M3.4 — Cost Control & Variance Analysis

M3.4 — Cost Control & Variance Analysis

1. Practical outcome

By the end of this section, the learner should be able to use the approved cost baseline and the reliable current cost position established in M3.3 to determine where cost performance differs from plan, explain why it differs, judge whether the movement is temporary or structural, assess the likely effect on final cost, and convert the analysis into proportionate management action. They should be able to distinguish timing effects from genuine cost-performance problems; analyse period and cumulative movements; decompose material variances into quantity, rate, productivity, mix, rework, price and other drivers; challenge favourable as well as adverse results; and maintain the distinction between performance variance, authorised change, contingency and forecast.

The objective is not to produce a red/green variance table. Cost control is a closed-loop management process: establish a trustworthy current position, compare it with the correct control reference, identify the real drivers, decide what can still be influenced, take action, and test in the next reporting cycle whether the action worked. A variance that is calculated but not understood or acted upon is only accounting history.

M3.4 therefore sits deliberately between M3.3 and M3.5. M3.3 establishes what cost has actually been incurred and committed at the cut-off. M3.4 explains the resulting variances and controls the response. M3.5 develops the forward-looking cost forecast and commercial/contract interfaces in greater depth.

A useful question throughout this Knowledge Area is: “What is different from the authorised cost plan, why is it different, is the cause likely to continue, and what decision or action should follow?”

2. Core concepts

Cost control is more than cost reporting

Cost reporting describes the current numbers. Cost control uses those numbers to influence the outcome. The control cycle links baseline, current cost, variance analysis, forecast, corrective action and governance. The practitioner should be able to explain not only that a package is 0.20m CU adverse, but what caused the movement, which part is recoverable, who owns the response, what the response may cost, and when evidence will show whether it is working.

This means cost control is inseparable from delivery reality. A cost-saving action that delays the fixed opening date, degrades quality or transfers unmanaged risk elsewhere may be a poor project decision even if the local cost line improves. Cost analysis should inform integrated decisions, not encourage isolated optimisation.

The comparison must use the correct reference

A variance has meaning only when both sides of the comparison refer to the same scope, cut-off, coding structure, currency/price basis and control status. Comparing current incurred cost with an obsolete budget, a different WBS definition or a different reporting date can create an apparently precise but invalid result.

The current authorised cost baseline is normally the primary control reference. If approved change has legitimately altered the baseline, the practitioner should preserve the original and prior positions but analyse current performance against the current authorised baseline. Unapproved change, anticipated claims and management targets should not be silently inserted into the baseline merely to improve the variance.

Budget, incurred cost, commitment and forecast answer different questions

The baseline answers: what cost was authorised and when was it planned to be incurred or earned? Incurred cost answers: what cost belongs to work performed by the cut-off? Commitment answers: what contractual or ordered exposure remains? Forecast answers: what do we now expect the final cost to be? These views are connected but they are not interchangeable.

M3.3 is critical here. If 0.25m CU of valid accruals is missing, a package can appear favourable simply because invoices are late. Variance analysis should therefore start from the reconciled incurred-cost position, not from whatever cash or invoice value happens to be visible in the ledger on the day the report is produced.

Period and cumulative variance tell different stories

A current-period variance shows what changed this reporting cycle and is useful for identifying emerging behaviour. A cumulative variance shows the net position to date. A large adverse movement this month can be hidden inside a previously favourable cumulative balance, while a persistent small monthly variance can become more important than a single large one-off event.

Good control therefore examines both. The practitioner should also compare the current period with recent periods so that deterioration, improvement and recurring patterns are visible rather than treated as isolated monthly explanations.

Spend against the time-phased baseline is useful, but it is not automatically cost performance

A simple comparison of incurred cost with the time-phased cost baseline can reveal whether expenditure is ahead of or behind the planned profile. It cannot by itself explain efficiency because the amount of work accomplished may also differ from plan. Spending less than planned can be favourable, or it can simply mean that less work has been delivered.

For example, if the project planned to incur 1.62m CU by Month 12 but has incurred 1.50m CU, the 0.12m CU apparent favourable movement is not enough to conclude that cost performance is good. If the completed scope should only have cost 1.30m CU on the approved basis, the work actually delivered may still be materially inefficient.

Compare incurred cost with the budgeted value of accomplished scope where the control system supports it

Where the project has reliable scope, progress and time-phased budget integration, the practitioner can compare the incurred cost of completed work with the budget value assigned to that accomplished scope. This removes much of the timing distortion caused by being ahead or behind plan and gives a stronger view of underlying cost efficiency.

M3.4 uses this concept to teach cost-control judgement, but it does not duplicate the formal integrated cost/schedule performance methods in M5.3. M5.3 develops Earned Value concepts, indices and integrated interpretation. The M3.4 principle is simpler: do not call lower spend a saving until you have considered how much authorised scope has actually been delivered for that spend.

Variance is a signal; root cause is the analysis

A headline variance describes the size and direction of a difference. It does not explain the cause. “Labour is 0.10m CU adverse” is still only a symptom if the real issue is low productivity caused by restricted access, excessive rework, an incorrect crew mix or a supplier learning curve.

The analysis should continue until the driver is specific enough to support a decision. The depth should be proportionate: a 0.002m CU one-off stationery variance does not need the same investigation as a recurring 0.20m CU productivity loss on a near-critical installation package.

Common cost drivers should be separated rather than blended

Useful cost-driver categories include scope/quantity, unit rate or price, labour productivity, resource mix, overtime or premium working, rework, waste, logistics, mobilisation, escalation, currency, procurement strategy, subcontractor performance, change, schedule delay and accounting/coding effects. The categories should reflect how the organisation can act on the result rather than being chosen only because they are easy to extract from a ledger.

A material variance may contain several drivers at once. For example, labour cost can be adverse because more hours were used than planned, the average rate was higher, and some of the hours were spent on rework. Combining all three into “labour overspend” hides the management response.

Quantity, rate and productivity analysis turns a total variance into controllable components

For unit-based work, a simple first-principles baseline is budget quantity multiplied by budget unit rate. Actual cost then reflects actual quantity and actual effective rate. The practitioner can separate the effect of delivering more or fewer units from the effect of paying a different rate. Where labour or plant effort is important, productivity adds another dimension: how much output was achieved per hour, crew-day or machine-hour compared with the basis used in the estimate and baseline.

Illustrative logic: baseline cost = authorised quantity × baseline unit cost. Cost movement can then be explored through quantity change, unit-rate change, productivity/resource-use change and identifiable exceptional items such as rework. The exact decomposition method should be defined consistently so the parts reconcile to the total variance.

The purpose is not to force every variance into a mathematically perfect driver tree. It is to create a defensible explanation. If two drivers interact, the organisation should use a consistent convention for allocating the interaction rather than presenting different answers depending on who performs the analysis.

Timing variance should not be confused with permanent cost variance

Cost can move between periods without changing final expected cost. An invoice may post early, a mobilisation milestone may occur before the original cash profile, or work may be resequenced. These movements may create a period variance but have no adverse effect on Estimate at Completion. The practitioner should identify timing differences explicitly so they do not trigger unnecessary corrective action.

The opposite is also true: a favourable period result can hide a future problem. Delayed procurement or slow mobilisation may reduce current spend while increasing future schedule and cost exposure. A robust narrative explains whether the movement is timing, efficiency, scope, price, change or forecast deterioration.

Authorised change and performance variance must remain separate

When scope is legitimately changed and the authorised baseline is adjusted through governance, the new baseline becomes the current control reference from its approved effective date. Before approval, the expected cost of the change may belong in the forecast or change register, but it should not be used to rewrite history or erase an adverse performance variance.

This distinction protects accountability. A productivity overrun should not be relabelled as change simply because additional budget is requested. Equally, a genuine client-directed scope addition should not be treated as poor cost performance by the delivery team when the control basis has not yet caught up with the authorised decision.

Contingency and reserve do not make adverse performance disappear

M3.2 established that contingency and reserve should be visible and governed. A cost variance caused by inefficiency, rework or unsupported rate growth should remain visible even if management later approves funding to deal with the consequence. Reserve drawdown changes the authorised funding/control position under defined rules; it does not prove that the original work was delivered efficiently.

The practitioner should therefore report both the performance story and the governance story: what variance arose, whether a qualifying uncertainty materialised, whether contingency or reserve is eligible, what authority approved any drawdown, and what residual forecast exposure remains.

Cost-to-date variance and forecast-at-completion variance are not the same

A package can be adverse to date but recover part of the position through changed methods, negotiated rates or learning-curve improvement. It can also look acceptable to date while the remaining work is forecast to overrun badly. The most decision-useful analysis therefore connects historical variance with the expected cost of remaining work.

M3.4 should at least test whether the identified driver is likely to continue and what it means for final cost. M3.5 then develops ETC/EAC forecasting, commercial exposure and forecast assurance in greater depth.

Commitment analysis can reveal variance before actual cost arrives

A new purchase order, supplier quotation, subcontract award or contract modification can show that future cost has moved before the work is performed. Comparing commitments with the corresponding budget allowance can therefore provide an early warning. A package may still have low actual cost but already be commercially committed above budget.

The practitioner should avoid simplistic comparisons where scope or terms differ. A higher commitment may include additional authorised scope, different payment terms, transferred risk or quantities that were not in the original allowance. The variance needs a like-for-like commercial explanation.

Materiality and trend determine the depth of control response

Not every variance deserves escalation. Organisations should define thresholds that consider both absolute and percentage movement, but thresholds should not replace judgement. A small variance on a critical scarce resource, a recurring rate problem or a leading indicator of a much larger forecast movement may deserve attention even when it is below a formal reporting threshold.

Trend is often more informative than one point. Repeated adverse productivity movements, a declining forecast each month or a growing difference between committed cost and budget may indicate a structural problem before the final overrun becomes obvious.

Variance analysis should finish with action, ownership and a test of effectiveness

A complete cost-control response states what will be done, by whom, by when, what cost or schedule consequence the action itself may have, and what evidence will show whether it worked. “Monitor closely” is not a control action unless the monitoring has a defined trigger and response.

The next reporting cycle should test the action. If the same driver continues, management should know whether the original action was not implemented, was implemented but ineffective, or was overwhelmed by another factor. This closes the control loop and creates learning rather than repetitive commentary.

3. Step-by-step method — how to control cost and analyse variance

1. Confirm the reporting cut-off and control basis. Establish the period, data date, currency/price basis and organisational accounting rules. Confirm that schedule/progress and cost data are aligned or that any cut-off difference is explicitly reconciled.

2. Confirm the current authorised cost baseline. Identify the approved budget, time phasing, control-account ownership, approved changes and held contingency/reserve. Preserve the original baseline and prior authorised positions so the history of movement remains visible.

3. Preserve the previous reporting snapshot. Retain the prior period actual/incurred cost, commitments, ETC/EAC, variances, actions and explanations. Variance control is stronger when the practitioner can explain not only where the project is now but what changed since the last cycle.

4. Establish the reliable current incurred-cost position. Use the M3.3 process to reconcile ledger actuals, valid accruals, credits, reversals and coding corrections. Do not begin variance interpretation while material cost is still missing or duplicated.

5. Validate current commitments and major commercial movements. Reconcile open purchase orders/contracts, approved contract changes, cancelled items and significant new awards. Identify commitments that signal a future variance even if no actual cost has yet posted.

6. Align cost and budget to the same control structure. Map baseline, incurred cost, commitments and forecast to the same WBS/CBS/control account and accountable owner. Resolve material miscoding before interpreting performance.

7. Remove or explain known distortions. Identify cut-off differences, advance payments, retentions, tax treatment, foreign-exchange effects, one-off accounting journals, transfers and other items that could make a numerical movement look like delivery performance when it is not.

8. Calculate current-period and cumulative expenditure variances. Compare incurred cost with the time-phased baseline at the same cut-off. Record both amount and percentage where useful, but treat this first result as a diagnostic rather than the final performance conclusion.

9. Test whether the expenditure variance is caused by delivery timing. Compare planned and actual physical accomplishment, procurement timing and major milestones. Identify whether lower or higher spend simply reflects less/more work, resequencing or commercial timing.

10. Compare incurred cost with the budgeted value of accomplished scope where reliable integration exists. Use approved progress-measurement rules and baseline cost assignments to determine whether the work actually completed has cost more or less than its authorised cost basis. Keep this analysis traceable to the underlying scope and progress evidence.

11. Decompose material variances into controllable drivers. Analyse quantity, unit rate, productivity, resource mix, overtime/premium working, rework, waste, logistics, escalation/currency, supplier performance, change and other relevant causes. Reconcile the driver explanation back to the headline variance, allowing a clearly labelled residual where exact allocation would create false precision.

12. Analyse commitment and procurement variances. Compare committed values with the baseline allowance for equivalent scope. Identify whether variance is caused by market price, quantity, terms, scope, risk transfer, procurement strategy or pending change, and whether it is already reflected in the forecast.

13. Classify each material movement correctly. Distinguish timing, performance, authorised change, unapproved/pending change, risk/uncertainty, accounting/coding correction and forecast judgement. The classification should determine the appropriate governance route rather than being chosen to improve reported performance.

14. Review contingency and reserve eligibility without netting the variance away. Test whether a materialised event meets the approved drawdown rules. Show the performance variance before any drawdown, the authorised transfer separately, and the residual exposure after the decision.

15. Assess the effect on ETC and EAC. Determine whether the cause is complete, temporary, recurring or expected to worsen. Update the forward cost view sufficiently to show the likely final consequence, while recognising that M3.5 provides the fuller cost-forecasting method.

16. Compare against prior forecast and recent trend. Identify new variance, deterioration/improvement, repeated drivers and forecast movement. Check whether earlier corrective actions produced the expected effect.

17. Define proportionate corrective or protective actions. Assign an owner, due date, expected cost/schedule/risk consequence and success measure. Consider alternatives rather than assuming the cheapest local response is best for the project.

18. Approve, communicate and preserve the analysis. Reconcile headline figures, record assumptions and evidence, obtain the required review, publish a concise management narrative and retain the period snapshot so future cycles can distinguish correction, change, trend and genuine performance.

4. Worked Reference Project example — Community & Learning Centre

Use a separate Month-12 variance scenario for the WBS 1.4 Technology package. It retains the 3.800m CU authorisation, comprising 3.711m CU base control budget and 0.089m CU package contingency, but uses a different incurred-cost dataset from the 1.834m CU reconciliation case in M3.3. Apply the M3.3 method to the 1.790m CU position below. M3.5 continues this variance scenario.

Step A — establish the correct Month-12 cost position

The finance ledger shows 1.540m CU of recognised actual cost. M3.3 review identifies 0.250m CU of valid incurred-but-not-yet-invoiced cost that must be accrued. The correct incurred-cost position is therefore 1.790m CU. This single adjustment demonstrates why cost control cannot rely on invoice timing.

Recognised actuals 1.540m CU + valid accruals 0.250m CU = incurred cost 1.790m CU.

The Month-12 time-phased base cost baseline planned 1.620m CU of cost by the cut-off. If the project compared the ledger actual alone with the baseline, it would report 0.080m CU favourable (1.540m versus 1.620m). Once the valid accrual is included, the expenditure position becomes 0.170m CU adverse (1.790m versus 1.620m). The apparent favourable result was a billing-timing illusion.

Step B — separate timing/progress from underlying cost efficiency

Progress evidence shows that the budget value assigned to the scope physically accomplished by Month 12 is 1.480m CU. In other words, the project planned to have accomplished 1.620m CU of budgeted work by this date but has accomplished only 1.480m CU. This is principally a delivery-timing/progress issue and should not be confused with cost efficiency.

Planned budgeted accomplishment to Month 12: 1.620m CU. Budget value of scope actually accomplished: 1.480m CU. The project is 0.140m CU behind the planned accomplishment profile on this cost basis.

Now compare incurred cost with the budget basis of the work actually accomplished. The project has incurred 1.790m CU to deliver scope that carried 1.480m CU of baseline cost. The underlying cost difference on accomplished scope is therefore 0.310m CU adverse. This is a much more serious message than the original ledger-only 0.080m CU favourable result.

Incurred cost 1.790m CU − budget value of accomplished scope 1.480m CU = 0.310m CU adverse underlying cost movement.

M5.3 will formalise integrated earned-value measures and indices. At M3.4, the professional lesson is that cost should be interpreted against what has actually been delivered, not only against the calendar date.

Step C — decompose the 0.310m CU adverse movement

The cost engineer works with the Technology Lead, planner, commercial manager and package team to trace the 0.310m CU movement to evidence. The agreed analysis is:

·       0.120m CU adverse — labour productivity below the estimating/baseline assumption during early installation and configuration.

·       0.065m CU adverse — rework following failed inspection and room-interface changes within the existing authorised scope.

·       0.045m CU adverse — premium logistics and expedited deliveries used to protect access dates.

·       0.035m CU adverse — temporary higher-cost resource mix and overtime during mobilisation/learning.

·       0.025m CU adverse — supplier unit-rate and quantity effects on consumable components.

·       0.020m CU adverse — smaller residual items, each evidenced but below the level at which separate management action is useful.

The driver values reconcile to 0.310m CU. More importantly, the categories suggest different responses. Rework requires quality/interface control; productivity requires work-method and learning-curve intervention; premium logistics may need procurement planning; resource mix may reduce as the team stabilises. Calling the entire 0.310m CU “supplier overspend” would produce the wrong action.

Step D — test whether the causes will continue

The team reviews the remaining work rather than assuming the adverse cost-to-date will repeat mechanically. Early productivity was poor, but the last two reporting periods show improving installation output. Rework controls have been strengthened and the room-interface issue is now resolved. Premium logistics is expected to reduce after the next delivery lot. However, some supplier rate pressure remains.

The remaining unfinished scope carries 2.231m CU of base budget (3.711m less the 1.480m budget value of accomplished scope). The current evidence-based ETC is 2.130m CU. This implies that management expects part of the historic adverse position to be offset by better performance on the remaining work, but not all of it.

Current EAC = incurred cost 1.790m CU + ETC 2.130m CU = 3.920m CU.

Against the 3.711m CU base control budget, the current EAC is 0.209m CU adverse. Against the total 3.800m CU authorised package budget, it is 0.120m CU adverse. These are different but both useful comparisons. The first shows performance against the planned-work control basis; the second shows the expected funding/authorisation exposure for the package.

Step E — do not use held contingency to hide the performance story

The package still holds 0.089m CU contingency. The principal adverse drivers identified above are productivity loss, rework and premium execution choices, not automatic qualifying events under the package-contingency rules. The cost report therefore does not simply subtract 0.089m CU from the 0.120m CU forecast overrun and declare the package almost on budget.

If a separately identified uncertainty later materialises and meets the approved drawdown criteria, the project can process that decision through the M3.2 governance route. The variance history remains visible, and the budget movement is recorded separately from the operational cause.

Step F — compare with the prior forecast and identify the trend

At Month 11 the package EAC was 3.760m CU. The Month-12 EAC of 3.920m CU is therefore a 0.160m CU adverse forecast movement in one reporting cycle. Management should not focus only on the 0.120m CU current overrun against authorisation; the rapid month-on-month deterioration is itself a control signal.

The team checks which part of the forecast movement is genuinely new. 0.065m CU comes from rework recognised this period, 0.045m CU from premium logistics, and the remainder principally from revised productivity/resource assumptions. This allows the forecast movement to be tied to the same root-cause story as the cost-to-date variance.

Step G — define management actions rather than commentary

The control response is agreed as a short set of testable actions:

·       Technology Lead: introduce a two-stage room-readiness and inspection hold point before installation to prevent repeat rework; effectiveness measured by first-time-pass rate and rework hours.

·       Installation Manager: track points installed per crew-day weekly and demonstrate a sustained productivity recovery against the revised achievable target before further overtime is approved.

·       Commercial Manager: renegotiate the next two logistics lots and remove premium freight where it no longer protects a critical interface.

·       Resource Manager: reduce the temporary premium resource mix as trained internal capability becomes available, while checking that the change does not endanger the fixed opening date.

·       Cost Engineer: maintain a driver reconciliation from the 0.310m CU adverse cost-to-date position through ETC/EAC and report whether each action is reducing the relevant driver.

None of these actions changes the baseline automatically. If management later approves a genuine scope change, contingency drawdown or budget transfer, that decision will be shown separately so the project can still see whether the original corrective actions improved delivery performance.

Step H — produce a decision-useful management statement

“At Month 12, WBS 1.4 has incurred 1.790m CU against 1.620m CU time-phased baseline, a 0.170m CU adverse expenditure variance. Because the budget value of accomplished scope is only 1.480m CU, the underlying cost of delivered work is 0.310m CU adverse, driven mainly by early productivity loss (0.120m), rework (0.065m) and premium logistics (0.045m). Current EAC is 3.920m CU, 0.120m above the 3.800m authorised package total and 0.160m worse than last month. The 0.089m package contingency has not been assumed as an offset because the principal drivers are performance-related. Actions are in place on rework prevention, productivity, logistics and resource mix; the next review will test whether those actions reduce the identified drivers without jeopardising the fixed opening date.”

This statement is useful because it separates data quality, timing, accomplished scope, root cause, forecast consequence, contingency treatment and action. A statement that simply says “Technology is 4.5% over budget; recovery plan in progress” would conceal nearly all of the control information management needs.

5. Interpretation and professional judgement

Ask five questions in sequence

For every material movement, the practitioner should ask: Is the variance real? Is it timing or performance? What caused it? Will the cause continue? What decision or action follows? Skipping one of these questions commonly produces poor analysis — especially when a numerical result is accepted before data quality or delivery progress has been checked.

A favourable variance can be bad news

Lower spend is not automatically a saving. It may indicate delayed mobilisation, unplaced orders, missing accruals, lower delivered scope, deferred maintenance or failure to perform planned work. Favourable variances should be challenged with the same seriousness as adverse ones when they are material or unexplained.

Do not focus only on sunk cost

Cost already incurred cannot usually be recovered. Management attention should distinguish the historical explanation from the controllable future. The most valuable response may be to prevent recurrence on the remaining 60% of work rather than spending excessive effort proving the final decimal of a past variance.

This does not mean history is unimportant. Reliable root cause is needed to avoid repeating the same problem and to improve future estimating, planning and commercial decisions. The balance is to understand the past sufficiently to improve the future.

A recovery action can legitimately cost more

The cheapest action is not always the best project decision. Additional resource, expediting or overtime may increase package cost but protect a fixed opening date whose delay would create greater project cost or strategic harm. The practitioner should make the trade-off explicit rather than label all incremental recovery spend as poor control.

Conversely, expensive recovery should not be approved merely because a schedule date is threatened. The expected benefit, feasibility, resource constraints, risk and knock-on effects should be examined with planning, risk, commercial and project leadership.

Do not “variance-launder” through budget movement

Repeatedly moving budget to the area that overspent makes performance appear better but destroys the learning value of the baseline. Budget changes should follow authorised scope/change/contingency governance. Performance overruns remain visible as performance unless the underlying control basis was genuinely changed by an approved decision.

Distinguish one-off effects from recurring drivers

A one-off mobilisation premium may be complete and have no further forecast effect. Persistent productivity loss or supplier rate escalation is different. The same cost-to-date variance can therefore imply very different EAC outcomes. Forecast judgement should be based on whether the causal mechanism remains active, not on simple extrapolation.

Use thresholds as a filter, not a substitute for judgement

Formal thresholds help focus scarce management attention and create consistency. They should not suppress important emerging signals. A variance below threshold may still deserve escalation if it affects critical scope, repeats every period, threatens contingency adequacy or indicates that a key estimate assumption is invalid.

Keep cost and schedule stories connected

Cost movement often originates in schedule or delivery conditions: late access creates standby, acceleration creates premium shifts, resequencing changes resource efficiency, and delay extends indirect costs. The cost engineer should work with the planner and package team to determine whether the cost variance is a symptom of another control problem rather than treating it as a purely financial event.

Separate evidence, judgement and decision

A strong analysis distinguishes observed facts from interpretation. “Actual crew-hours were 18% above the baseline basis” is evidence. “The loss was caused by restricted access” is a judgement that requires supporting evidence. “Add a second supervisor and change the work-front release process” is a management decision. Keeping those layers visible makes challenge easier and improves auditability.

Residual variance is acceptable when false precision would be worse

It is not always possible or useful to allocate every small amount to a perfect driver. A clearly labelled residual, with a defined materiality threshold and evidence that no major issue is being hidden, is better than arbitrary allocation. The sum of explained drivers plus residual should still reconcile to the headline movement.

Good cost commentary answers “so what?”

Professional commentary should state the movement, the principal cause, whether it is timing or underlying performance, the forecast consequence, the management action and any decision required. It should avoid repeating figures already visible in the table unless the narrative explains their significance.

A practical test is: “If the variance table were removed, would the commentary still tell management what changed, why it matters and what they need to do?”

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.

Evidence to retain

·       approved current cost baseline and change history;

·       period and cumulative incurred-cost extracts, including accrual reconciliation;

·       progress/accomplishment evidence used to interpret spend timing;

·       variance calculations and driver reconciliations;

·       productivity, quantity, rate, resource-mix and rework evidence for material drivers;

·       commitment/contract award comparisons to budget allowance;

·       contingency/reserve eligibility and drawdown decisions;

·       prior-period forecast and variance comparison;

·       corrective-action log with owners, dates and success measures;

·       approved management narrative and period-end control snapshot.

7. Try-it-yourself practical exercise

Using the Community & Learning Centre Reference Project, perform a Month-12 cost-control review for WBS 1.4 Digital & Technology Systems. The purpose is to demonstrate that you can move from a raw financial position to a decision-useful explanation and action plan.

1. Start with the 3.800m CU authorised package total, 3.711m CU base control budget and 0.089m CU separately held package contingency established in M3.2.

2. Use a Month-12 time-phased base cost of 1.620m CU. Create ledger actuals of 1.540m CU and valid accruals of 0.250m CU. Calculate the correct incurred-cost position and show how the result would be misstated if the accruals were omitted.

3. Assume the budget value of physically accomplished scope is 1.480m CU. Calculate and explain: the ledger-only apparent variance; the expenditure variance after accrual; the difference between planned and accomplished budget; and the underlying cost difference on accomplished scope.

4. Decompose the underlying adverse cost movement into at least five evidenced drivers. Include productivity, rework and one rate/price or resource-mix effect. Reconcile the drivers to the headline movement and state any residual explicitly.

5. For one labour-intensive work package, create baseline quantity, baseline hours, baseline rate, actual quantity, actual hours and actual rate. Explain separately the quantity, productivity and rate effects without double counting.

6. Introduce one favourable current-period spend variance caused by delayed procurement. Explain why it should not be treated as a saving and identify the likely schedule/forecast consequence.

7. Introduce one new supplier commitment that is 0.070m CU above the equivalent budget allowance. Determine whether the movement is price, quantity, scope, terms or pending change and state how it should affect the forecast.

8. Include one genuine approved scope change and one performance overrun. Show how they follow different governance and reporting routes.

9. Test whether any adverse movement qualifies for the 0.089m CU package contingency. Do not assume eligibility; document the rule and evidence used.

10. Set current ETC to 2.130m CU and calculate EAC. Compare EAC with both the 3.711m CU base control budget and the 3.800m CU authorised package total. Explain why both comparisons are useful.

11. Set the prior-month EAC to 3.760m CU. Calculate the forecast movement and identify which drivers are new this period, which are continuing and which have improved.

12. Define at least four corrective/protective actions. Each must have an owner, due date, success measure and an explanation of any schedule, risk, quality or commercial trade-off.

13. Prepare a variance-driver bridge or reconciliation showing how the headline adverse movement is explained and how the remaining-work actions affect the forecast without rewriting the baseline.

14. Write a one-page management narrative that states what changed, why, whether the cause will continue, the forecast consequence, contingency treatment, action and any decision required. Avoid generic phrases such as “monitor closely”.

What good performance looks like

The exercise is successful when the learner can demonstrate that the cost comparison is made on a consistent basis; distinguish invoice timing from incurred cost; distinguish lower spend from genuine efficiency; quantify meaningful root causes; reconcile variance drivers; connect cost-to-date movement to ETC/EAC; preserve the distinction between change, contingency and performance; and propose actions that can be tested in the next cycle. A variance table with unexplained percentages is not sufficient.

8. Bridge to capability

The How-to teaches how cost evidence becomes an explanation and control response. At APPLIED and above, capability assessment should test whether the practitioner can analyse realistic variance, distinguish timing, change and performance, and justify action. Compatible comparisons, reconciled driver analysis and forecast consequences should support the explanation. 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 cost variance should be explained?

Capability question for APPLIED and above: Can you demonstrate a defensible driver analysis and recommend action that addresses the cause without disguising performance?