16.2 Quality Control and Success Criteria
Key Takeaways
- Success criteria define what a successful project looks like overall; acceptance criteria define what a specific deliverable must meet to be accepted.
- Both must be agreed in advance, or acceptance becomes a negotiation at the point of handover when leverage rather than evidence decides.
- Quality control techniques include inspection, testing, peer and technical review, and statistical sampling where full inspection is impractical.
- Quality planning defines the standard, quality control checks the deliverable against it, and quality assurance checks that the processes producing it are adequate.
- Quality control finds defects; it does not by itself prevent them, which is why planning and assurance matter alongside it.
Outcome 18b requires knowledge of how quality control techniques are used to determine whether success criteria are met. Two things are being tested at once: the techniques themselves, and the discipline of measuring against criteria that were agreed in advance rather than argued about at handover.
Success criteria and acceptance criteria
Be precise with language in exam answers:
- Success criteria — measures that define whether the project (or a major outcome) is successful, often including time, cost, quality, benefits readiness, or compliance goals agreed with the sponsor.
- Acceptance criteria — specific conditions a deliverable or requirement must meet to be accepted by the customer/user/operations.
Quality control determines whether outputs meet acceptance criteria and, aggregating evidence, whether quality-related success criteria are being met.
| Level | Example |
|---|---|
| Success criterion | "Go-live with zero critical defects open and user training ≥95% complete by 30 Sept" |
| Acceptance criterion | "Payment API rejects unauthenticated calls with HTTP 401 in 100% of security test cases" |
Quality control techniques
Quality control is the set of activities that determine whether products meet required standards and success/acceptance criteria. Techniques should be selected for product type, risk, and life-cycle stage.
Inspection
Inspection examines a product or material against a specification — visually or with instruments. Common in construction, manufacturing, and document checks (for example inspecting a completed installation against drawings, or inspecting goods received).
- Strengths: concrete verification of physical or documented attributes
- Limits: can be late if only end-point; may miss latent defects not visible at inspection time
- Planning need: checklists, standards, competent inspectors, quarantine of non-conforming items
Testing
Testing exercises the product under controlled conditions to see whether it behaves as required. Includes unit, integration, system, user acceptance, performance, security, and factory/site acceptance tests depending on domain.
- Strengths: reveals functional and performance failures before full live use
- Limits: tests only what is designed to be tested; poor requirements produce poor tests
- Planning need: test strategy, cases traced to requirements, environments, data, entry/exit criteria, defect management
Reviews
Reviews are structured examinations of documents, designs, plans, or code by peers or stakeholders (walkthroughs, technical reviews, inspections of documents, gate reviews). They catch defects early when cheaper to fix.
- Strengths: early detection; knowledge sharing; improves requirements and design quality
- Limits: effectiveness depends on preparation, skill, and psychological safety to raise issues
- Planning need: review types, checklists, exit criteria, action tracking
Sampling
Sampling checks a statistical or selected subset when inspecting or testing 100% is impractical or destructive. Used in manufacturing lots, large document populations, or high-volume transactions.
- Strengths: efficient control for large populations
- Limits: sample risk — defects outside the sample may escape; needs sound sampling design
- Planning need: sample size/method, acceptance quality limits, response when samples fail
| Technique | Typical question it answers | Example |
|---|---|---|
| Inspection | Does this item conform to the specified attributes? | Cable termination torque and labelling check |
| Testing | Does it behave correctly under defined conditions? | UAT script for booking a clinic appointment |
| Review | Is this design/document adequate before we proceed? | Design review of interface specification |
| Sampling | Is the batch/population likely within quality limits? | Random sample of 50 welded joints from 2,000 |
Often techniques combine: review the test plan, inspect the test environment build, run tests, sample production transactions after go-live.
Determining whether success criteria are met
QC does not exist for its own sake. Evidence from inspections, tests, reviews, and sampling is compared to planned criteria:
- Define criteria in quality planning (from requirements and success measures)
- Execute control activities and record results
- Judge pass/fail against criteria and severity rules
- Raise non-conformances/defects; correct or concede with authority
- Re-verify fixes
- Report status to governance (can we pass the gate / accept the product?)
If criteria are unmet, options include rework, de-scope with change control, delay acceptance, or — if residual risk is accepted by the proper authority — conditional acceptance with known limitations. The project manager must not silently redefine "success" to make a failing product look green.
Scenario C — choosing techniques
A bridge bearing delivery: inspection and certificates for materials; sampling of torque checks; review of installation method statements; load testing as specified. A mobile app: peer review of critical code; automated and manual testing; limited beta sampling of real user sessions; accessibility inspection against standards. Technique choice follows risk and product type.
Scenario D — success criteria vs wishful acceptance
Success criteria require fewer than five high-severity defects at go-live. Testing still shows twelve. Pressure mounts to "accept and fix live." Appropriate quality control governance: report the evidence, quantify residual risk and benefits impact, and escalate for a decision to delay, reduce scope, or add hypercare — not to hide results.
Quality planning vs control vs assurance (brief distinction)
| Focus | Core question | Example activity |
|---|---|---|
| Planning | What quality will we achieve and how? | Write quality management plan and acceptance criteria |
| Control | Does this product meet criteria? | Run UAT; inspect welds; sample batches |
| Assurance | Are quality processes adequate and followed, giving confidence to stakeholders? | Process audit; independent compliance check of the quality system; PMO health check on whether reviews actually happen |
Assurance and control support each other: assurance may find that testing processes are weak; control produces the product evidence sponsors need. On the PMQ, if asked about determining whether success criteria are met, emphasise control techniques and evidence; if asked about confidence in management of quality, bring in assurance.
Answer pattern for LO18 long responses
When a scenario shows defects, failed go-live, or arguments about "good enough":
- Define quality as fitness for agreed requirements/purpose, not perfectionism.
- Show quality planning contents: standards, criteria, indicators, methods, roles, timing.
- Link indicators and criteria to the business case and benefits assumptions.
- Select QC techniques (inspection, testing, reviews, sampling) suited to the product and risk.
- Explain how evidence judges success/acceptance criteria, with governance options if failed.
- Briefly separate QA if the question asks about process confidence or independent checks.
That structure demonstrates professional quality management aligned to APM PMQ LO18 rather than a vague promise to "ensure high quality."
A manufacturer receives 10,000 identical components and cannot practically inspect every unit in full. Which quality control approach is most appropriate as part of the control regime?