GM593 Unit 9 Forecast to Completion Example

Seventy Percent of Items, Fifty-Two Percent of the Work: A Unit 9 Forecast to Completion for a Lab Relocation

[Student Name]

Purdue University Global

GM593: Project Execution With Monitoring and Control

Unit 9 Forecast to Completion

[Instructor Name]

[Date]

The university, its relocation project and every figure are composite and illustrative; the methods cited are real.

Summary

The Hollis Hall relocation is forecast to finish in about 13.5 months against a 12-month plan, within a range of 13.0 to 14.0 months. The final cost is forecast at about $2,627,600 against a budget at completion of $2,360,000, within a range of $2,593,400 to $2,675,300. The recommended points assume the corrective actions hold sign-off capacity near its current level and that Hollis Hall stays open one month past its planned closing.

Forecasting the Finish: Three Methods

Counting items. At the end of month eight, 22,150 of 31,400 items have been scanned, leaving 9,250. Over the last four weeks, the project has scanned about 820 items a week. At that rate the remaining items would take about 11.3 weeks, finishing about six weeks before the planned end. This method treats every item as the same, which is the problem.

Counting equivalent units. Weighted by effort, the whole inventory comes to about 43,440 equivalent units, of which about 22,590 are done, or 52 percent. The remaining items carry about 20,850 units, an average of 2.25 per item, compared with about 1.0 for the items already moved. The project has demonstrated about 1,030 units a week since the corrective actions took effect. At that rate the remaining work would take about 20.2 weeks, finishing packing about three weeks after the planned end, with transport and receiving still to follow.

The labs moved first held mostly instruments; the labs still waiting hold most of the freezers and hazardous chemicals, so counting items makes the hardest work look like the smallest part.

Earned schedule. Walt Lipke's earned schedule method converts earned value into time, by finding the date on the baseline when the work now earned was planned to be complete (Lipke, 2003). At the end of month eight, the project's earned value matches the baseline for about month 7.0, giving a time-based schedule performance index, SPI(t), of about 0.874. Dividing the planned 12 months by 0.874 forecasts about 13.7 months. Unlike the cost-based SPI, which drifts toward 1.0 as a late project nears its end, SPI(t) keeps its meaning to completion.

The cost-based SPI tells a different and misleading story here. At 0.912 it suggests the project is less than 10 percent behind. But that index is a ratio of dollars, and at the end of any project earned value must equal planned value, so the cost-based SPI rises toward 1.0 even for a project that finishes months late. Lipke designed earned schedule to avoid this drift, which matters most in the last third of a project, where this relocation now sits.

The recommended finish of 13.5 months sits between the equivalent unit forecast, which covers packing only, and the earned schedule forecast for the whole project.

Forecasting the Final Cost: Four Estimates

At the end of month eight, earned value is $1,587,500, actual cost $1,744,500, the cost performance index 0.910 and the cost-based schedule performance index 0.912.

If past overruns were one-off. EAC equals actual cost plus the remaining budgeted work: $1,744,500 plus $772,500, or $2,517,000. This assumes future work will run exactly to budget, which the record does not support.

At the current CPI. EAC equals the budget at completion divided by the CPI: $2,360,000 divided by 0.910, or about $2,593,400.

With both indexes acting. EAC equals actual cost plus the remaining work divided by the product of CPI and SPI: $1,744,500 plus $772,500 divided by 0.830, or about $2,675,300. This assumes schedule pressure will raise costs as well, which fits a project now paying for extra signing help.

Bottom-up. Each control account's remaining work was re-estimated by its owner, and one month of holdover cost for Hollis Hall was added: utilities, security and the demolition contractor's standby charge, about $57,000. The total is about $2,627,600.

Frank Anbari notes that the index-based formulas are useful checks but rest on assumptions about future performance, and that a bottom-up estimate is worth the effort when conditions have changed (Anbari, 2003). The corrective actions changed the project's conditions in month six, so the bottom-up figure is the recommended point.

Checking the Forecasts Against Each Other

The three date methods and four cost estimates do not need to agree, but their disagreements should be explainable. The item count disagrees with the other two date methods because it ignores effort, and it is set aside for that reason. The equivalent unit forecast and earned schedule differ by about a month, because the first covers packing alone and the second the whole project, including transport and receiving. On cost, the one-off estimate is the outlier because it assumes a sudden return to budget that eight months of data contradict. The bottom-up figure falls between the CPI estimate and the combined-index estimate, which is what a reader would expect from a project that has fixed its main cause but is still paying for the fix.

Assumptions Behind the Recommended Points

The recommended finish assumes weekly sign-off stays at or above 1,030 equivalent units, that the contract safety specialist stays for the full ten weeks and that no lab requests a further delay. The recommended cost assumes one month of holdover, no second month, and recovery of none of the fan rework cost, which is still in dispute with the design firm. Each assumption has an owner who will report on it weekly.

What Would Move the Forecasts

Three conditions would move the forecasts most. If EHS sign-off capacity falls again, for example if the contract safety specialist leaves early, the finish would slip by about a week for every two weeks of reduced capacity. If the remaining ultra-low freezers fail qualification at the rate the first two did, biological labs would wait, and the finish could reach the top of the range. If the demolition contractor will not accept a one-month delay, the university would face either overtime to finish on time or a larger standby charge, and the cost forecast would rise.

Quentin Fleming and Joel Koppelman recommend reporting a forecast as a range with a stated basis, rather than a single figure, so that decision makers can see how much the estimate depends on its assumptions (Fleming & Koppelman, 2010). The steering group should decide now whether to negotiate the month of holdover with the demolition contractor, while the forecast still allows a choice.

References

Anbari, F. T. (2003). Earned value project management method and extensions. Project Management Journal, 34(4), 12-23. https://doi.org/10.1177/875697280303400403

Fleming, Q. W., & Koppelman, J. M. (2010). Earned value project management (4th ed.). Project Management Institute.

Lipke, W. (2003). Schedule is different. The Measurable News, 31-34.