LEARN. PRACTICE. GROW.Read the obligation before applying a contract label
The lab uses invented conditions to practice decisions rather than interpret a real contract. A fixed price covers described scope. If the proposal names ten applications, two additional ones require the agreed change; dividing price by ten does not create an accepted rate. Distinguish price, supplier-incurred cost, fee, and expenditure authority. A fee ceiling can coexist with variable total cost. Before recommending purchase, identify conditions affecting timing, quality, customer dependencies, and changes. The PM coordinates analysis with procurement and competent owners within their authority.
Calculate a fee with explicit bounds
In the first model, all values are thousands of euros. Target cost is 100 and target fee 10. Calculate raw fee=10+0.2×(100−actual cost); apply a floor 6 and ceiling 14; payment=actual cost+fee. At cost 80, the fee is 14 and payment 94. At cost 130, the formula gives 4, but the floor makes the fee 6 and payment 136. The 14 bound affects fee; it does not establish a 114 cap on total payment. Operation order and units are part of the stem. These calculations assume neither additional clauses nor taxes.
Compare with a total-price ceiling
The second model differs: price=the lesser of 125 and [actual cost+10+0.2×(100−actual cost)]. It does not include the first model’s fee floor. At cost 130, preliminary price is 134, but the ceiling yields 125. Supplier result is 125−130=−5. Do not carry conditions between exercises merely because both mention incentives. A potential loss can affect behavior, quality, or collaboration; monitoring remains necessary. Financial risk allocation does not remove delivery failure’s impact on the buyer’s business.
Separate eligibility, scoring, and tie-breaking
Night coverage is mandatory in this fictional procurement. A proposal excluding it does not become eligible through low price. Apply clarification and eligibility rules before weighted comparison. For eligible proposals, technical 50%, operations 30%, commercial 20% weights yield 7.6 for both A=[8,6,9] and B=[7,9,7]. The tie requires the predefined procedure. Do not adjust weights after learning the desired winner. A legitimate criteria change requires authority, traceability, and consistent handling of affected proposals.
Observe behavior produced by the incentive
Paying per migrated server can focus attention on creating machines, while the intended outcome requires a working application and rehearsed recovery. Reconciling reported units with acceptance avoids counting activity as outcome. If incentives create adverse effects, discuss adjustments with procurement without unilaterally changing the contract. For the fictional 40-thousand milestone,20% is retained until final testing: 32 thousand may become eligible after acceptance, but still require approval. None of these calculations independently establishes complete-service operating readiness.
Negotiate constraints and check common dependencies
An eight-hour window contains three preparation hours that can occur beforehand without downtime. Separating them reduces the technical window to five, still above the business’s accepted four. Bring the remaining gap to a decision with executable options. Also confirm promised capability: two suppliers may depend on the same subcontracted specialist. If that person is booked for another client, two independent capabilities have not been established. Identify qualified alternatives, coverage, and sequencing before the window. Do not quantify risk reduction merely by contract count.
Guided practice: recommendation to procurement
Recalculate 94 and 136 in the first model,125 and loss 5 in the second, and 32 of eligible milestone payment. Explain why these results are not interchangeable. Then apply mandatory coverage and calculate the 7.6 tie. Write a recommendation containing the desired outcome, eligibility evidence, financial exposure, dependencies, alternatives, and required decision. For a scope discrepancy, compare records and use the agreed mechanism with authorized representatives. The report should enable a decision rather than merely count tickets or show busy hours.
Synthetic contract fixtures; thousands of EUR; no real contract interpretation.
model_A_fee = clamp(10 + 0.20 * (100 - actual_cost), 6, 14)
model_A_payment = actual_cost + model_A_fee
cost_80 -> fee_14, payment_94
cost_130 -> fee_6, payment_136
model_B_price = min(125, actual_cost + 10 + 0.20 * (100 - actual_cost))
cost_130 -> price_125, supplier_result_minus_5
weights = [0.50, 0.30, 0.20]
A = [8, 6, 9] -> 7.6
B = [7, 9, 7] -> 7.6
eligible_milestone_payment = 40 * (1 - 0.20) -> 32Two proposals tie at 7.6, but only proposals meeting mandatory coverage reach tie-breaking. A favorable price does not create that coverage.
Common pitfalls
Importing clauses from another model; confusing fee with total price; compensating mandatory gates with points; counting contracts as independent capability.
Related topics: Suppliers and acceptance · Value and closure · Governance and risk
Supplier decisions combine conditions, evidence, and risk. Calculation helps when units, formula, and scope are explicit.
Reference: The special challenges of project management under fixed-price contracts · PMP ECO July 2026; DR PMP 2026.7