Triple Constraint
The triple constraint is the model that links three constraints of a project: scope, what it delivers; schedule, the date by which it delivers it; cost, what it spends to deliver it. All three depend on the same work. What is fixed determines what absorbs change. Quality sits at the centre of the triangle, as a fourth constraint or as a threshold. The PMBOK speaks of competing constraints. For adaptive approaches it presents two options: fix scope and let budget or schedule vary; fix budget and schedule and let scope vary.
Goal
The triple constraint is a project management model that links scope, schedule and cost. The model produces no figures. It forces an answer, before the first surprise, to a question that sponsors put off: when something changes, which of the three gives way?
The answer belongs to the sponsor. The project manager prepares it with figures but does not decide it. The sponsor gives it at launch and signs it in the project charter. The project manager checks every change request against it: if the change goes in, an adjustable constraint pays the price.
The deliverable fits on one page: the status of each constraint, fixed or adjustable, the quantified margin of the adjustable constraints, a verifiable quality threshold and the name of the arbiter. It is used to accept or reject a change, to choose the lever when the project slips and to name what the project will sacrifice if a surprise exceeds the margins.
Usage
When to use
- Project launch: decide which constraint gives way before the first surprise and write it into the charter.
- Change request: quantify its effect on the three constraints before accepting it.
- Delay or overrun: choose between paying, pushing back the date or cutting scope.
- Date imposed by a third party: a law coming into force, the end of support for a software product, the year-end closing. The schedule is fixed; classify the other two.
- Commitment credit approved by parliament: any overrun requires an additional credit. Cost is fixed; classify the other two.
- Choice of contract type: fixed price, time and materials or a budget per iteration allocate the constraints between client and supplier.
- Sponsor who demands all three: show with figures that one constraint must give way and which one.
When not to use
- Judging whether a project is worth its cost or comparing projects: the model measures execution; use the business case and net present value.
- Estimating a duration or a cost: the model says which constraint moves, not by how much; use estimation.
Description
Three constraints linked by the work
Scope is the set of deliverables and features that the project commits to produce. Schedule is the end date or the date of an imposed milestone. Cost is the sum committed: salaries, external services, licences, hardware. For a Swiss public project it often takes the form of a commitment credit voted by a parliament.
The link runs through the work and the team's capacity. Scope determines the amount of work. Divided by capacity, it gives a duration. Capacity multiplied by duration gives most of the cost. Fixing two constraints therefore amounts to calculating the third. A date and a budget determine how much capacity the project can buy over the period and so how much work gets done.
Yet the model misleads anyone who reads it as an equation. An added person costs money from the first day and gives time back only once up to speed. On a software project that is already late, Brooks's law predicts that adding people makes it later. Money spent off the critical path does not bring the end date forward.
Quality: at the centre or a fourth constraint
The original triangle had different corners. Martin Barnes, a founding member of the UK's Association for Project Management (APM), drew a triangle of time, cost and quality. He explained in 2012, in the APM's journal, that he wanted someone to answer for the quality of the product, alongside the engineers who held the budget and the schedule. The quality, cost, delivery (QCD) triangle, common in French-speaking industry, carries the same three corners. The usual version in project management puts scope at the top and quality at the centre.
The PMBOK places quality within scope. The strength of a bridge, its service life or its ease of maintenance is paid for in cost and schedule, like a feature. The PMBOK also measures the effectiveness of project management by compliance with constraints such as cost, scope, schedule and quality.
Treat quality as a threshold. The sponsor sets it at launch as verifiable criteria, and no trade-off goes below it without the sponsor's explicit decision. Quality left out of the classification becomes the constraint that gives way in silence.
What is fixed, what varies
In a predictive approach, the PMBOK derives the schedule and cost baselines from the scope to be delivered. Together the three form the integrated baseline. For adaptive approaches, the PMBOK presents two options, which it calls the inverted triangle. In the first, scope is fixed and budget or schedule absorb unforeseen changes. In the second, budget and schedule are fixed and scope absorbs the changes.
The same reasoning holds for any project with a constraint imposed from outside, whatever the approach: what is fixed determines what absorbs change.
| Fixed constraints | Constraint that gives way | Lever | Technique |
|---|---|---|---|
| Scope and schedule | Cost or increased risk | Add resources on the critical path or overlap activities planned in sequence | Crashing and fast tracking |
| Schedule and cost | Scope | Deliver first what carries the most value, defer the rest | Time boxing and budgeting, prioritization |
| Scope and cost | Schedule | Recalculate the end date with the available capacity | Critical path method |
| All three | Quality (shortened testing, deferred documentation, technical debt) | None: release a constraint | None: scope and constraints were not set properly at launch |
The contract has often already decided
With a supplier, the contract type allocates the constraints. A fixed-price contract fixes the price for a defined scope, usually by an agreed date. The supplier carries the risk of exceeding its costs and builds it into its price. For the client, all three constraints are then fixed: the pressure lands on quality unless the contract writes down its acceptance criteria (step 3 below). A time-and-materials contract bills the hours worked: cost becomes the adjustable constraint, and the client carries the risk. An agile contract with a fixed budget per iteration fixes the cost and the duration of each iteration. The number of iterations determines the total; scope is left to the client's prioritization. Reading the contract before classifying the constraints avoids reopening what the signed contract already settles.
Running the trade-off
- Name the origin of each constraint. A constraint is fixed when a third party imposes it: a law coming into force, a contractual penalty, a credit approved by parliament, a market window. A constraint that nobody imposed is a preference and stays adjustable.
- Classify. Each constraint receives a status: fixed or adjustable within a quantified margin, for example 10% of the budget or six weeks. At most two constraints are fixed. If all three are, the sponsor releases one before launch.
- Set the quality threshold. The sponsor writes it as criteria that acceptance testing can verify: acceptance criteria, test coverage, an accessibility standard, data protection requirements.
- Record the classification and the arbiter. The project manager records the statuses, the margins and the name of the arbiter in the project charter. A steering committee can hold this role.
- Apply the rule to every change. Every change request quantifies its effect on the three constraints. The change goes in if an adjustable constraint absorbs it within its margin. Otherwise it goes up to the arbiter, who rejects it or revises the classification.
- Review at each milestone. A classification changes when its causes change: an additional credit approved by parliament, a date pushed back by the legislator, a competitor who launches its product first. The review is recorded as a decision of the sponsor.
What makes the trade-off fail
All three constraints fixed. A project launched this way has no margin. Officially, the first surprise is not allowed to affect scope, date or budget, and quality gives way: shortened testing, deferred documentation, technical debt recorded nowhere. The defect shows up in operation, once the project is closed and its budget met.
Implicit classification. A sponsor who has never said which constraint gives way decides it during the crisis, with no margin and under the pressure of the date. The project manager then makes the trade-offs alone, week after week, through small concessions that nobody approved.
A scope that cannot be split. Letting scope vary presupposes splitting it at launch into elements that can be delivered separately and are ranked by value. An indivisible block is delivered whole or not at all.
Mistaking a project that met its scope, date and budget for a success. The PMBOK judges a project's success on two dimensions: the value of its outcome and the effectiveness of its management. The triangle measures only the second. The Sydney Opera House was meant to cost 7 million Australian dollars and take four years. It cost 102 million and took fourteen. Its management counts as a failure; its outcome is on UNESCO's World Heritage List. In Montreal, a well-managed viaduct on Autoroute 15, at a cost close to 11 million Canadian dollars, was demolished a year after it was built: it did not fit the redevelopment of the neighbouring Champlain Bridge.
AI considerations
The PMBOK lists baseline optimisation among the uses of AI in project management: AI strengthens the analysis of trade-offs between scope, schedule and cost. An assistant reads a change request, matches it against the estimated backlog and the schedule, then drafts its effect on the three constraints with its assumptions. It produces the three variants of the same change: pay, postpone, reduce. It also reviews a charter and flags a contradictory classification, for example three fixed constraints.
It does not choose the constraint that gives way. That choice commits the sponsor, who answers for it to a parliament, a client or a regulator. The assistant's figures are only as good as the estimates it receives: a model that fills in a missing effort estimate from projects outside the organisation produces false precision. Contract amounts and supplier rates are confidential. The project manager checks the organisation's policy and the FADP before entering them into an online service.
Examples
Concept illustrated: when schedule and cost are fixed, scope absorbs the change. A canton puts its building permit procedures online in time for its revised law to come into force. At launch, the sponsor signs the classification of the constraints.
| Constraint | Status | Origin | Margin |
|---|---|---|---|
| Schedule | Fixed | Revised law in force on 1 January 2027 | None |
| Cost | Fixed | Commitment credit of CHF 1'200'000 voted by the cantonal parliament | None: any overrun requires an additional credit |
| Scope | Adjustable: 14 online procedures, 9 of them required by 1 January | Revised law (9 procedures) and requests from the department (5) | The 5 procedures not required, which can be deferred to a second batch funded outside this credit |
| Quality | Threshold | Accessibility to WCAG 2.1 level AA, data protection, acceptance by the building department | None |
| Arbiter | Steering committee, chaired by the head of the building department | ||
In the fifth month, the legal department asks for a procedure that the revised law requires and that the specification had left out: the application for an exemption.
| Schedule (weeks on the critical path) | Cost | Scope on 1 January | |
|---|---|---|---|
| Exemption application, added | +5 weeks | +CHF 70'000 | +1 required procedure |
| Permit extension, deferred to the second batch | −3 weeks | −CHF 40'000 | −1 non-required procedure |
| Notice of completion of works, deferred to the second batch | −2 weeks | −CHF 35'000 | −1 non-required procedure |
| Balance | 0 | −CHF 5'000 | 13 procedures, including the 10 required |
The committee debated neither the date nor the credit: the classification signed at launch had settled them. The meeting was about which two procedures to defer.
Visualisations
Whether scope alone is fixed or schedule and cost are, quality remains a threshold at the centre. Only the number of constraints that absorb change varies: two in the first case, one in the second.
Cost
| Phase | Level | Justification |
|---|---|---|
| Preparation | Medium | Trace the origin of each constraint to the law, the contract or the credit, then obtain statuses and quantified margins from the sponsor. |
| Execution | Low | One meeting at launch, then one line of analysis per change request, once the effort is estimated. |
| Documentation | Low | One page in the project charter and one column per constraint in the change log. |
Tooling
A page of the project charter or a spreadsheet is enough to classify the constraints. The change request log, kept in a spreadsheet or a tracking tool such as Jira, gets one column per constraint. Scheduling software such as MS Project or ProjectLibre compares a variant with the baseline schedule and shows the effect of a change on the end date.
In an adaptive approach, a burn-up chart plots total scope above completed work. Each addition raises the scope line, and the sponsor reads what it costs in date or in deferred features.
Sources
- Project Management Institute, The Standard for Project Management (PMBOK Guide, 8th edition), PMI: the term competing constraints (§2.1); the two dimensions of project success, with the Sydney Opera House and the Montreal viaduct (§2.1.2); the integrated baseline of predictive approaches and the constraint options of adaptive approaches, with the inverted triangle of figure 4-3 (§4.2).
- Project Management Institute, A Guide to the Project Management Body of Knowledge (PMBOK Guide), 8th edition, PMI: quality as a characteristic of scope (§2.2.1, key concepts of the Scope Performance Domain); AI in the analysis of trade-offs between scope, schedule and cost (Appendix X3, table X3-1).
- Association for Project Management, A tribute to APM founder, Dr Martin Barnes CBE (2022): the origin of the time, cost and quality triangle in the work of Martin Barnes and the reason he gave for drawing it.