Modern Project Controls: Earned Value Management in Practice

Ask a project manager how the job is going and you will often hear "we've spent about 60 percent of the budget and we're roughly 60 percent done." It sounds reassuring. It is also almost meaningless, because it conflates two very different questions: how much have we spent, and how much have we actually accomplished. Earned Value Management (EVM) exists to separate those questions and, in doing so, to give you an honest early-warning system that most cost reports simply cannot.

The three numbers everything rests on

EVM sounds intimidating because of its acronyms, but at its heart it tracks three quantities over time, all expressed in the same currency.

• Planned Value (PV): the budgeted cost of the work you scheduled to have done by now.
• Earned Value (EV): the budgeted cost of the work you have actually completed — physical progress priced at the plan's rates.
• Actual Cost (AC): what you have genuinely spent to achieve that progress.

The magic is in the comparisons. Compare EV against PV and you learn whether you are ahead of or behind schedule. Compare EV against AC and you learn whether the work you completed cost more or less than budgeted. Two variances, two independent truths, from three numbers.

Reading the variances and indices

From those three values flow the metrics that steer decisions. The Schedule Variance (SV = EV minus PV) and Cost Variance (CV = EV minus AC) tell you, in money, how far off you are. Negative is bad in both cases. More useful for trend-watching are the ratios: the Schedule Performance Index (SPI = EV divided by PV) and the Cost Performance Index (CPI = EV divided by AC). An index of 1.0 means on plan; 0.9 means you are getting ninety paise of value for every rupee or dollar of position.

What makes CPI powerful is its stability. On most projects, once roughly 15 to 20 percent of the work is done, the CPI rarely improves dramatically without deliberate intervention. If you are running at 0.85 a fifth of the way through, betting the rest of the job will be flawless enough to recover is usually wishful thinking. This is why EVM is an early-warning tool: it tells you the trajectory long before the final invoices confirm it.

Forecasting the finish honestly

The most valuable output of EVM is not where you are but where you are heading. The Estimate at Completion (EAC) projects the final cost. The simplest and often most reliable form assumes current efficiency continues: divide the total budget (BAC) by the CPI. If your budget is 100 crore and your CPI is 0.90, your project is trending toward roughly 111 crore — a figure worth surfacing to stakeholders months before it becomes an unpleasant surprise.

You can also compute a To-Complete Performance Index (TCPI), which tells you the efficiency the remaining work must achieve to still hit the original budget. When the TCPI climbs well above your demonstrated CPI, that is the data telling you the target is no longer realistic and the conversation should shift from "can we recover" to "how do we replan."

Where EVM goes wrong on site

EVM fails not because the math is hard but because the inputs are gamed or sloppy. A few disciplines keep it honest.

• Measure physical progress objectively. The single biggest error is inflating EV by claiming progress that is not real. Use rules of credit — for example, percentage by installed quantity — not a foreman's gut feel.
• Align AC with EV timing. If costs are booked when invoiced rather than when work is done, your CV becomes noise. Accruals matter.
• Break work down sensibly. A control account should be small enough to measure but not so granular that tracking becomes a full-time job. Match the work breakdown structure to how the project is actually delivered.
• Avoid the 50/50 trap for long tasks. Crediting half the value at start and half at finish distorts trends on activities that span many reporting periods.

Scaling it to the project

Full EVM is overkill for a small fit-out and essential on a large infrastructure program. The right response is proportionality. On a modest job, tracking EV at the level of major work packages monthly is plenty. On a billion-rupee program with financing covenants and multiple stakeholders, weekly control-account-level EVM integrated with the schedule earns its keep many times over. In India's public infrastructure space, where funding tranches and audit scrutiny are realities, a defensible earned-value trail is as much a governance asset as a management one.

Takeaways

Earned value is not bureaucracy for its own sake — it is the difference between measuring effort and measuring achievement. Track the three base numbers rigorously, watch CPI and SPI as trends rather than snapshots, and use EAC to tell stakeholders the truth early. The teams that benefit most are not the ones with the fanciest dashboards; they are the ones whose progress claims are honest and whose costs are booked in step with the work.

Done well, EVM turns the vague comfort of "we're about 60 percent done" into a specific, actionable statement: here is what we have earned, here is what it cost, and here is where we will land if nothing changes. That clarity, delivered while there is still time to act, is the whole point of project controls.