7 Project Management Formulas and Quantitative Techniques

Start with assumptions and units

Project management calculations turn estimates and performance data into inputs for decisions. Before calculating, identify the question’s assumptions, units, and measurement date. A formula supports judgment; it does not replace it.

Estimate work and uncertainty

Choosing an estimating approach

Analogous estimating uses results from a similar past project or activity, adjusted for differences. It is relatively quick, but depends on how comparable the work is.

Parametric estimating applies a statistical rate to a quantity. For example, at $3$ labor-hours per unit, $240$ units require an estimated $720$ labor-hours before adjustments for setup, rework, or other factors.

Bottom-up estimating estimates detailed activities or work packages and adds them to obtain a higher-level estimate. Detail can improve the estimate when scope is sufficiently defined, but bottom-up estimating takes more effort.

Expert judgment uses informed experience, ideally alongside historical data and other estimating methods.

Three-point estimates

Three-point estimating uses an optimistic estimate $O$, a most likely estimate $M$, and a pessimistic estimate $P$. Check that all three use the same units and refer to the same scope.

The triangular expected estimate is $E = \frac{O + M + P}{3}$. The PERT expected estimate, which weights the most likely estimate more heavily, is $E = \frac{O + 4M + P}{6}$. For PERT, the standard deviation is $\sigma = \frac{P - O}{6}$, and the variance is $\sigma^2 = \left(\frac{P - O}{6}\right)^2$.

For $O = 4$ days, $M = 7$ days, and $P = 16$ days, the PERT estimate is $\frac{4 + 4(7) + 16}{6} = 8$ days. The standard deviation is $\frac{16 - 4}{6} = 2$ days.

For a path of independent activities, add the expected durations to estimate path duration and add activity variances to estimate path variance. Take the square root of the total variance to obtain path standard deviation. This is an approximation: activity dependencies and the project’s actual duration distributions can affect the result.

Estimates, budgets, and reserves

An estimate predicts the resources or cost needed to perform defined work; a budget is the authorized amount. In common predictive-project terminology, the cost baseline includes contingency reserve for identified risks. Management reserve addresses work within project scope that is not yet specifically planned and is not part of the cost baseline. Follow the definitions given in the problem because terminology and reserve treatment can vary by context.

Measure earned value performance

Earned value management (EVM) compares three values at a status date. Use the same scope, currency, and reporting period for each value.

• Planned Value (PV): Budgeted value of work scheduled by the status date.

• Earned Value (EV): Budgeted value of work actually completed by the status date.

• Actual Cost (AC): Cost actually incurred for that completed work.

• Budget at Completion (BAC): Total authorized budget for the baseline scope.

Variances and performance indices

Cost variance is $\mathrm{CV} = \mathrm{EV} - \mathrm{AC}$. A negative result indicates that the work performed is over budget; a positive result indicates that it is under budget.

Schedule variance is $\mathrm{SV} = \mathrm{EV} - \mathrm{PV}$. A negative result indicates that less work has been completed than planned; a positive result indicates that more work has been completed than planned.

The Cost performance index (CPI) is $\mathrm{CPI} = \frac{\mathrm{EV}}{\mathrm{AC}}$. A value below $1$ indicates cost inefficiency relative to plan, while a value above $1$ indicates cost efficiency relative to plan.

The schedule performance index is $\mathrm{SPI} = \frac{\mathrm{EV}}{\mathrm{PV}}$. A value below $1$ indicates progress behind the planned rate; a value above $1$ indicates progress ahead of the planned rate.

For $\mathrm{PV} = \$50{,}000$, $\mathrm{EV} = \$40{,}000$, and $\mathrm{AC} = \$45{,}000$, the results are $\mathrm{CV} = -\$5{,}000$, $\mathrm{SV} = -\$10{,}000$, $\mathrm{CPI} \approx 0.89$, and $\mathrm{SPI} = 0.80$. The project has earned about $89$ cents of planned value per dollar spent and has completed $80\%$ of the work planned by this date. EVM schedule variance is expressed in value units, not calendar time.

Forecast cost and required performance

Forecasting cost

Estimate to Complete (ETC) is the forecast cost of the remaining work. Estimate at Completion (EAC) is the forecast total cost at completion. Variance at Completion (VAC) is $\mathrm{VAC} = \mathrm{BAC} - \mathrm{EAC}$; a negative result forecasts an overrun against $\mathrm{BAC}$.

Choose an EAC method according to the cause of past variance and the assumptions in the question:

1. If past variance is considered atypical and remaining work is expected to follow plan, use $\mathrm{EAC} = \mathrm{AC} + (\mathrm{BAC} - \mathrm{EV})$.

2. If current cost performance is expected to continue, use $\mathrm{EAC} = \mathrm{AC} + \frac{\mathrm{BAC} - \mathrm{EV}}{\mathrm{CPI}}$. This is algebraically equivalent to $\frac{\mathrm{BAC}}{\mathrm{CPI}}$ when $\mathrm{CPI}$ is cumulative $\frac{\mathrm{EV}}{\mathrm{AC}}$ and the same values and scope are used.

3. If both cost and schedule performance are assumed to affect remaining cost, use $\mathrm{EAC} = \mathrm{AC} + \frac{\mathrm{BAC} - \mathrm{EV}}{\mathrm{CPI} \times \mathrm{SPI}}$.

The third method is an assumption-based forecast, not a universal rule: schedule delays do not always translate directly into cost inefficiency. For any selected EAC, calculate $\mathrm{ETC} = \mathrm{EAC} - \mathrm{AC}$.

Required performance on remaining work

To-Complete Performance Index (TCPI) measures the cost efficiency required on remaining work to meet a target. To meet $\mathrm{BAC}$, use $\mathrm{TCPI} = \frac{\mathrm{BAC} - \mathrm{EV}}{\mathrm{BAC} - \mathrm{AC}}$. To meet a revised $\mathrm{EAC}$, use $\mathrm{TCPI} = \frac{\mathrm{BAC} - \mathrm{EV}}{\mathrm{EAC} - \mathrm{AC}}$.

A higher required $\mathrm{TCPI}$ indicates that better cost performance is needed on the remaining work. If a denominator is zero or negative, inspect the scenario rather than applying the formula mechanically.

Calculate schedule dates and float

The Critical path is the longest-duration path through a logically linked schedule network. Under the modeled logic and durations, it determines the earliest possible project completion. A schedule can have more than one critical path.

Forward and backward passes

For a basic activity-on-node network using a consistent time convention, calculate early and late dates as follows:

• Early Finish (EF): $\mathrm{EF} = \mathrm{ES} + \text{duration}$.

• Early Start (ES): The largest $\mathrm{EF}$ among immediate predecessors; for a starting activity, use the schedule’s defined start time.

• Late Start (LS): $\mathrm{LS} = \mathrm{LF} - \text{duration}$.

• Late Finish (LF): The smallest $\mathrm{LS}$ among immediate successors; for a terminal activity, use the project completion date or modeled finish.

The forward pass calculates early dates from the start; the backward pass calculates late dates from the finish. State whether the schedule counts elapsed time from time zero or uses inclusive calendar-day dates, because the convention can change day-count arithmetic.

Total float is $\mathrm{LS} - \mathrm{ES}$ or $\mathrm{LF} - \mathrm{EF}$. It is the amount an activity can slip without delaying the modeled project completion date. Free float is $\text{earliest ES of immediate successors} - \mathrm{EF}$; it is the amount an activity can slip without delaying an immediate successor’s early start.

Activities with zero total float are critical under the calculated schedule. Constraints, calendars, resource limits, and changes in logic can alter the result. Float is not automatically a promise of usable contingency.

Evaluate risk and decisions

Expected monetary value

For a chance event, multiply each possible monetary outcome by its probability and add the results: $\mathrm{EMV} = \sum (\text{probability of outcome} \times \text{monetary outcome})$. Probabilities for mutually exclusive and exhaustive outcomes should total $1$. For a threat, expected monetary exposure is often calculated using positive cost impacts; for an opportunity, use positive benefit impacts. EMV is an average over possible outcomes, not a guarantee of what will happen.

A response that costs $\$2{,}000$ and avoids a $25\%$ chance of a $\$10{,}000$ loss has an expected avoided loss of $\$10{,}000 \times 0.25 = \$2{,}500$. Its simplified expected net benefit is $\$2{,}500 - \$2{,}000 = \$500$. Whether to act also depends on constraints, risk appetite, and other consequences.

Decision tools

A decision tree lays out choices and chance outcomes. Calculate expected values from terminal outcomes backward through chance nodes, then compare alternatives using the decision objective, such as highest expected benefit or lowest expected cost.

Sensitivity analysis varies an important assumption to see how the decision changes. Monte Carlo simulation samples uncertain inputs repeatedly to produce a range or probability distribution of possible outcomes; it does not make weak input data reliable.

Apply other quantitative relationships

The number of potential pairwise communication channels among $n$ people who could communicate directly with one another is $\frac{n(n - 1)}{2}$. With $6$ people, there are $\frac{6 \times 5}{2} = 15$ potential channels. This counts possible links, not how many messages will be sent.

A productivity or unit rate is $\frac{\text{output}}{\text{input}}$, such as units completed per labor-hour. Confirm that the rates being compared use consistent definitions and conditions.

Percentage complete by budgeted value is $\frac{\mathrm{EV}}{\mathrm{BAC}} \times 100\%$. This is an earned-value measure, not necessarily a count of physical units unless the earning rules support that interpretation.

Use a reliable calculation process

Use a consistent calculation process:

1. Identify what the question asks for: an estimate, variance, index, forecast, or required performance.

2. Write down the relevant values and units; distinguish planned, earned, actual, and budgeted amounts.

3. Select the formula whose assumptions match the scenario.

4. Substitute values carefully and retain precision until the final rounding step.

5. Interpret the sign, ratio, units, and decision implication. A correct number without the correct interpretation is not a complete answer.

Across these techniques, clear scope, consistent data, and stated assumptions matter. Estimating methods and three-point formulas represent work and uncertainty; EVM measures performance and supports forecasts; network passes and float describe schedule flexibility and critical work; and risk and decision techniques provide additional quantitative perspectives.

• Expected monetary value (EMV)