Ordo scores every programme against two recognised standards — the global DCMA 14-point assessment and the UK's CIOB Planning Protocol 2021 — plus fourteen further integrity checks. You don't need to be a planner to understand any of them. Here's what each one looks for, why it matters, and the threshold a programme is held to.
The result is a single A–F grade, a CIOB PP21 compliance verdict, and every finding traced back to the activities behind it.
The industry-standard health check, developed by the US Defense Contract Management Agency to tell, quickly and objectively, whether a schedule is built well enough to be trusted before anyone relies on its dates.
Whether activities are properly joined up, so the programme reacts when something moves.
Every activity should have something before it and something after it. An activity with no predecessor or successor is "dangling" — if it slips, nothing downstream moves, so the programme quietly lies about the end date.
≤ 5% of activitiesA lead is a negative lag — telling a task to start before its predecessor finishes. It's almost always a modelling trick to force a date, and it distorts the critical path. There should be none.
0A lag is waiting time built into a link (e.g. "wait 14 days"). A few are legitimate — concrete curing, deliveries — but heavy use hides real work that should be a tracked activity.
≤ 5% of linksMost links should be simple Finish-to-Start: finish one job, start the next. Too many Start-to-Start or Finish-to-Finish links make a programme hard to follow and often mask sequencing problems.
≥ 90% Finish-to-StartWhether dates fall out of the plan naturally, or are being held in place artificially.
A hard constraint pins an activity to a fixed date regardless of the logic. Used heavily, it hides slippage — the programme can't show a delay because the date is nailed down.
≤ 5% of activitiesFloat is spare time before an activity affects the end date. A lot of activities with very high float (over ~44 days) usually means missing logic — work that isn't properly joined into the network.
≤ 5% over 44 daysNegative float means an activity is already behind where the logic says it needs to be — the programme is telling you it can't hit a date. Any negative float needs explaining and resolving.
0Whether the building blocks of the programme are realistic and clean.
Very long activities (over ~44 working days) are hard to track and hide risk — you can't tell if a 60-day job is on course until it's too late. They should be broken into smaller pieces.
≤ 5% over 44 daysForecast dates in the past, or actual dates in the future, relative to the data date (the "as-of" date of the update). These are data errors that make the whole update untrustworthy.
0If the programme is meant to be resource-loaded, every activity with a duration should carry the labour, plant or cost that drives it. Missing resources mean the plan isn't costed or levelled.
0 missing (if resourced)Whether progress and the critical path stack up against the baseline.
Activities that have slipped past their baseline dates. A high missed count is the clearest early sign a programme is falling behind its original plan.
≤ 5% of activitiesA sanity check on the network: push a big delay into a critical activity and the finish date should move by the same amount. If it doesn't, the critical path is broken and can't be trusted.
must passA ratio of how much time is needed versus how much is left. Above 1.0 is healthy; below means the remaining critical work has to run faster than planned just to hit the date.
≥ 0.95The rate work is actually being completed versus the baseline plan. Below 0.95 means tasks are finishing slower than planned — a leading indicator of overall delay.
≥ 0.95The 14-point pass is necessary but not sufficient. Ordo runs fourteen further integrity checks for the faults that quietly break a network even when the headline checks pass — including several that Asta Powerproject and other planning tools also flag.
Activities with no logic at all — floating free of the network.
No real predecessor, so the task can drift earlier than it should.
No real successor, so a slip never pushes anything downstream.
The rarest link type, almost always a mistake when it appears.
Links already implied by a longer path — clutter that hides the real drivers.
Work reported done ahead of its predecessors — the logic no longer matches reality.
Long waits buried in links instead of shown as trackable activities.
Many activities landing on one — risk concentrated on a single date.
Unmanaged waiting time sitting directly on the completion date.
A milestone should be a moment, not a task with length.
Scheduling work to the last moment removes all contingency.
Two activities sharing a name in one section — ambiguous in reports and claims.
Too few links per activity — the network is under-developed.
Non-hard date constraints that still pin dates logic should drive.
DCMA came out of US defence contracting. Closer to home, the Chartered Institute of Building publishes the Planning Protocol 2021 (CIOB PP21) — 15 pass/fail "Stress Tests" with tolerance thresholds. Ordo scores your programme against both, so you get the globally-recognised DCMA view and the UK professional-body view side by side.
PP21 is stricter in places — several tests are zero-tolerance (no hard constraints, no leads, no negative float). It's pass/fail by design: the protocol exists so a programme can be fairly rejected and reissued. For "Standard Projects" tests 4, 5 and 13 may be omitted; "Complex Projects" run all 15.
No open ends — every activity tied in.
0No link overlaps by a negative duration.
0No Finish-Start link carrying a lead.
0Start-Start / Finish-Finish links kept low.
< 10%Dates driven by logic, not pinned.
0No total float over ~2 reporting periods.
≤ 44dNothing already behind the logic.
0No activity over 44 working days.
0 over 44dProgress consistent with the data date.
0Programme reflects the full scope.
≤ 10%Milestones for the dates that matter.
presentWorking calendars applied correctly.
reflectiveEvery activity has a unique ID.
0 dupThe network recalculates cleanly.
recalculatesA critical path to each completion.
≥ 1Upload a P6, Asta or Excel file and get all 25 checks scored to a single grade — free.
Review a programme →Further reading: the DCMA 14-point assessment is documented publicly by Stantec and TenSix, among others. Thresholds shown are the widely-used defaults; they can be tuned per contract or client standard.