Earned Value Management has been the gold standard for integrated cost and schedule performance measurement for over five decades. And yet, embedded within the EVM methodology is a blind spot that most practitioners have learned to live with — even though it consistently produces misleading signals at the moment they are needed most.
The blind spot is the Schedule Performance Index expressed in cost terms (SPIcost).
The Problem With SPI(cost)
In conventional EVM, the Schedule Performance Index is calculated as:
SPI = EV ÷ PV (Earned Value ÷ Planned Value)
An SPI below 1.0 means the project is behind schedule. An SPI above 1.0 means it is ahead. This seems intuitive — and for much of the project life cycle it is a useful indicator.
But here is the problem: at the end of a project, SPI always converges to 1.0, regardless of how late the project finishes.
Why? Because at project completion, EV (total earned value) must equal BAC (Budget at Completion) — assuming all scope is delivered. And PV (planned value) also equals BAC. So SPI = BAC ÷ BAC = 1.0. A project that finishes six months late will show SPI = 1.0 at completion.
This means that SPI(cost) becomes increasingly unreliable as a schedule indicator in the later stages of a project — precisely when stakeholders are most focused on delivery date certainty. In the last 20% of project duration, SPIcost can actually improve as the project falls further behind schedule.
This is not a theoretical concern. It affects the accuracy of completion date forecasts for real projects, and it is a risk that has been well-documented in AACE International literature since the early 2000s.
Enter Earned Schedule
Earned Schedule (ES) is a method developed by Walt Lipke that addresses this limitation directly. Rather than expressing schedule performance in cost terms, Earned Schedule converts earned value into a time-based equivalent.
The core calculation:
ES = AT(cumulative) + fraction (where EV falls between two PV cumulative data points)
In plain language: Earned Schedule answers the question "at what point in the planned schedule did we actually earn what we have earned so far?" — and expresses the answer in time units (months, weeks, days), not currency.
This single shift unlocks a set of far more reliable performance indicators:
- SPI(t) — Schedule Performance Index based on time. Calculated as ES ÷ AT (Actual Time elapsed). Does not converge to 1.0 at completion; remains a valid indicator throughout the project life cycle.
- TSPI(t) — To-Complete Schedule Performance Index based on time. Measures the efficiency needed in the remaining time to meet the target completion date — analogous to TCPI in cost forecasting.
- IEAC(t) — Independent Estimate at Completion expressed in time. Probabilistic completion date forecast derived from current schedule performance.
Why This Matters in Practice
The practical implications of adopting Earned Schedule in a project controls framework are significant:
Better completion date forecasting. IEAC(t) derived from SPI(t) consistently outperforms forecasts derived from SPI(cost) in late-project scenarios, particularly on projects experiencing schedule slippage. Published research on large defence and infrastructure programmes supports this consistently.
More credible performance reporting. When a project sponsor or lender asks "when will this finish?", an answer derived from time-based schedule performance metrics is more defensible than one derived from cost-weighted performance that is mathematically converging to 1.0.
Better contractor performance assessment. When evaluating contractor schedule performance in an EPC environment, SPI(t) provides a cleaner signal than SPI(cost), which can be distorted by cost-loading patterns on the contractor's schedule.
Integration with schedule risk analysis. Time-based performance metrics integrate naturally with Monte Carlo schedule risk modelling, enabling confidence-range completion forecasts that reflect both baseline uncertainty and actual performance to date.
Implementing Earned Schedule
Earned Schedule is not a replacement for EVM — it is an extension of it. Any project with a robust EVM framework already has the data required to calculate ES metrics. The additional computational overhead is minimal.
The barriers to adoption are primarily awareness and methodology. Many project controls teams are unfamiliar with ES, and many project management information systems do not calculate it natively — though it can be computed externally from standard EVM data extracts.
Faolan Consulting routinely incorporates Earned Schedule into our project controls frameworks, particularly on capital projects in mining, energy, and infrastructure where completion date certainty carries significant commercial and regulatory weight.
If your controls framework is using SPI(cost) as your primary schedule performance indicator in the latter half of your project, you are flying with an unreliable instrument. Let's fix that.
Controls. Intelligence. Delivered.