Concept 7 of 15

Project cost management

4 questions test this

Cost management is three activities that get collapsed into one. Estimating asks what the work will cost, budgeting turns those estimates into a baseline spread across time, and control compares what has been spent against what that baseline said and does something about the difference.

An estimate is a range that gets reported as a number

Analogous estimating scales a completed project of similar shape. It is quick, available before anyone understands the work, and it inherits every peculiarity of the project it borrowed from. Parametric estimating multiplies a measured rate by a quantity, and it is exactly as good as the rate and the quantity are. Bottom up estimating costs each package of work and adds them up, which is the most accurate method available and impossible until the work has been decomposed.

Three point estimating asks for an optimistic, a most likely and a pessimistic figure and combines them. It is the only one of the four that returns a shape rather than a point, and that matters because the figure reaching the sponsor is one value pulled out of a distribution. Handing over the mean without the spread throws away what the sponsor needed in order to decide how much cover to hold.

The cost baseline and the funding requirement

Aggregating the estimates gives the cost of the work. Adding contingency reserve gives the cost baseline, which is the authorised, time phased spend the project is measured against. Adding management reserve on top gives the total project budget, and the funding requirement is that budget laid out against the dates on which the money actually has to be available.

Management reservefor what was never identifiedreleased by the sponsorContingency reservefor risks already identifieddrawn on by the project managerAggregated activity estimatesthe priced cost of the work itselfoutside the baselineCost baselinewhat variance iscalculated fromTotal projectbudgetthe authorisedfunding

Estimates aggregate into the cost of the work, and adding contingency reserve gives the cost baseline that every variance is calculated from. Management reserve sits above that line and inside the total budget, which is why releasing any of it moves the baseline rather than spending against it.

The distinction decides what counts as a variance. The baseline is what variance is calculated from, so everything inside it is spend the project promised to make, and money released from above the line arrives as a change to the promise.

Where the two reserves sit and who holds them

Contingency reserve covers identified risks. It lives inside the cost baseline because those risks were analysed and priced when the baseline was built, and the project manager draws on it without asking permission.

Management reserve covers what was never identified. It sits outside the cost baseline and inside the total budget, it belongs to the sponsor or the governing body rather than to the project manager, and releasing any of it is a change that moves the baseline. That is the practical test for telling the two apart.

Direct and indirect cost

A direct cost is incurred for this project and would not exist without it, such as the people assigned to it and the licences bought for it. An indirect cost is shared across several pieces of work and allocated to each, such as premises, shared services and management overhead.

Indirect cost matters to a project manager because it is allocated by a formula nobody on the project chose, so it moves for reasons unconnected to how the project is performing. An overrun caused by a change in the overhead recovery rate is a real overrun that arrived from the finance system rather than from delivery, and a project manager who cannot separate the two spends the review defending work that was never the problem.

A budget with no reserve is a forecast presented as a commitment

Every estimate carries uncertainty, so a total built out of estimates carries it too. Stripping the reserve out leaves the uncertainty exactly where it was and removes the funded response to it, and the first risk that materialises then gets paid for out of scope, out of quality or out of somebody else's project.

This is why a reserve deleted in a budget negotiation reappears later as a change request, an unpaid overtime bill or a quietly lowered acceptance standard. The honest version of that conversation states the point estimate, states the confidence attached to it, and names the amount being held against the risks already on the register. A sponsor is entitled to refuse the reserve. No sponsor is entitled to refuse the reserve and keep the confidence level that came with it.

Funding limit reconciliation

Money rarely arrives in one lump. It is released annually, quarterly or at agreed gates, and each period has a ceiling. Funding limit reconciliation compares planned spend against those ceilings period by period and reschedules work when the plan wants more money than the period allows.

It is a scheduling constraint that arrives from finance rather than from the work. If the plan spends 400,000 in a quarter that has 300,000 available, something moves, and moving it changes the schedule and usually the cost. Finding that out while the baseline is still being set makes it an ordinary planning decision, and finding it out when an invoice cannot be paid makes it an incident.

Common misconceptions

Contingency reserve and management reserve are two names for the same buffer.

Contingency sits inside the cost baseline and the project manager spends it against risks that were identified and priced. Management reserve sits outside the baseline and inside the total budget, belongs to the sponsor, and is released for work nobody foresaw. Spending contingency is expected behaviour. Releasing management reserve moves the baseline.

Spending less than planned means the project is under budget.

It means less money has left the account. A project four weeks behind has not bought the work it planned to buy, so a comfortable cost position is often a schedule problem that has been read as good news. Cost is only interpretable against the work actually completed.

4 questions test this concept

A project manager aggregates the work package estimates for a fleet electrification project and reaches 2,400,000. The risk analysis prices the identified risks at 300,000, and the governing body holds a further 200,000 against events nobody has foreseen. What is the cost baseline, and what is the total project budget?

  • AA baseline of 2,700,000 and a budget of 2,900,000.
  • BA baseline of 2,400,000 and a budget of 2,900,000.
  • CA baseline of 2,900,000 and a budget of 2,900,000.
  • DA baseline of 2,400,000 and a budget of 2,700,000.
Check whether it stuck.

One per page, with a worked explanation.

Start the set
Related material
Book
Software Estimation: Demystifying the Black Art, On producing an estimate that survives being questioned.
Book
How Big Things Get Done, On why large budgets fail and what reference class data does about it.
Book
Identifying and Managing Project Risk, On sizing a reserve and defending it to a sponsor who wants it removed.
Book
A Guide to the Project Management Body of Knowledge (PMBOK Guide), Seventh Edition, On the performance domains behind planning and measurement.