ChatGPT Prompts for Business Analysts

<context> I am a business analyst preparing to elicit requirements for [PROJECT]. The primary interviewee is [STAKEHOLDER], whose role and goals I will describe below. The current pain point we are solving is the inefficiency of [PROCESS]. Known context: [paste project background, scope boundaries, and any constraints] </context> <task> 1. Produce 12-15 open-ended elicitation questions tailored to [STAKEHOLDER], grouped into themes: current-state workflow, pain points, desired outcomes, edge cases, and success metrics. 2. For each question, add a one-line note on what requirement signal it is designed to surface. 3. Flag 3 follow-up probes I should use if answers stay vague. 4. End with 3 questions specifically designed to uncover unstated or assumed requirements. </task>

<context> I just finished a requirements workshop for [PROJECT] with [STAKEHOLDER]. Below are my raw, unstructured notes. Raw notes: [paste messy notes, bullet fragments, and quotes here] </context> <task> 1. Extract every distinct requirement and classify each as functional, non-functional, or business rule. 2. Rewrite each in the format: 'The system shall...' with a unique ID (REQ-001, REQ-002...). 3. Assign a MoSCoW priority (Must/Should/Could/Won't) and note the assumption behind that priority. 4. List any ambiguities, conflicts, or gaps that need a follow-up with [STAKEHOLDER]. 5. Output as a markdown table: ID | Requirement | Type | Priority | Source | Open Question. </task>

<context> Here is a draft requirements list for [PROJECT]. Requirements: [paste your current requirement statements or table] </context> <task> 1. Review the set for gaps using a checklist: error handling, security, performance, data retention, access control, reporting, and edge cases. 2. List any requirements that conflict or overlap, and explain the conflict. 3. Identify requirements that are untestable or ambiguous, and rewrite each to be measurable. 4. Suggest 5 requirements commonly needed for projects like this but absent from the list. 5. Output gaps, conflicts, and suggestions as three separate prioritized sections. </task>

<context> I need acceptance criteria for a requirement on [PROJECT]. Requirement: [paste the single requirement statement] Relevant business rules: [paste any rules or constraints] </context> <task> 1. Write acceptance criteria in Given/When/Then (Gherkin) format covering the happy path. 2. Add criteria for at least 3 negative or edge cases (invalid input, boundary values, permission denial). 3. Make every criterion independently testable and free of implementation detail. 4. Flag any data, integration, or business-rule dependency the criteria assume. 5. Output as a numbered Gherkin list. </task>

<context> I am building a requirements traceability matrix for [PROJECT]. Requirements: [paste requirement IDs and statements] Business objectives: [paste the goals these requirements support] </context> <task> 1. Map each requirement to the business objective(s) it supports. 2. Add columns for: linked test case (placeholder ID), source stakeholder, design artifact, and status. 3. Flag any requirement that does not trace to a business objective (potential scope creep). 4. Flag any objective with no supporting requirement (potential gap). 5. Output as a markdown traceability matrix plus a short summary of orphaned items. </task>

<context> I am mapping the current-state of [PROCESS] within [PROJECT]. Below is how a [STAKEHOLDER] described it to me. Description: [paste the narrative of how the process works today] </context> <task> 1. Reconstruct the as-is process as a numbered step-by-step flow, identifying actor, action, input, and output for each step. 2. List all decision points and their branches. 3. Identify handoffs between roles or systems and flag where ownership is unclear. 4. Note any steps that appear manual, redundant, or out of sequence. 5. Output the flow as a numbered list and a separate table of decision points. </task>

<context> Here is the current-state map of [PROCESS] for [PROJECT]. Process steps: [paste the as-is steps, including any timing or volume data] </context> <task> 1. Identify bottlenecks, delays, rework loops, and non-value-adding steps. 2. For each issue, estimate the likely impact (time, cost, error risk) and the root cause. 3. Apply a lean lens: classify waste using the categories of waiting, overprocessing, defects, and handoffs. 4. Rank the issues by impact and ease of fixing (quick wins vs structural changes). 5. Output as a prioritized table: Issue | Type | Impact | Root Cause | Fix Effort. </task>

<context> I need to design the to-be version of [PROCESS] for [PROJECT]. Current-state issues: [paste the bottlenecks and pain points identified] Constraints: [paste budget, system, headcount, or compliance constraints] </context> <task> 1. Propose a redesigned to-be process as a numbered step flow, removing or automating the identified waste. 2. For each change, state the benefit and which current-state issue it resolves. 3. Highlight which steps require new tooling, integration, or policy change. 4. Note risks introduced by the redesign and a mitigation for each. 5. Output the to-be flow, a before/after comparison table, and a change-impact summary. </task>

<context> I need a SIPOC and a swimlane breakdown for [PROCESS] in [PROJECT]. Process overview: [paste a short description of the process and the roles involved] </context> <task> 1. Build a SIPOC table: Suppliers, Inputs, Process (5-7 high-level steps), Outputs, Customers. 2. Then break the process into swimlanes by actor or system, listing each step under its owning lane. 3. Mark every cross-lane handoff explicitly. 4. Identify any lane that is overloaded or any handoff with no clear trigger. 5. Output the SIPOC table, the swimlane breakdown, and a short observations note. </task>

<context> A recurring failure is happening in [PROCESS] for [PROJECT]. Problem statement: [describe the symptom, frequency, and observed impact] What we know: [paste any data, logs, or stakeholder observations] </context> <task> 1. Apply the 5 Whys technique to drill from symptom to likely root cause(s); show each layer. 2. Build a fishbone (Ishikawa) breakdown across categories: People, Process, Technology, Data, Policy. 3. Distinguish probable root causes from contributing factors. 4. Recommend the evidence I should gather to confirm each candidate root cause. 5. Output the 5 Whys chain, the fishbone breakdown, and a verification checklist. </task>

<context> I am running stakeholder analysis for [PROJECT]. Stakeholders: [list names/roles, e.g. [STAKEHOLDER] and others, with what you know about each] </context> <task> 1. Place each stakeholder on a power/interest grid (High/Low power x High/Low interest) and explain the placement. 2. For each, recommend an engagement strategy (manage closely, keep satisfied, keep informed, monitor). 3. Note each stakeholder's likely concerns, success criteria, and potential objections. 4. Identify hidden or missing stakeholders I may have overlooked. 5. Output a stakeholder table plus a quadrant-by-quadrant engagement plan. </task>

<context> I need a RACI matrix for [PROJECT]. Key activities/deliverables: [list the decisions and deliverables] People and roles: [list roles including [STAKEHOLDER]] </context> <task> 1. Build a RACI matrix mapping each activity to roles as Responsible, Accountable, Consulted, or Informed. 2. Enforce exactly one Accountable per activity and flag any violation. 3. Highlight activities with no Responsible owner or too many Consulted parties. 4. Note where ambiguity could cause delays and recommend a fix. 5. Output the RACI matrix and a short list of clarity risks. </task>

<context> I need to communicate an update on [PROJECT] to [STAKEHOLDER]. The message: [paste what needs to be communicated, e.g. a scope change or delay] Stakeholder profile: [role, what they care about, tolerance for detail] </context> <task> 1. Rewrite the message for [STAKEHOLDER], matching their concerns, seniority, and preferred level of detail. 2. Lead with the impact most relevant to them, not the background. 3. Anticipate their top 2 likely questions or objections and pre-empt each. 4. Provide a 1-line TL;DR version and a fuller version. 5. Keep tone professional and free of jargon they would not use. </task>

<context> I have a tense conversation coming up with [STAKEHOLDER] about [PROJECT]. The situation: [describe the tension, e.g. conflicting priorities, a rejected requirement, or a missed expectation] </context> <task> 1. Summarize the likely position and underlying interest of [STAKEHOLDER]. 2. Identify the common ground and the points where we genuinely diverge. 3. Script 3 ways to open the conversation that de-escalate and stay solution-focused. 4. Prepare responses to their 3 most likely pushbacks. 5. Recommend the outcome to aim for and a fallback if agreement is not reached. </task>

<context> Two stakeholders on [PROJECT] want conflicting things. Stakeholder A wants: [paste position] Stakeholder B (e.g. [STAKEHOLDER]) wants: [paste position] Project constraints: [paste relevant constraints] </context> <task> 1. Restate each position and the underlying business need behind it. 2. Analyze where the conflict is real vs a misunderstanding of terms. 3. Propose 3 resolution options (compromise, phased delivery, escalate) with trade-offs. 4. Recommend the option best aligned with the project objectives and explain why. 5. Draft a neutral summary I can circulate to both parties to confirm alignment. </task>