6 Quality Planning and Delivery

What project quality means

Project quality is the degree to which a deliverable and the processes used to create it meet agreed requirements and stakeholder needs. Quality does not simply mean high-end performance: a product is fit for purpose when it meets its defined requirements and can be used as intended.

Quality management involves planning how to achieve quality, checking both processes and deliverables, and responding to evidence of problems or opportunities to improve.

Set objectives and acceptance criteria

Quality objectives turn desired quality outcomes into guidance for project decisions and a basis for evaluation. Make them specific and measurable. For each objective, identify the requirement or stakeholder need it addresses, an observable or measurable metric, a target or threshold for success, and the measurement method, timing, and owner.

For example, “the service should be reliable” is too vague to assess. A measurable objective could state that, during a 30-day pilot, the service will be available for at least $99.5\%$ of scheduled operating time, excluding approved maintenance. The plan should also explain how availability is calculated and who reviews the results.

Acceptance criteria define the evidence that demonstrates a requirement has been delivered. A requirement describes what is needed; the criteria state what a deliverable must satisfy to be accepted by the designated customer or approver. Write criteria before work begins, link them to requirements, and make them clear enough for different reviewers to reach consistent decisions.

Criteria can address function, performance, accuracy, safety, accessibility, compliance, or other relevant attributes. If a criterion is difficult to measure, specify the review method and who has authority to approve the result. Keep requirements, criteria, tests, and approvals traceable so the team can confirm that the agreed scope has been met.

Examples of objectives and checks include:

• Online registration form: The objective is to collect complete, valid registrations. A valid submission is stored with all required fields, while invalid email formats are rejected with a clear message. Check this with test cases and a review of stored records.

• Training guide: The objective is to help new staff complete a task correctly. At least eight of ten pilot users should complete the task without assistance, as assessed through an observed usability test.

Plan quality activities

The quality management plan records how the project will meet its quality objectives. It typically identifies standards and procedures, roles and approval authority, quality metrics, review and test activities, required records, and how nonconforming work and changes will be handled. Match the plan’s level of detail to the project’s risk, size, and regulatory or contractual obligations.

Plan assurance and control activities early, include them in the schedule, and assign the necessary time and resources. For each check, define what will be examined, the method or sampling approach, when it will occur, who will perform it, what evidence will be recorded, and what will happen if the result does not meet the criterion. Use more rigorous or frequent checks when failure would have greater consequences. Independent review can be useful for high-risk or regulated work.

Distinguish assurance, control, and evaluation

Quality assurance (QA) is preventive and process-focused: it examines whether processes are suitable and followed. Examples include auditing whether reviews occur as planned, checking that suppliers follow agreed procedures, and improving a test process that repeatedly misses errors.

Quality control (QC) is product-focused: it checks whether work products meet requirements. It includes inspection, testing, measurement, and review. For example, a team might test whether a feature works across supported browsers or inspect a shipment for damage and compliance with specifications.

Verification and validation answer different but related questions. Verification checks whether a deliverable conforms to its specified requirements—“Did we build it right?” Validation checks whether it meets users’ needs in its intended context—“Did we build the right thing?” A project may need both, such as technical testing followed by a user pilot.

Manage defects and nonconforming work

A defect is a flaw or failure to meet a requirement. Nonconforming work is any output that does not satisfy an applicable criterion. When a problem is found, use an authorized response:

1. Record and assess it. Describe the issue, the affected requirement or criterion, its severity, and the available evidence. Prioritize according to impact and risk.

2. Contain the impact. Prevent the defective item from being used, released, or passed to later work until its status is clear.

3. Decide what to do. Depending on authority and project rules, correct or rework the item, replace it, or request a formally approved change or concession. Do not quietly lower the acceptance standard.

4. Verify the resolution. Recheck corrected work against the criterion and consider whether related work may also be affected.

5. Investigate recurring or serious issues. Look beyond the immediate symptom for possible causes in requirements, design, training, tools, suppliers, or processes.

Keep defect records visible to the people who need to act. A useful record links the issue to its requirement, owner, priority, status, resolution, and verification result. If defects recur, improve the process that produces them or fails to detect them rather than relying only on final inspection.

Measure and improve quality

Quality measures help a team determine whether its approach is working. Depending on the project, useful measures may include first-pass acceptance rate, defect count and severity, rework time, test coverage, supplier nonconformance, or user-reported problems. Define measures consistently and review trends rather than isolated results. Avoid rewarding a metric in a way that encourages hidden defects or unnecessary testing.

Use a deliberate improvement cycle when results fall short or a process could work better:

1. Define the problem or improvement goal.

2. Collect relevant evidence and look for patterns or root causes.

3. Select and implement a change.

4. Measure whether the change produced the intended result.

5. Incorporate effective practices into routine work and share the learning.

For example, if reviews repeatedly find the same data-entry errors, the team might examine the form design and instructions, test a revised version, and compare error rates. An improvement is not successful merely because it was implemented; its effects should be checked. Lessons learned can inform remaining project work and future projects.