M5.3 — Integrated Cost & Schedule Performance

M5.3 — Integrated Cost & Schedule Performance

1. Practical outcome

The learner should be able to connect authorised scope, time-phased budget, evidence of accomplishment and incurred cost at a common cut-off. They should calculate and interpret basic Earned Value Management measures, reconcile them to source records, distinguish cost efficiency from expenditure timing, recognise the limitations of budget-based schedule indicators, challenge aggregation and use performance evidence to test remaining-work forecasts.

EVM is a cross-cutting method within the Open Standard. M1 provides scope, baseline, coding and progress rules; Module 2 provides the credible schedule; Module 3 provides budget, actual cost and accruals; Module 4 provides uncertainty; M5 integrates interpretation and decision support. This chapter teaches the underlying professional logic. It does not certify that a project satisfies a contractual earned value management system (EVMS) standard, and software-generated indices do not establish compliance.

Ask: “How much authorised work should have been accomplished, how much has actually been accomplished, what did it cost, and what does that imply for the work still to do?”

2. Core concepts

The three views must describe the same work

Planned Value (PV) is the budgeted value of authorised work scheduled to be accomplished by the cut-off. Earned Value (EV) is the budgeted value of work actually accomplished by that cut-off, measured using the approved progress method. Actual Cost (AC) is the incurred cost associated with the work performed by the cut-off, including valid accruals under the project’s accounting basis.

These are comparable values expressed in cost units, but they answer different questions. EV is not revenue, invoice value, market value or customer benefit. AC is not necessarily cash paid. PV should not be derived from a payment schedule when payment and accomplishment differ.

State whose cost is being measured. In a contractor EVMS, the contractor’s fee/profit is outside its cost-based performance baseline; applicable direct and indirect costs belong on a consistent basis. An owner’s acquisition cost may include the price paid to that contractor, including its fee. Do not mix contractor cost-based EV with owner price-based AC, or strip fee from an owner’s cost view without a corresponding, justified reconciliation of scope and basis.

The integrated baseline provides the denominator

Budget at Completion (BAC) is the total budget for the scope being measured. The Performance Measurement Baseline (PMB) is the controlled time-phased budget used to measure performance. Its scope, schedule, responsibilities, budgets and progress rules should be compatible.

The 50.0m CU project authorisation is not automatically the BAC for an EVM calculation. Under formal EVM, the PMB includes budgets for authorised work, including planning packages and undistributed budget, but excludes management reserve. Undistributed budget is authorised scope awaiting distribution, not risk reserve. Management reserve cannot fund added contract scope or erase accumulated overruns. Reconcile the measured-scope BAC to the relevant PMB and project total; reserve earns no value merely with time.

Common EVM terminology also uses BCWS for PV, BCWP for EV and ACWP for AC. Check that equivalent labels cover the same scope, cost basis and reporting period before combining datasets.

Earned value depends on evidence and the chosen method

Units complete, weighted milestones, fixed-formula techniques and other approved methods convert accomplishment into budgeted value. Select them before results are known and apply them consistently. Their design is taught in M1.4; M5.3 uses the resulting accomplishment to analyse performance.

A procurement package can earn value through approved manufacturing or delivery gates without pretending that payment proves accomplishment. A 0/100 method may suit short, objectively complete tasks but produces lumpy period results on long tasks. A method that earns too much early can give reassuring indices while difficult acceptance remains.

Level of effort requires separate visibility

Genuine level-of-effort (LOE) work is normally earned in relation to the planned passage of time under the defined method. Its schedule index can therefore stay near one while discrete deliverables deteriorate. It should not be used to describe work that has measurable outputs merely because LOE is easier to administer.

Identify LOE, discrete and other work types in the dataset. Where material, show discrete-work performance separately so support activity does not conceal delivery failure. Budget-weighted progress remains one legitimate view; it is not a substitute for mandatory readiness gates.

Cost and schedule variances use budget values

Cost Variance (CV) = EV − AC. Positive is favourable and negative is adverse under this convention. It asks whether accomplished work cost more or less than its budgeted value.

Schedule Variance (SV) = EV − PV. Positive means more budgeted work has been accomplished than planned by the cut-off; negative means less. SV is expressed in CU, not working days. It is not the difference between forecast and baseline completion dates, and it does not show whether the shortfall lies on a controlling path.

Indices normalise the comparison

Cost Performance Index (CPI) = EV/AC. A CPI of 0.80 means 0.80 CU of budgeted accomplishment has been earned per CU incurred. Schedule Performance Index (SPI) = EV/PV. An SPI of 0.80 means earned budget value is 80% of planned budget value at the cut-off. It does not mean the project is 20% late in calendar terms.

If AC or PV is zero, the corresponding ratio is undefined. Investigate the reason rather than substituting zero, one or an arbitrarily large value. Early-period and milestone-based results can be volatile. Always review raw values, method and maturity beside indices.

Period performance can reveal what cumulative indices conceal

Cumulative indices describe the full history to date. Period indices describe accomplishment and cost within the reporting interval. A three-period CPI can be calculated as the sum of EV earned in those periods divided by their summed AC; it is not the arithmetic mean of three CPI values.

Recent indices can reveal deterioration or improvement, but can also be distorted by milestone lumpiness, accrual corrections and work mix. Reconcile period movements to cumulative values and inspect the evidence before extrapolating recent performance.

Correct aggregation uses the underlying values

Project CPI = sum of EV/sum of AC; project SPI = sum of EV/sum of PV. Averaging package indices gives a small package the same influence as a large one unless a mathematically valid weighting is used. Prefer the underlying totals.

Aggregation can still hide important exceptions. A large package performing well can outweigh a small package that controls opening. Show material control accounts, critical milestones and readiness gates alongside overall indices. Portfolio aggregation also requires consistent currency and control definitions; a combined index across incompatible bases can be meaningless.

Schedule logic remains essential

Once all measured scope is complete and the baseline planned finish has been reached, EV = PV = BAC, so conventional SV is zero and SPI is one even if completion was late. If work finishes early, PV may still be below BAC until its planned date. These measures become weak indicators of lateness near completion; retain critical-path analysis, milestone date variance, remaining-duration evidence and readiness conditions.

Budget values also weight expensive work more heavily than low-cost interfaces. A critical approval can threaten completion while total SPI stays close to one. An optional time-based earned schedule method can provide additional analysis, but should be taught with its own definitions and limitations; it is not required to interpret the basic EVM measures here.

Performance-based EAC calculations are conditional tests

Estimate to Complete (ETC) is the expected cost of remaining work. Estimate at Completion (EAC) = AC + ETC. A bottom-up ETC should consider remaining quantities, resources, rates, productivity, commercial exposure and uncertainty as developed in M3.5.

Several simple EAC tests help challenge that forecast:

·       EAC = AC + (BAC − EV): remaining work costs its original budget; past variance is retained, but future efficiency is assumed to be one.

·       EAC = AC + (BAC − EV)/CPI: cumulative cost efficiency continues on remaining work. With the same cumulative data, this equals BAC/CPI.

·       EAC = AC + (BAC − EV)/(CPI × SPI): a composite sensitivity assumes schedule performance also affects remaining cost. Use only when the mechanism and scope make the assumption defensible.

These are conditional projections, not statistical confidence bounds or universal predictions. They do not automatically include all commercial changes, risks or reserve needs. M5.4 reconciles them with a delivery-based forecast and Module 4 uncertainty analysis.

TCPI tests the efficiency required by a goal

To-Complete Performance Index (TCPI) against a specified cost goal = (BAC − EV)/(goal − AC). With goal = BAC, it asks what remaining efficiency is required to finish within the original measured-scope budget. With goal = a proposed EAC, it tests the remaining efficiency implied by that forecast.

A required efficiency substantially above demonstrated performance needs an evidenced explanation: a changed method, different remaining work mix, secured rates or removal of a known constraint. The ratio does not make the goal achievable. If remaining funds are zero or negative while work remains, the ratio is undefined or not operationally meaningful; report the shortfall directly.

Baseline changes and data corrections require transparency

Approved change can alter scope, budget and future phasing under governance. An updated baseline should not silently erase past adverse performance. Preserve the original/current baseline relationship, approval, effective date and change bridge as required by M6.

Correct erroneous progress and cost records with traceable adjustments. A correction to EV can alter CPI and SPI without new physical work; an accrual correction can alter CPI without changed productivity. Reports should identify these effects rather than credit them to a delivery intervention.

3. Step-by-step method

1. Define the measured scope and decision. Identify the control accounts or packages covered, their responsible owners and what performance judgement is required.

2. Confirm integrated-baseline readiness. Reconcile WBS/CBS, scope, time-phased budget, schedule, progress rules and authorisation. Resolve material mismatches before presenting precise indices.

3. Reconcile BAC to project authority. State what is measured, how undetailed scope is handled and how contingency/reserve sits inside or outside the performance baseline.

4. Establish the common cut-off. Align schedule status, evidence, cost incurrence and accrual basis. Record source versions and permitted reporting lags.

5. Validate accomplishment. Apply approved methods to verified outputs; separate rejected work, partial acceptance, LOE and corrections. Preserve source evidence.

6. Extract PV and calculate EV. Use controlled phasing and budget weights for the same scope. Check that cumulative EV does not exceed authorised BAC without a documented reason requiring correction.

7. Establish AC. Reconcile ledger records, valid accruals, duplicate charges and coding. Do not add outstanding commitments to incurred AC as if both were costs to date.

8. Calculate period and cumulative CV, SV, CPI and SPI. State units and signs; handle zero denominators explicitly. Reconcile movements to preserved snapshots.

9. Aggregate underlying values. Retain meaningful breakdowns by discrete work, LOE, package, supplier and critical interface.

10. Investigate drivers. Use M5.2 reasoning and discipline-specific evidence; distinguish efficiency, accomplishment timing, method effects, correction and authorised change.

11. Test milestone and path implications. Compare SPI/SV with the actual schedule forecast, near-critical paths, acceptance gates and working-day milestone variance.

12. Test the remaining cost forecast. Compare bottom-up ETC/EAC with appropriate performance-based projections and TCPI. Challenge improvement assumptions and omitted exposure.

13. Integrate uncertainty and action. Carry coherent assumptions into M5.4, identify decisions and response owners, and preserve the basis for the next review.

14. Communicate limits. Explain what the indices demonstrate, what they cannot demonstrate and which evidence still qualifies the conclusion.

4. Worked Reference Project example — Community & Learning Centre

Context — a measured subset rather than the whole budget

Continue the 100-point Technology acceptance subset. For this EVM teaching calculation, 1.000m CU of the existing Technology base budget is assigned to a discrete acceptance work package containing 100 equal-budget units. Each unit earns 0.010m CU when its defined acceptance condition is met. Any earlier installation-stage credit belongs to separately budgeted work and is not also earned in this acceptance work package.

The example makes an illustrative allocation within existing authorised scope. It does not create a new 1.000m CU budget, replace the 3.8m CU Technology package authorisation or use the 50.0m CU project cap as this work package’s BAC. Its AC covers only this measured subset; the broader 1.58m CU incurred-cost example in M5.2 covers different package scope and is not inserted here.

Step A — establish controlled values

At Month 15, 40 units were planned and accepted; incurred cost was 0.420m CU. At Month 16, the plan requires 80 units, 64 are accepted and reconciled incurred cost is 0.800m CU. No approved baseline change or historical correction occurs between these snapshots.

PV at Month 16 = 80 × 0.010 = 0.800m CU. EV = 64 × 0.010 = 0.640m CU. CV = 0.640 − 0.800 = −0.160m CU. SV = 0.640 − 0.800 = −0.160m CU. CPI = 0.640/0.800 = 0.800. SPI = 0.640/0.800 = 0.800.

The equal CV and SV are a coincidence of these teaching values, not a general relationship. Cost efficiency is adverse and budgeted accomplishment is behind plan. The recent period is weaker than the cumulative result; its CPI is 0.240/0.380 = 0.632 and SPI is 0.240/0.400 = 0.600.

Step B — connect the arithmetic to delivery

M5.2 identified access restrictions, failed inspection, configuration approval and inspection capacity as acceptance drivers. The controls team checks where their incurred cost sits and whether the progress method recognises accomplishment consistently. It also asks whether some AC precedes acceptance because of the approved 0/100 method. That timing can contribute to short-term volatility without removing the real acceptance backlog.

M5.1 found that nine units per week are now required versus six observed. The scheduler tests the unfinished acceptance sequence and operational handover dependencies. SPI of 0.800 is not converted to a 20% time delay or used alone to move the public opening forecast.

Step C — test EAC assumptions

Remaining budgeted value = BAC − EV = 1.000 − 0.640 = 0.360m CU. The following tests use the same measured scope and Month 16 cut-off.

The bottom-up ETC of 0.340m CU implies remaining efficiency of 0.360/0.340 = 1.059, above the cumulative CPI of 0.800 and recent CPI of 0.632. Management must explain why remaining work can be delivered more efficiently. Secured access, a validated defect correction and cheaper remaining work may support that conclusion; a promise of recovery does not.

TCPI to finish within BAC = 0.360/(1.000 − 0.800) = 1.800. This would require 1.80 CU of budgeted accomplishment per CU of remaining spend. That is much stronger than demonstrated performance and should not be accepted without exceptional evidence. The team preserves the original BAC and reports the forecast challenge.

Step D — demonstrate correct aggregation

Consider a separate two-account illustration. Account A has EV 0.100m CU and AC 0.080m CU, so CPI = 1.25. Account B has EV 0.900m CU and AC 1.200m CU, so CPI = 0.75. The unweighted mean is 1.00, which would imply neutral performance. Correct combined CPI = (0.100 + 0.900)/(0.080 + 1.200) = 1.000/1.280 = 0.781.

These two accounts are a mathematical aggregation illustration and are not added to the acceptance subset or project budget. Even the correct combined result would need critical-path context if the smaller account controlled a mandatory opening gate.

Step E — management interpretation

“The Technology acceptance subset has earned 0.640m CU against 0.800m CU planned at an incurred cost of 0.800m CU. Cumulative CPI and SPI are both 0.800; recent-period performance is weaker. The bottom-up EAC of 1.140m CU requires improved remaining efficiency that has not yet been fully evidenced. Continuing cumulative efficiency would indicate 1.250m CU. The package owner must validate the access, quality and inspection assumptions; the scheduler must test the acceptance path and opening interface. The indices do not quantify a calendar delay or provide a risk confidence level.”

5. Interpretation and professional judgement

Reliable inputs matter more than precise indices

Three decimal places do not compensate for unsupported EV or missing accruals. If material data is unverified, qualify the result and resolve the source problem. EVM integrates the evidence; it does not improve poor evidence automatically.

The remaining work may differ from completed work

Cumulative CPI extrapolation is less persuasive when procurement is complete and the remaining work is testing, or when a known one-off mobilisation cost has ended. Explain the changed work mix and rebuild ETC. Conversely, difficult remaining integration work can make a seemingly conservative extrapolation optimistic.

CPI below one does not explain cause

It identifies adverse cost efficiency on the chosen measurement basis. Rate, productivity, rework, early cost incurrence, measurement lumpiness and coding effects need separate diagnosis. Controls should not assign blame solely from the ratio.

SPI near one can coexist with a late opening forecast

Expensive non-critical work can dominate the value-based result while a low-cost approval controls completion. Late completed scope also drives SPI back towards one. Trace the milestone path and readiness conditions before making a schedule statement.

Reserve authority does not repair performance

An approved budget transfer may change a future control basis under governance. It does not demonstrate that earlier work was efficient. Preserve the performance history and reserve movement separately, using the convention adopted by the organisation.

Performance-based EAC tests are challenge tools

The spread between 1.140 and 1.363m CU above is a spread between assumptions, not an 80% confidence interval. A probabilistic result requires a defined uncertainty model with credible dependencies and inputs as taught in M4.3.

6. Practice diagnostic

Use this table to challenge the quality of the approach or deliverable. The rows consolidate the most important recurring red flags and checks; they are prompts for professional judgement rather than a substitute for the method above.

7. Try-it-yourself practical exercise

Use the 1.000m CU acceptance work package and the Month 15/16 snapshots.

·       1. Recalculate cumulative and period CV, SV, CPI and SPI. State units, signs and the exact question each result answers.

·       2. An additional valid 0.060m CU accrual is discovered for work performed by Month 16. Recalculate cumulative CPI and CV, and explain why EV, PV and SPI do not change solely because of that accounting correction.

·       3. A separate 0.100m CU support work package has PV = EV = 0.100m CU and AC = 0.090m CU. Aggregate it with the original acceptance values and compare the combined SPI with discrete-work SPI. Explain what the aggregate conceals.

·       4. Reproduce the three performance-based EAC tests and TCPI to BAC. Compare the bottom-up ETC of 0.340m CU with the efficiency it requires.

·       5. Produce an evidence checklist that would support or reject the improved remaining-efficiency assumption. Include access, quality, supplier terms and remaining work mix.

·       6. The acceptance subset eventually completes three weeks after its required milestone. Calculate final PV/EV/SPI for that completed scope and explain why milestone variance must remain visible.

·       7. Write a 150-word management interpretation that includes an EAC challenge and a schedule limitation without claiming the formulas guarantee an outcome.

What good performance looks like

With the extra accrual, AC becomes 0.860m CU, CPI = 0.640/0.860 = 0.744 and CV = −0.220m CU. Using the original values for the separate support illustration, combined PV = 0.900, EV = 0.740 and AC = 0.890m CU: SPI = 0.822 and CPI = 0.831. Support improves the aggregate SPI relative to discrete-work SPI of 0.800 without removing the acceptance shortfall. On completion, conventional SPI becomes one even when the milestone was late. Strong work preserves each scope/correction boundary and explains the operational implications.

Final challenge: “What could make these indices mathematically correct but operationally misleading?”

8. Bridge to capability

The How-to teaches how scope, planned value, accomplishment and incurred cost are compared. At APPLIED and above, capability assessment should test whether the practitioner can reconstruct realistic inputs, interpret results and challenge misleading indices. Reconciled PV, EV and AC, sound progress rules and connected schedule and forecast evidence should support the analysis. Supporting work products are illustrative and should be assessed with the practitioner’s reasoning. Sample learning questions illustrate the judgement at each recognition level; they are discussion prompts rather than validated assessment items.

Learning question: Do you understand how planned value, accomplishment and cost should be compared?

Capability question for APPLIED and above: Can you demonstrate a reliable integrated comparison and explain what its indices do and do not support?