The Three Numbers Every Project Manager Must Know
- Adebowale Oyinleye
- 11 minutes ago
- 7 min read
TWO NUMBERS FEEL LIKE CONTROL. THEY ARE NOT THE SAME MEASUREMENT. |
WHAT MOST SITES TRACK
Percentage complete, cash certified, calendar days used None of these, alone, says whether the job is in trouble
WHAT THE THREE NUMBERS TRACK
Planned Value, Earned Value and Actual Cost — same units, same date, compared
Together they say how much trouble, and where
WHAT HAPPENED: ILLUSTRATIVE COMPOSITE
A contractor reports a residential block at 60 per cent complete at month nine of a fifteen-month programme. Roughly 60 per cent of the ₦300 million budget has been certified and paid. Both numbers look reasonable side by side, and the monthly site meeting moves on.
Neither number was measured against the other. The 60 per cent complete was a site walk-round estimate, rounded to a figure nobody challenged. The 60 per cent paid included advance payment on materials not yet built in. The work actually in place, valued at its planned rate, came to nearly 45 per cent.
By month twelve the gap is impossible to round away. The project is both behind schedule and over budget, and there are three months left to find money and time that no longer exist.
“Percentage complete is an estimate. Earned value is a measurement — and the two are not the same number wearing different clothes.”
What the three numbers actually measure
Planned Value. The budgeted cost of the work that was scheduled to be done by today — not what has happened, what the baseline said should have happened. PMI’s own guidance defines it as the approved budget for work scheduled to be performed by a given date; its older name is Budgeted Cost of Work Scheduled.
Earned Value. The budgeted cost of the work actually completed by today, valued at the rate the baseline planned for it — not money spent, work done. A team that has spent very little but built a great deal has high earned value; a team that has spent freely on rework and mobilisation has low earned value however much cash has left the account.
Actual Cost. What was genuinely spent to produce that work — labour, plant, materials and site overheads actually incurred, matched to the same activities and the same date as the other two.
Why three, not two. Percentage complete on its own is a schedule opinion with no cost content. Cash spent on its own is a cost fact with no schedule content. Earned value is what lets the other two be compared on the same footing — against the budget, and against each other — instead of read separately and reconciled by instinct.
Cost Performance Index — value for every Naira spent
The formula. Cost Variance is Earned Value less Actual Cost; Cost Performance Index is Earned Value divided by Actual Cost. A CPI of 1.0 means every Naira spent bought exactly a Naira of planned work. Above 1.0 is favourable — more value delivered than spent. PMI’s guidance is explicit on the direction: a CPI below one signals an unfavourable cost condition.
Worked through the composite. At month nine the planned value for the programme to that date was ₦180 million (60 per cent of the ₦300 million budget). The work actually in place — earned value — came to ₦135 million (the 45 per cent figure). Actual cost to date was ₦189 million. CPI is 135 ÷ 189, or 0.71: every Naira spent has returned about seventy-one Kobo of planned work, not the Naira it was supposed to.
What that means in practice. A CPI of 0.71 does not describe a project running slightly hot. Using the standard Estimate at Completion formula — Budget at Completion divided by cumulative CPI — ₦300 million ÷ 0.71 is roughly ₦423 million: left uncorrected, this trajectory implies a final cost around forty per cent above the approved budget, an arithmetic a percentage-complete report never surfaces, because it never compares spend to the value of what was actually built.
Schedule Performance Index — on pace, in money terms
The formula. Schedule Variance is Earned Value less Planned Value; Schedule Performance Index is Earned Value divided by Planned Value. It measures schedule progress in the same currency as cost, which a bar chart cannot do — an SPI of 1.0 or above is favourable, below 1.0 unfavourable.
Worked through the composite. Earned value of ₦135 million against planned value of ₦180 million gives an SPI of 0.75: the project has delivered about three-quarters of the work it should have banked by this date, expressed in the same Naira terms as the cost figures above — not “a bit behind”, but a specific, calculable shortfall.
Reading the two together. CPI and SPI answer different questions, and a project can fail either one without failing the other. High CPI with low SPI is a project spending efficiently but too slowly — cost-disciplined, calendar-late. Low CPI with high SPI is a project moving fast by spending its way through problems — on programme, hemorrhaging money. Low on both, as in the composite above, is the condition a percentage-complete report is least equipped to catch early, because it only ever reports the schedule half of the picture.
“A project can be on schedule and losing money or under budget and dangerously late. Percentage complete cannot tell the two apart. CPI and SPI can — because they are the same measurement, taken twice, against two different baselines.”
The case against earned value management
The obvious objection. Earned value looks precise because it produces a number to two decimal places. Practitioners who have run it have a serious qualification to add.
The argument. The whole system stands on one input: how earned value itself gets measured. Where it is assessed by a site team’s own percentage-complete estimate, EVM specialists have documented the same failure repeatedly — progress is optimistically estimated through the middle of a work package, the indices look acceptable for months, and the shortfall only surfaces close to completion, by which point the cost and schedule variance it reveals can no longer be corrected.
The industry’s own fix concedes the point. The standard mitigations — capping claimed progress at fifty per cent until a task is fully complete, or at ninety per cent until it is signed off — exist specifically because raw percentage-complete self-reporting could not be trusted on its own. A technique that needs a cap on its own optimism has not solved subjectivity; it has bounded it.
Where that leaves it. Earned value management is not a substitute for competent progress verification, it is a multiplier on it. Feed it a self-reported percentage from a party with an interest in the answer, and it will report false comfort with impressive precision. Feed it progress measured against physical, milestone, or quantity-based criteria by someone without that interest, and it becomes the earliest reliable warning most projects will get.
Setting it up on a Nigerian site
First. Build the baseline before work starts, not once it is already under way. A cost-loaded programme — the full budget spread across activities and time — is the planned value curve, and there is no earned value to compare against a baseline that does not exist yet.
Second. Agree how earned value will be measured, and by whom, before the first valuation. Physical measurement or milestone achievement against the bill of quantities, assessed by the supervising consultant, holds up; the contractor’s own percentage-complete estimate, taken at face value, is the exact failure mode the objection above describes.
Third. Capture actual cost against the same activity codes as planned and earned value, not as a single running total. A lump total tells you what left the account; it cannot tell you which activity is bleeding money, which is the number a recovery plan actually needs.
Fourth. Calculate CPI and SPI at every valuation, not at year-end or at final account. The whole value of the two indices is early warning; run at completion, they are a post-mortem.
Fifth. Put the trend, not just the number, in front of the project team at every progress meeting. A CPI of 0.71 read once is a fact for the file; a CPI that has moved from 0.94 to 0.84 to 0.71 across three valuations is a decision waiting to be made, and it is the direction of travel – more than the reading itself – that tells a site whether to act now or to keep watching.
THE BROADER POINT
Percentage complete is an estimate. Earned value is a measurement.
Cost estimates on public infrastructure work are wrong far more often than they are right — nine projects in ten run over their approved budget, by twenty-eight per cent on average, on the research record. Most of what gets watched on a Nigerian site — percentage complete, cash certified — was never built to catch that coming; it can only confirm it once the money is already gone. Planned Value, Earned Value and Actual Cost, read together through CPI and SPI, can catch it while there is still programme left to recover in. Build the baseline before the first valuation, measure earned value against something other than the builder’s own optimism, and read both indices every time — because a project can be sixty per cent complete and sixty per cent paid and still be badly over budget and badly behind schedule, and only the three numbers, not the two already on the report, will tell you which. A recovery plan built on that distinction starts three months earlier than one built on a percentage-complete report alone — and three months, on a fifteen-month programme, is the difference between a manageable overrun and a project that never gets back on budget.
← Week 3 Why Inflation Ate Your Project· Projects Associates· www.projectsassociates.com ·Week 5 →Why Projects Go Wrong in Month Two
CLAUSE REFERENCES Project Management Institute: definitions and formulas for Planned Value, Earned Value, Actual Cost, Cost Variance, Schedule Variance, Cost Performance Index and Schedule Performance Index, per PMI’s published guidance (pmi.org). · Earned Value Management “Gold Card” reference, U.S. National Science Foundation / Defense Acquisition University (nsf.gov), corroborating the formulas and the Estimate at Completion variants (EAC = BAC ÷ cumulative CPI). · Flyvbjerg, B., Holm, M. S. and Buhl, S. (2002) “Underestimating Costs in Public Works Projects: Error or Lie?”, Journal of the American Planning Association, 68(3): 279–295 — 258 transportation infrastructure projects, cost overruns in 86% of cases, average 28% (rail 45%, fixed-link 34%, road 20%). · Percent-complete subjectivity and the 50/50 and 90/10 measurement conventions: Humphreys & Associates EVM commentary, corroborated against general project-controls practice. · The site scenario (₦300 million programme) is an illustrative composite, not a real project. |
Comments