
| Project end | ▼ −4 days |
| Scope changed | +316 / −0 activities |
| Newly critical | 122 activities |
| Compression | 26 durations cut |
| Key dates moved | 84 milestones |
Quality review, drift, critical path, recovery — free. Drop in a .xer,
.xml or .xlsx and see what it finds.