ChatGPT prompt to enhance quality

Create a quality-improvement plan for [PROCESS OR SERVICE] using the evidence below. I need a plan people can execute, not a general list of ideas.

Problem and affected customers or users: [QUALITY PROBLEM]
Evidence, defect data, examples, and current baseline: [EVIDENCE AND BASELINE]
Current workflow, controls, and constraints: [CURRENT PROCESS]
Available people, authority, budget, and target date: [RESOURCES AND DEADLINE]

Start with a concise problem statement that defines the defect, where it occurs, who is affected, baseline rate or frequency if known, and desired outcome. Then distinguish observed facts from likely causes. Do not call something a root cause unless the supplied evidence supports it; list untested causes as hypotheses.

Produce these sections:
1. A cause-and-control table: failure point, evidence, likely cause or hypothesis, immediate containment, permanent corrective action, preventive control, owner, due date, and proof of completion.
2. A 30/60/90-day plan, adjusted to the stated deadline, with no more than five actions per period.
3. A measurement plan: leading measures, outcome measures, data source, baseline, target, sampling cadence, and escalation threshold. Do not invent numbers; mark missing baselines as “to establish.”
4. A RACI-style decision list for actions needing approval or cross-functional support.
5. A short communication note for affected staff explaining the changed standard and how exceptions are reported.

Prioritize containment before permanent fixes when the issue could harm customers, safety, compliance, or revenue. Include a check that the proposed control fits the actual workflow rather than adding an approval step with no owner. Ask up to 3 clarifying questions only if a required input is missing.

Before answering, verify every action has one accountable owner and a completion test, and flag recommendations that need specialist, legal, safety, or regulatory review.

Build a service-quality recovery plan for [SERVICE]. Customer complaints and examples are [COMPLAINT EVIDENCE]. Our current service workflow is [WORKFLOW], and available owners are [OWNERS]. Produce a customer-impact triage, a cause-and-control table, a four-week action plan, response standards for open complaints, and metrics for recurrence and resolution. Separate confirmed failures from hypotheses, and do not invent complaint volumes or customer promises. Ask up to 3 questions only if needed. Check that urgent customers have containment actions and each permanent fix has an accountable owner and proof of completion.

Create a corrective and preventive action plan for [PRODUCT OR LINE]. The defect description, lot information, and inspection results are [DEFECT DATA]. Process steps and controls are [PROCESS DETAILS]. Regulatory, safety, and release requirements are [REQUIREMENTS]. Produce a containment decision, traceability and investigation steps, a cause-hypothesis log, corrective and preventive actions, verification sampling, and release criteria. Do not diagnose a root cause beyond the evidence or replace required engineering or quality review. Ask up to 3 questions if necessary. Check that affected lots are accounted for and every action has evidence of effectiveness.

Design a content-quality improvement plan for [CONTENT TYPE]. Examples of weak and strong work are [EXAMPLES]. The current creation and approval workflow is [WORKFLOW], and the intended audience and standard are [AUDIENCE AND STANDARD]. Produce a usable scorecard with weighted criteria, a pre-publish checklist, calibration process for reviewers, revision workflow, and monthly measures. Define observable criteria such as factual support, audience fit, structure, and accessibility; do not use taste as the sole standard. Ask up to 3 questions only if needed. Check that the scorecard can be applied consistently by two reviewers and that every criterion has a pass condition.