Earned Value at the Halfway Gate
Metro Transit is migrating its fare system to a hosted platform. The approved budget is $4.8 million over 12 months. At the end of month 6, the control account shows planned value of $2.0 million, earned value of $1.6 million, and actual cost of $1.9 million. The critical-path data-conversion activity is 80% complete but three weeks late because source records contain duplicate identifiers. A quality audit found that 15% of converted records require rework. Forty percent of the remaining budget is planned for the final three months, and the vendor has offered two additional conversion engineers at $38,000 each per month. The sponsor wants a firm go-live date; the CFO wants an updated estimate at completion. A change request for $250,000 of additional data cleansing is pending, and the team believes the original conversion estimate was optimistic.
1. Calculate the cost and schedule variances, CPI, SPI, EAC, ETC, VAC, and TCPI. State which EAC method you would use and why.
Read the worked answer
CV = EV − AC = 1.6 − 1.9 = −$0.3M (over budget). SV = EV − PV = 1.6 − 2.0 = −$0.4M (behind schedule). CPI = 1.6/1.9 = 0.84; SPI = 1.6/2.0 = 0.80. If current cost performance continues, EAC = BAC/CPI = 4.8/0.84 = $5.70M; ETC = 5.70 − 1.9 = $3.80M; VAC = 4.8 − 5.70 = −$0.90M. TCPI to finish within BAC = (4.8 − 1.6)/(4.8 − 1.9) = 3.2/2.9 = 1.10, meaning the team must deliver 10% more value per dollar than planned-unlikely without a change. If the variance were judged one-time or atypical, EAC = AC + (BAC − EV) = 1.9 + 3.2 = $5.1M; but the data-quality rework and optimistic estimate suggest the CPI method is more defensible. Both forecasts exceed BAC, so the $250,000 cleansing request must be evaluated as a scope change, not hidden in reserve.
2. Recommend corrective actions for the remaining six months. Address crashing, fast-tracking, root cause, and the use of reserves.
Read the worked answer
The root cause is source-data quality, so crashing the conversion without fixing it would accelerate rework. Corrective actions: (1) contain the defect-add data profiling and cleansing before conversion, not just more testers; (2) fast-track only where dependencies allow, such as starting conversion of clean record sets while cleansing the rest; (3) crash selectively with the two vendor engineers at $76,000 per month only if they work on the critical path and the forecast benefit exceeds the cost; (4) re-sequence the final three months, because 40% of budget in that window leaves little slack; (5) submit the $250,000 cleansing change request with a schedule and cost impact. Do not use management reserve to fund a known scope change; reserve is for unidentified risks. If the go-live date is fixed by contract, present a re-baselined date or a phased cutover with reduced scope. Update the risk register with data quality as a realized risk and revise the estimate using actual productivity.
3. What should the project manager tell the sponsor and CFO, and what governance actions should follow?
Read the worked answer
Tell the sponsor and CFO one integrated message: current CPI 0.84 and SPI 0.80; EAC about $5.70M versus $4.8M BAC; VAC about −$0.90M; and the date is at risk because the critical path is three weeks late. Give a range, not a false point estimate: $5.1M to $5.7M depending on whether the variance is one-time, with the higher figure more likely. Explain that the $250,000 cleansing request is additional scope, so approving it raises the baseline to $5.05M before the performance variance; rejecting it means accepting rework and schedule risk. Ask for a decision at a gate: approve funding and re-baseline, or reduce scope or phase the rollout. Publish the revised forecast in the status report and update the cost baseline only after approval. Keep TCPI in the conversation-1.10 within the old BAC is not credible, so pretending the original budget is intact would mislead governance.