19.2 Communicating Cost Estimates and Cost/Schedule Key Performance Indicators
Key Takeaways
- An estimate must be communicated with its class and accuracy range stated before the point value, because a decision-maker anchors on the first number they hear.
- Exclusions that are documented but never surfaced to the decision-maker become unfunded scope, which is a governance failure rather than an estimating error.
- SPI is an earned-value measure in cost units, not a schedule measure: as a project approaches completion EV and PV both approach BAC, so SPI drifts toward 1.0 however late the project is.
- TCPI is a warning rather than a target; empirical program data shows cumulative CPI stabilises by roughly 15 to 20% completion and rarely improves by more than about 0.10 afterward.
- Confirming estimate adherence to contract terms covers scope alignment, pricing mechanism, escalation entitlement, currency risk, markup base, owner-furnished items, retention, and provisional sums — with markup applied to marked-up subcontractor totals the most frequent commercial finding.
19.2 Communicating Cost Estimates and Cost/Schedule Key Performance Indicators
Blueprint tasks 2.B communicate cost estimates to stakeholders, 2.G communicate cost and schedule key performance indices (KPIs), and 2.H confirm estimate adherence to contract scope and terms cover the outbound half of the interface domain. The technical work is done; this is about whether anyone can act on it.
1. Communicating an Estimate
An estimate communicated as a single number will be treated as a commitment, whatever caveats surround it. Communication therefore has to carry four things:
| Element | Why it must travel with the number |
|---|---|
| The class and accuracy range | Sets the reader's expectation of precision before they anchor on the point value |
| The basis, in one paragraph | What scope, what date, what currency, what method |
| The exclusions | Unstated exclusions become unfunded scope |
| The decision the estimate supports | A screening estimate misused as a control budget is a governance failure, not an estimating error |
The anchoring problem
Once a decision-maker hears a number, every subsequent discussion is anchored to it. This is why the order of presentation matters: state the class and range first, then the point value. "This is a Class 4 estimate, so the range is $95M to $155M, with $120M as the most likely" lands very differently from "$120 million, plus or minus."
[!TIP] Present the range as a picture. A simple horizontal range bar with the point value marked, and the previous class's range shown above it, communicates convergence far better than a table. The audience sees uncertainty narrowing as definition matures — which is the actual message of the classification system.
2. Cost and Schedule KPIs (Task 2.G)
A KPI set is a communication device, not an analysis. Its job is to make the project's condition legible at a glance and to prompt the right question.
| KPI | Formula | Reads as healthy when | Common misread |
|---|---|---|---|
| CPI | EV / AC | Greater than or equal to 1.0 | Early-period CPI inflated by missing accruals |
| SPI | EV / PV | Greater than or equal to 1.0 | Approaches 1.0 at completion even on a late project |
| TCPI (BAC) | (BAC − EV) / (BAC − AC) | Close to CPI | Values far above CPI are treated as targets rather than warnings |
| VAC | BAC − EAC | Positive | Depends entirely on which EAC method was used |
| Productivity factor | Budget hours / actual hours | Greater than or equal to 1.0 | Reported without earned quantity behind it |
| Commitment ratio | Committed / total forecast | Rising steadily | High ratio on a poorly priced scope is false comfort |
| Contingency drawdown | Drawn / original contingency | Below percent complete | Drawdown tracked without linking to risk retirement |
| Float consumption | Float lost / float available | Below percent complete | Recovered by re-logic rather than production |
| Change volume | Count and value per period | Flat or declining | Value tracked without frequency (see 18.3) |
Two rules the exam rewards
SPI is not a schedule measure. It is an earned-value measure expressed in cost units. As a project approaches completion, EV approaches BAC and PV approaches BAC, so SPI approaches 1.0 no matter how late the project is. Schedule status comes from the schedule — critical path and float — with SPI used only as a supporting indicator early in execution.
TCPI is a warning, not a target. When TCPI(BAC) is 1.18 and cumulative CPI has been 0.85 for six months, the honest reading is that the original BAC is unachievable. Presenting 1.18 as "the efficiency the team will now achieve" is the single most common misuse of the index. Empirical research on completed programs shows cumulative CPI stabilises by roughly 15–20% completion and rarely improves by more than about 0.10 thereafter.
3. Building a Report Set That Gets Read
- Three levels, one source. A one-page executive summary, a project-level report, and control-account detail — all generated from the same data with no manual re-keying between them.
- Trend beats snapshot. A single period's CPI is noise. Six periods of CPI is information.
- Exception reporting. Highlight control accounts breaching a defined threshold rather than presenting all of them equally.
- Fixed format, fixed cadence. Readers learn where to look. Changing the layout each month destroys that.
4. Confirming Estimate Adherence to Contract Scope and Terms (Task 2.H)
Task 2.H is the check that an estimate — or a contractor's price — actually matches what the contract says. It is where cost engineering meets commercial terms.
| Check | Question | Typical finding |
|---|---|---|
| Scope alignment | Does the priced scope match the contract scope of work exactly? | Priced exclusions that the contract does not permit |
| Pricing mechanism | Are rates consistent with the contract's schedule of rates or change-pricing formula? | Change priced at open-market rates when the contract fixes them |
| Escalation | Does the contract permit escalation, on what index, from what date? | Escalation claimed under a firm-fixed-price contract |
| Currency | Which currency, and who carries exchange risk? | Exchange exposure silently transferred |
| Markups | Are overhead and profit within the contracted percentages, and applied to the right base? | Markup applied to markup |
| Owner-furnished items | Are items the owner supplies excluded from the contractor's price? | Owner-furnished equipment priced by both parties |
| Retention and bonding | Are commercial terms reflected in the cash profile? | Retention ignored in the funding forecast |
| Allowances and provisional sums | Are they identified separately and adjusted on a defined basis? | Provisional sums treated as firm |
[!IMPORTANT] Markup on markup is the most frequent commercial finding in change pricing. A subcontractor adds overhead and profit, the main contractor then applies its own percentage to the subcontractor's marked-up total, and the owner pays profit on profit. Most contracts limit the base to which each markup applies. Confirming that base is a routine part of task 2.H and a recurring exam scenario.
A project is 92% complete, is four months behind its contractual completion date, and reports SPI of 0.97. The sponsor concludes the schedule position is nearly recovered. What is wrong with that reading?
A project reports cumulative CPI of 0.85, stable for six reporting periods, and TCPI to the original BAC of 1.18. The project manager proposes reporting to the board that the team will now deliver at 1.18 efficiency and finish within the original budget. How should the cost engineer respond?
A main contractor submits a change order in which a subcontractor's cost of $400,000 carries the subcontractor's 15% overhead and profit, and the main contractor then applies its contracted 10% markup to the resulting $460,000. What should the cost engineer check, and what is the likely finding?