Claude Prompt Library

30 Claude Prompts Built for Business Analysts

30 copy-paste prompts

Every prompt below asks for a finished BRD, process map, or stakeholder report, not a paragraph of tips. Fill in the bracketed inputs and Claude hands back a document you can put in front of a sponsor today. Not "give me some advice".

In short: This page contains 30 copy-paste ready prompts, organized into 6 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

By Louis Corneloup ยท Founder, Techpresso
Last updated ยทHand-curated & tested by the AI Academy team

Requirements & BRDs

5 prompts

Business Requirements Document (BRD) Draft

1/30

You are a senior business analyst who has led requirements gathering for enterprise software projects. <context> A stakeholder has asked for a formal Business Requirements Document to kick off a new internal project. You only have rough notes from early conversations, not a polished brief. </context> <inputs> - Project name: [PLACEHOLDER] - Business problem or opportunity: [PLACEHOLDER] - Sponsor and key stakeholders: [PLACEHOLDER] - Rough notes on what the solution should do: [PLACEHOLDER] - Timeline or budget constraints, if any: [PLACEHOLDER] </inputs> <task> Turn the rough notes into a complete BRD: executive summary, business objectives, scope (in scope and out of scope), stakeholder list, high-level requirements grouped by theme, assumptions, and open questions. </task> <constraints> - Number every requirement (BR-1, BR-2, and so on) so it can be referenced later. - Flag any input that is missing or unclear as an open question instead of inventing detail. - Keep the executive summary under 150 words. </constraints> <format> A structured document with headers matching the sections above, ready to paste into Word or Confluence. </format>

Turns rough project notes into a numbered, stakeholder-ready Business Requirements Document.

๐Ÿ’ก

Pro tip: Paste your actual meeting notes into the rough notes field, even if messy. Claude organizes them better than a blank prompt would.

User Story Backlog With Acceptance Criteria

2/30

You are a senior business analyst who writes backlogs that development teams can estimate without asking clarifying questions. <context> A feature has been approved and needs to become a backlog of user stories before sprint planning. </context> <inputs> - Feature or epic name: [PLACEHOLDER] - Target users: [PLACEHOLDER] - Business goal: [PLACEHOLDER] - Known constraints (technical, compliance, timeline): [PLACEHOLDER] - Number of stories wanted: [PLACEHOLDER] </inputs> <task> Write a backlog of user stories in "As a / I want / so that" format, each with 3 to 5 acceptance criteria in Given/When/Then style. </task> <constraints> - Number the stories so they can be referenced in a sprint board. - Keep every acceptance criterion testable, no vague wording like "should work well". - Flag any story that depends on another story being done first. </constraints> <format> A numbered list of stories, each followed by its acceptance criteria as a sub-list. </format>

Converts an approved feature into a numbered backlog of testable user stories.

๐Ÿ’ก

Pro tip: List your real constraints, especially compliance ones. Claude will fold them into the acceptance criteria instead of leaving them implicit.

Functional Requirements Specification

3/30

You are a senior business analyst who writes functional specs developers can build directly from. <context> A high-level business requirement has been approved and needs to become a detailed functional specification for the engineering team. </context> <inputs> - System or module name: [PLACEHOLDER] - Business requirement it supports: [PLACEHOLDER] - User roles involved: [PLACEHOLDER] - Relevant business rules: [PLACEHOLDER] - Existing system it must integrate with: [PLACEHOLDER] </inputs> <task> Produce a functional requirements spec with numbered functional requirements (FR-1, FR-2, and so on), the business rules behind them, and a short data flow description. </task> <constraints> - Every FR must be independently testable. - Use consistent role names throughout, matching the inputs exactly. - Never combine two behaviors into a single FR. </constraints> <format> A table with columns ID, Requirement, Business Rule, Priority, followed by a short data flow paragraph. </format>

Breaks a business requirement into numbered, independently testable functional requirements.

๐Ÿ’ก

Pro tip: Name the exact integration system if there is one. It changes which FRs Claude flags as needing an interface contract.

Requirements Traceability Matrix

4/30

You are a senior business analyst supporting an audit or compliance review of a project's requirements. <context> Every requirement needs to be traceable back to a business objective and forward to a test case before sign-off. </context> <inputs> - List of requirements or a requirements doc excerpt: [PLACEHOLDER] - Business objectives they map to: [PLACEHOLDER] - Test case IDs, if known: [PLACEHOLDER] </inputs> <task> Build a traceability matrix linking each requirement to its originating business objective, the relevant design or spec section, and a test case reference. </task> <constraints> - The number of rows must match the number of input requirements exactly. - Mark a test case as "TBD" if none was provided, do not invent one. - No duplicate requirement IDs. </constraints> <format> A table with columns Requirement ID, Description, Business Objective, Design Reference, Test Case, Status. </format>

Links every requirement to its business objective, design reference, and test case in one auditable table.

๐Ÿ’ก

Pro tip: If you don't have test case IDs yet, still list requirement IDs cleanly. You can update the Test Case column once QA writes its cases.

MoSCoW Prioritization Matrix

5/30

You are a senior business analyst facilitating a scope prioritization session for a release with a fixed deadline. <context> Stakeholders disagree on what must ship in the first release and need an objective priority pass before the next planning meeting. </context> <inputs> - List of candidate requirements or features: [PLACEHOLDER] - Release deadline: [PLACEHOLDER] - Known technical constraints: [PLACEHOLDER] - Stakeholder priorities already voiced, if any: [PLACEHOLDER] </inputs> <task> Sort every requirement into Must have, Should have, Could have, or Won't have this release, with a one line rationale for each placement. </task> <constraints> - Every input requirement must appear exactly once. - Each rationale must reference a concrete reason: deadline, dependency, revenue, or risk. - Flag any placement that is a close call and needs a sponsor decision. </constraints> <format> Four headed lists (Must, Should, Could, Won't), each requirement followed by its rationale in parentheses. </format>

Sorts a contested feature list into Must/Should/Could/Won't with a defensible reason for every call.

๐Ÿ’ก

Pro tip: Include any priorities stakeholders have already stated out loud. Claude will use them to break ties instead of guessing.

Process Mapping & Workflow Analysis

5 prompts

Current-State Process Narrative

6/30

You are a senior business analyst documenting a process that has never been written down before. <context> A process runs informally today with no documentation, and it needs a written as-is walkthrough before it can be mapped visually. </context> <inputs> - Process name: [PLACEHOLDER] - Department(s) involved: [PLACEHOLDER] - Trigger event that starts the process: [PLACEHOLDER] - End event: [PLACEHOLDER] - Known steps and systems used (bullet notes): [PLACEHOLDER] </inputs> <task> Write a step by step as-is process narrative, numbered, noting the actor and system involved at each step. </task> <constraints> - One step equals one action, never bundle two actions together. - Note every handoff between roles or systems explicitly. - Flag any step where the input notes are ambiguous instead of guessing. </constraints> <format> A numbered step list, each step tagged with its actor and system in parentheses. </format>

Turns informal notes about an undocumented process into a numbered as-is walkthrough.

๐Ÿ’ก

Pro tip: List the systems by their real names (the actual CRM, spreadsheet, or tool). It makes the handoffs Claude flags much more useful.

Future-State Process Redesign

7/30

You are a senior business analyst leading a process improvement initiative. <context> The current-state process has documented bottlenecks and leadership wants a redesigned target state before the next planning cycle. </context> <inputs> - Current-state process (paste narrative or bullet steps): [PLACEHOLDER] - Known bottlenecks or pain points: [PLACEHOLDER] - Constraints on the redesign (budget, systems that must stay): [PLACEHOLDER] - Goal of the redesign: [PLACEHOLDER] </inputs> <task> Produce a future-state process, numbered, explicitly noting which steps were removed, added, or automated compared to the current state, and the expected benefit of each change. </task> <constraints> - Every removed, added, or changed step must state why. - Do not propose a new system or tool unless the constraints allow it. - Keep the future-state step count in the same order of magnitude as the current state. </constraints> <format> A two-column comparison table (Current Step vs Future Step) followed by a short "Why this changes" note per row. </format>

Redesigns a documented current-state process into a future state with a stated reason for every change.

๐Ÿ’ก

Pro tip: State which systems absolutely cannot change. Claude will design around that constraint instead of proposing a rip-and-replace.

RACI-Based Swimlane Description

8/30

You are a senior business analyst preparing input for a process diagramming tool. <context> A process needs to be described as a swimlane, showing exactly who does what, before it gets drawn in a diagramming tool. </context> <inputs> - Process name: [PLACEHOLDER] - Roles or departments in the lanes: [PLACEHOLDER] - Process steps in order: [PLACEHOLDER] - Decision points, if any: [PLACEHOLDER] </inputs> <task> Describe the process as a swimlane, listing each step under its lane (role), including decision diamonds and their branches. </task> <constraints> - Every step must belong to exactly one lane. - Mark decision points clearly with their Yes/No (or equivalent) branches. - Keep lane names consistent with the inputs throughout. </constraints> <format> A lane-by-lane outline (one heading per lane, steps listed under it in sequence), plus a short legend of decision points. </format>

Describes a multi-role process as a lane-by-lane outline ready to hand to a diagramming tool.

๐Ÿ’ก

Pro tip: List roles in the order you want the lanes to appear left to right. Claude keeps that order in the outline.

SIPOC Diagram

9/30

You are a senior business analyst scoping a process improvement project. <context> Before deep-diving into a process improvement project, the team needs a high-level SIPOC (Suppliers, Inputs, Process, Outputs, Customers) to align on scope. </context> <inputs> - Process name: [PLACEHOLDER] - Process owner: [PLACEHOLDER] - Rough description of what happens: [PLACEHOLDER] - Who supplies the inputs: [PLACEHOLDER] - Who receives the outputs: [PLACEHOLDER] </inputs> <task> Build a SIPOC table identifying suppliers, inputs, the 5 to 7 high-level process steps, outputs, and customers. </task> <constraints> - Process steps must stay high-level, no more than 7 steps. - Every supplier or customer must be a named role or system, not "various". - Do not add detail beyond what a SIPOC needs, save the deep detail for a full process map. </constraints> <format> A five-column table: Suppliers, Inputs, Process, Outputs, Customers. </format>

Builds a scoping-level SIPOC table before a full process mapping effort starts.

๐Ÿ’ก

Pro tip: If you genuinely don't know a supplier or customer yet, say so. Claude will mark it as an open question rather than filling in a guess.

Process Cycle-Time Analysis Table

10/30

You are a senior business analyst quantifying delay in a multi-step process. <context> A process has known steps with rough timing estimates, and leadership wants to see exactly where time is being lost. </context> <inputs> - Process steps in order: [PLACEHOLDER] - Estimated time per step (or rough range): [PLACEHOLDER] - Steps that involve waiting or a handoff: [PLACEHOLDER] - Target total cycle time, if known: [PLACEHOLDER] </inputs> <task> Build a cycle-time table showing time per step, cumulative time, and flag any step where wait time exceeds work time. </task> <constraints> - Use the actual numbers given, do not invent precise timings if the input only gives a range. - Highlight the top 3 most time-consuming steps. - Compare the resulting total to the target cycle time if one was given. </constraints> <format> A table with columns Step, Work Time, Wait Time, Cumulative Time, Flag, followed by a 3 bullet summary of the biggest opportunities. </format>

Turns rough step timings into a cycle-time table that shows exactly where delay is concentrated.

๐Ÿ’ก

Pro tip: Mark which steps involve a handoff between teams. Those are almost always where wait time hides.

Gap Analysis & Current-State Assessment

5 prompts

Current-State vs Future-State Gap Analysis Report

11/30

You are a senior business analyst assessing what stands between where a business is today and where it needs to be. <context> Leadership wants a clear picture of every gap between the current state and the target future state before committing budget to close them. </context> <inputs> - Current-state description: [PLACEHOLDER] - Future-state or target description: [PLACEHOLDER] - Business driver for the change: [PLACEHOLDER] - Timeframe: [PLACEHOLDER] </inputs> <task> Produce a gap analysis identifying each gap between the current and future state, its business impact, and a recommended closing action. </task> <constraints> - Group gaps by category: people, process, technology, data. - Rank each gap by impact: High, Medium, or Low. - Only list a gap that is actually implied by the two state descriptions given. </constraints> <format> A table with columns Category, Gap, Impact, Recommended Action. </format>

Identifies and ranks every gap between a documented current state and target future state.

๐Ÿ’ก

Pro tip: Write the future-state description with the same level of detail as the current-state one. Mismatched detail levels produce a lopsided gap list.

Project Risk Register

12/30

You are a senior business analyst supporting project planning. <context> A project is entering planning and needs an initial risk register before the kickoff meeting. </context> <inputs> - Project name: [PLACEHOLDER] - Scope summary: [PLACEHOLDER] - Known constraints: [PLACEHOLDER] - Stakeholders: [PLACEHOLDER] - Risks already raised informally, if any: [PLACEHOLDER] </inputs> <task> Identify likely risks across scope, schedule, resources, technical, and stakeholder categories, each with a likelihood, an impact, and a mitigation. </task> <constraints> - List a minimum of 8 risks. - No duplicate risks. - Every mitigation must be a concrete action, "monitor closely" alone is not acceptable. </constraints> <format> A table with columns Risk ID, Description, Category, Likelihood, Impact, Mitigation, Owner (mark Owner as [PLACEHOLDER] if not provided). </format>

Builds a starter risk register across scope, schedule, resource, technical, and stakeholder categories.

๐Ÿ’ก

Pro tip: Add any risk that's already been raised informally, even a one-liner. It anchors the rest of the register in reality instead of generic risks.

Cost-Benefit Analysis Table

13/30

You are a senior business analyst building the financial case for a proposed project. <context> A sponsor wants a rough cost-benefit case before approving a project, using whatever numbers are currently available. </context> <inputs> - Project name: [PLACEHOLDER] - Estimated one-time costs: [PLACEHOLDER] - Estimated ongoing costs: [PLACEHOLDER] - Expected benefits (quantified or qualitative): [PLACEHOLDER] - Time horizon for the analysis: [PLACEHOLDER] </inputs> <task> Build a cost-benefit table over the given time horizon and a short recommendation on whether the numbers justify the project. </task> <constraints> - Use only the numbers provided. - If benefits are only qualitative, say so explicitly instead of inventing a dollar figure. - State any assumption used to arrive at an estimate. </constraints> <format> A table with columns Year, Costs, Benefits, Net, followed by a short recommendation paragraph. </format>

Builds a year-by-year cost-benefit table and a plain recommendation from whatever numbers you already have.

๐Ÿ’ก

Pro tip: If benefits are qualitative only, say that up front. Claude will label the recommendation as directional rather than presenting a false ROI number.

Feasibility Study Summary

14/30

You are a senior business analyst preparing a go/no-go input for a proposed initiative. <context> An initiative needs a short feasibility study before it can be greenlit. </context> <inputs> - Proposed initiative: [PLACEHOLDER] - Technical constraints: [PLACEHOLDER] - Budget range: [PLACEHOLDER] - Timeline: [PLACEHOLDER] - Organizational appetite or context: [PLACEHOLDER] </inputs> <task> Assess feasibility across technical, operational, financial, and schedule dimensions, each rated feasible, feasible with conditions, or not feasible, with a one line reason. </task> <constraints> - Every rating needs a stated reason tied directly to the inputs. - If there's no input to judge a dimension, mark it "insufficient information" rather than guessing a rating. - Keep the overall recommendation to 2 to 3 sentences. </constraints> <format> A four-row table (Dimension, Rating, Reason) plus a short overall recommendation. </format>

Rates an initiative's feasibility across four dimensions with a reason for every rating.

๐Ÿ’ก

Pro tip: Be honest about budget range even if it's a wide guess. A wide range still lets Claude flag a financial feasibility concern accurately.

Root Cause Analysis (5 Whys and Fishbone)

15/30

You are a senior business analyst investigating a recurring operational problem. <context> A recurring problem keeps resurfacing and needs a documented root cause analysis before a fix is proposed. </context> <inputs> - Problem statement: [PLACEHOLDER] - When or how often it occurs: [PLACEHOLDER] - Who is affected: [PLACEHOLDER] - Any data already gathered on it: [PLACEHOLDER] </inputs> <task> Run a 5 Whys chain from the problem statement to a root cause, then organize likely contributing factors into a fishbone with categories People, Process, Technology, and Environment. </task> <constraints> - Each "why" must logically follow from the previous answer. - Fishbone categories should only include factors implied by the input, not invented dramatic causes. - If the data given is thin, say so rather than presenting speculation as fact. </constraints> <format> A numbered 5 Whys list, followed by a 4-category fishbone outline. </format>

Walks a recurring problem through a 5 Whys chain and organizes contributing factors into a fishbone.

๐Ÿ’ก

Pro tip: Include any data you already have, even partial. Claude will lean on it instead of the more generic causes.

These prompts give you the what. Tutorials give you the why.

Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. 300+ tutorials and growing.

Try AI Academy Free

Stakeholder Interviews & Elicitation

5 prompts

Stakeholder Interview Question Guide

16/30

You are a senior business analyst preparing for a requirements elicitation interview. <context> You're about to interview a business stakeholder to gather requirements and want a guide that gets past surface-level answers. </context> <inputs> - Stakeholder name and role: [PLACEHOLDER] - Project or topic: [PLACEHOLDER] - What you already know: [PLACEHOLDER] - What you specifically need to learn from this interview: [PLACEHOLDER] </inputs> <task> Write an interview guide with opening context-setting questions, core elicitation questions grouped by theme, and 2 to 3 probing follow-ups per theme. </task> <constraints> - No leading or yes/no-only questions in the core section. - Keep the guide answerable within 45 minutes. - End with a wrap-up question asking if anything important was missed. </constraints> <format> A sectioned guide (Opening, Theme 1, Theme 2, and so on, Wrap-up) with questions as a bullet list under each. </format>

Builds a themed interview guide with probing follow-ups so the session surfaces real requirements, not small talk.

๐Ÿ’ก

Pro tip: Be specific about what you already know. Claude will skip re-asking it and spend the interview time on genuine gaps.

Stakeholder Analysis and RACI Matrix

17/30

You are a senior business analyst mapping stakeholders before a project kickoff. <context> Multiple stakeholders have different levels of interest and influence, and the team needs them mapped before assigning responsibilities. </context> <inputs> - Project name: [PLACEHOLDER] - Stakeholders with rough role or title: [PLACEHOLDER] - Known interests or concerns per stakeholder, if any: [PLACEHOLDER] - Major project phases: [PLACEHOLDER] </inputs> <task> Classify each stakeholder by influence and interest, placing them on a power/interest grid, then assign a RACI role per major project phase. </task> <constraints> - Every stakeholder must get exactly one RACI role per phase. - No more than one "Accountable" per phase. - Flag any stakeholder with high influence but low current engagement. </constraints> <format> Table 1: Stakeholder, Influence, Interest, Grid Quadrant. Table 2: a Stakeholder by Phase RACI grid. </format>

Places every stakeholder on a power/interest grid and assigns a clean RACI role per project phase.

๐Ÿ’ก

Pro tip: Note any stakeholder who has been disengaged so far. Claude will flag that as a risk in the grid, not just log it as low interest.

Interview Notes Synthesized Into Requirements

18/30

You are a senior business analyst turning raw interview notes into structured requirements. <context> Several stakeholder interviews have been conducted, and the raw notes now need to become structured requirements before the next design step. </context> <inputs> - Raw interview notes or bullet summaries (one or more interviews): [PLACEHOLDER] - Project context: [PLACEHOLDER] - Requirement numbering convention, if any: [PLACEHOLDER] </inputs> <task> Extract discrete requirements from the notes, group them by theme, and flag any conflicting statements between interviewees. </task> <constraints> - Only extract what is actually stated or clearly implied in the notes. - Never merge two different stakeholders' asks into a single requirement. - List conflicts separately, quoting both source statements. </constraints> <format> A grouped requirement list by theme, followed by a "Conflicts to resolve" section. </format>

Extracts discrete, sourced requirements from messy interview notes and flags where stakeholders disagreed.

๐Ÿ’ก

Pro tip: Label which notes came from which stakeholder. Without that, Claude can't flag conflicts between them accurately.

Requirements Elicitation Survey

19/30

You are a senior business analyst gathering requirements input at scale. <context> There are too many stakeholders to interview individually, so a survey is needed to gather structured input that can actually inform prioritization decisions. </context> <inputs> - Project or topic: [PLACEHOLDER] - Target respondent group: [PLACEHOLDER] - Specific decisions the survey needs to inform: [PLACEHOLDER] - Desired length (number of questions): [PLACEHOLDER] </inputs> <task> Design a survey with a mix of closed rating or multiple choice questions and 1 to 2 open-ended questions that will produce data usable for prioritization. </task> <constraints> - Closed questions must use a consistent rating scale throughout. - Avoid double-barreled questions that ask two things at once. - Keep the total question count within the requested length. </constraints> <format> A numbered survey with the question type labeled next to each, for example "[1-5 scale]" or "[open text]". </format>

Designs a scale-appropriate requirements survey that actually produces data you can prioritize on.

๐Ÿ’ก

Pro tip: State the decision the survey needs to inform, not just the topic. It keeps Claude from writing generic satisfaction questions.

Meeting Minutes and Action Item Tracker

20/30

You are a senior business analyst who turns messy meeting notes into shareable minutes the same day. <context> A requirements or status meeting just happened and the notes need to become clean minutes with a trackable action list. </context> <inputs> - Meeting topic: [PLACEHOLDER] - Attendees: [PLACEHOLDER] - Rough notes or bullet points from the discussion: [PLACEHOLDER] - Decisions made, if any: [PLACEHOLDER] </inputs> <task> Produce clean meeting minutes with a decisions section and an action item tracker with owner and due date placeholders. </task> <constraints> - Keep "Decisions", "Discussion Summary", and "Action Items" as separate sections. - Every action item needs an owner field, marked [PLACEHOLDER] if not stated. - Discussion notes stay in bullet form, not prose. </constraints> <format> A three-section document: Decisions, Discussion Summary, and an Action Items table with Owner and Due Date columns. </format>

Converts rough meeting notes into clean minutes with a trackable, owner-assigned action list.

๐Ÿ’ก

Pro tip: Paste the raw notes even if they're just fragments. Claude will still separate decisions from discussion cleanly.

Reporting & Dashboards

5 prompts

Weekly Project Status Report

21/30

You are a senior business analyst who keeps a sponsor informed with a consistent weekly update. <context> A sponsor expects a weekly status update in the same format every time, and this week's raw notes need to become that report. </context> <inputs> - Project name: [PLACEHOLDER] - This week's progress notes: [PLACEHOLDER] - Blockers or risks: [PLACEHOLDER] - Upcoming milestones: [PLACEHOLDER] - Overall RAG status (Red/Amber/Green), if known: [PLACEHOLDER] </inputs> <task> Produce a weekly status report with overall status, progress this week, upcoming work, and risks or blockers, each section concise. </task> <constraints> - State the RAG status explicitly and justify it in one line. - Every blocker must name an owner or say [PLACEHOLDER]. - Keep the whole report to roughly one page. </constraints> <format> A status report with labeled sections: Status, Progress, Upcoming, Risks/Blockers. </format>

Turns this week's raw notes into a consistent, sponsor-ready RAG status report.

๐Ÿ’ก

Pro tip: If you're unsure of the RAG color, describe the situation plainly and let Claude propose one with its reasoning, then adjust if you disagree.

KPI Dashboard Specification

22/30

You are a senior business analyst writing the spec an analytics team will build a dashboard from. <context> The analytics team needs a written spec before building a dashboard, not just an informal request. </context> <inputs> - Business area the dashboard covers: [PLACEHOLDER] - Audience (executives, ops team, and so on): [PLACEHOLDER] - 3 to 6 KPIs wanted: [PLACEHOLDER] - Data sources available: [PLACEHOLDER] </inputs> <task> Specify each KPI (definition, formula, data source, refresh frequency, target or threshold) and recommend a layout grouping for the dashboard. </task> <constraints> - Every KPI needs an explicit formula, not just a name. - Flag any KPI whose data source is unclear from the input. - Group KPIs logically, no more than 4 per section. </constraints> <format> A table with columns KPI, Definition, Formula, Source, Frequency, Target, followed by a short layout recommendation. </format>

Specifies every KPI's formula and source before a dashboard gets built, so nothing is left to interpretation.

๐Ÿ’ก

Pro tip: List the data sources you actually have access to, not the ideal ones. Claude will flag KPIs that need a source you don't have yet.

Executive Summary Slide Content

23/30

You are a senior business analyst preparing a one-slide summary for executives who won't read the full report. <context> A completed analysis needs to become a single executive summary slide before the next leadership meeting. </context> <inputs> - Analysis topic: [PLACEHOLDER] - 3 to 5 key findings: [PLACEHOLDER] - Recommendation: [PLACEHOLDER] - Supporting numbers, if any: [PLACEHOLDER] </inputs> <task> Write slide-ready content: a headline takeaway, 3 to 4 bullet findings, and a clear recommendation with a next step. </task> <constraints> - The headline must be a specific claim, not a topic label. - Each bullet stays under 15 words. - The recommendation must state a concrete next action. </constraints> <format> Single slide content: Headline, Key Findings (bullets), Recommendation. </format>

Distills a full analysis into one slide's worth of headline, findings, and a concrete recommendation.

๐Ÿ’ก

Pro tip: Draft the headline last. Write the findings first, then ask Claude to state the headline as the single biggest implication of them.

Data Quality Assessment Report

24/30

You are a senior business analyst assessing whether a dataset is trustworthy enough to build on. <context> Before a project relies on a dataset, its quality needs a documented assessment, not just a gut check. </context> <inputs> - Dataset name or description: [PLACEHOLDER] - Known fields: [PLACEHOLDER] - Sample issues already noticed (nulls, duplicates, inconsistent formats): [PLACEHOLDER] - Intended use of the data: [PLACEHOLDER] </inputs> <task> Assess data quality across completeness, consistency, accuracy, and timeliness, using the noticed issues as evidence, and recommend remediation before the data is used. </task> <constraints> - Every quality dimension needs a rating (Good, Fair, or Poor) tied to a stated reason. - Do not claim an issue exists that wasn't mentioned or implied in the input. - Every remediation step must be concrete, not "clean the data". </constraints> <format> A table with columns Dimension, Rating, Evidence, Remediation, plus a one-line go/no-go recommendation for using the data as-is. </format>

Rates a dataset across four quality dimensions and gives a clear go/no-go call before it's used.

๐Ÿ’ก

Pro tip: List the specific issues you've already spotted, even one or two. Claude will use them as evidence instead of assuming the data is clean.

Business Case Document

25/30

You are a senior business analyst building the formal case for a proposed initiative. <context> An initiative needs a formal business case to secure funding or approval, comparing it honestly against the alternative of doing nothing. </context> <inputs> - Initiative name: [PLACEHOLDER] - Problem it solves: [PLACEHOLDER] - Options considered, including "do nothing": [PLACEHOLDER] - Estimated cost and benefit for each option: [PLACEHOLDER] - Recommended option: [PLACEHOLDER] </inputs> <task> Produce a business case with a problem statement, an options comparison, a recommendation, and expected outcomes with success measures. </task> <constraints> - "Do nothing" must be included as one option with its consequences stated plainly. - The recommendation must reference the comparison table, not restate the case from scratch. - Success measures must be specific and measurable. </constraints> <format> A document with sections Problem, Options Considered (comparison table), Recommendation, Success Measures. </format>

Builds a funding-ready business case that honestly weighs the cost of doing nothing.

๐Ÿ’ก

Pro tip: Always include a real cost estimate for doing nothing (lost revenue, ongoing manual effort). It's usually the number that gets a business case approved.

Most people use 10% of Claude. Tutorials unlock the rest.

AI Academy: 300+ hands-on tutorials on Claude, ChatGPT, Midjourney, and 50+ AI tools. New tutorials added every week.

Start Your Free Trial

Project & Change Documentation

5 prompts

Change Request and Impact Assessment

26/30

You are a senior business analyst assessing a scope change requested mid-project. <context> A scope change has been requested mid-project and needs a formal impact assessment before it can be approved. </context> <inputs> - Change requested: [PLACEHOLDER] - Reason for the change: [PLACEHOLDER] - Current baseline scope, schedule, and budget: [PLACEHOLDER] - Who requested it: [PLACEHOLDER] </inputs> <task> Document the change request and assess its impact on scope, schedule, budget, and risk, ending with a recommendation to approve, reject, or defer. </task> <constraints> - Address every dimension (scope, schedule, budget, risk) even if the impact is "none". - The recommendation must follow logically from the stated impacts, not contradict them. - Mark requester and date as [PLACEHOLDER] if not given. </constraints> <format> A change request form (Requester, Date, Description, Reason) followed by an Impact Assessment table and a Recommendation line. </format>

Documents a mid-project change and its impact across scope, schedule, budget, and risk in one form.

๐Ÿ’ก

Pro tip: State the current baseline numbers even roughly. Without a baseline, the impact assessment has nothing to measure against.

UAT Test Plan and Test Cases

27/30

You are a senior business analyst preparing user acceptance testing for a feature that's ready to test. <context> A feature or process is ready for user acceptance testing and needs a test plan and test cases tied directly to its requirements. </context> <inputs> - Feature or process being tested: [PLACEHOLDER] - Key requirements or acceptance criteria it must satisfy: [PLACEHOLDER] - User roles who will test it: [PLACEHOLDER] - Testing timeframe: [PLACEHOLDER] </inputs> <task> Produce a UAT test plan (scope, entry and exit criteria, roles) plus a set of numbered test cases with steps and expected results tied to the acceptance criteria. </task> <constraints> - Every acceptance criterion given must map to at least one test case. - Test case steps must be concrete and sequential. - Expected results must be specific and checkable, not "works correctly". </constraints> <format> A short test plan section, then a table of test cases with columns ID, Steps, Expected Result, Linked Requirement. </format>

Builds a UAT plan and test cases that map directly back to the original acceptance criteria.

๐Ÿ’ก

Pro tip: List your acceptance criteria exactly as written elsewhere in the project. Claude will map test cases one-to-one against them.

Training Material Outline for a New Process

28/30

You are a senior business analyst preparing end users for a process or system rollout. <context> A new process or system is rolling out and end users need training material before go-live. </context> <inputs> - Process or system name: [PLACEHOLDER] - Audience (role): [PLACEHOLDER] - Key tasks users need to be able to do after training: [PLACEHOLDER] - Format constraint (one-pager, slide deck, video script): [PLACEHOLDER] </inputs> <task> Outline training material covering each key task with a short walkthrough, ending with a quick-reference summary. </task> <constraints> - Cover every key task listed, nothing skipped. - Avoid jargon the target audience wouldn't already know. - The quick-reference summary must fit on roughly a single page. </constraints> <format> An outline with one section per key task, ending with a Quick Reference summary section. </format>

Outlines training material for a rollout, task by task, ending in a one-page quick reference.

๐Ÿ’ก

Pro tip: Name the audience's actual job title, not just 'users'. It changes how much background Claude explains versus assumes.

Post-Implementation Review Report

29/30

You are a senior business analyst closing out a project that has gone live. <context> A project has gone live and needs a formal look-back comparing what was promised to what actually happened before it's closed out. </context> <inputs> - Project name: [PLACEHOLDER] - Original objectives: [PLACEHOLDER] - What actually happened (outcomes, issues encountered): [PLACEHOLDER] - Lessons already noted by the team, if any: [PLACEHOLDER] </inputs> <task> Compare original objectives to actual outcomes, document lessons learned, and recommend follow-up actions. </task> <constraints> - Every original objective must be addressed as Met, Partially Met, or Not Met, with evidence. - Lessons learned must be actionable, "communicate better" alone is not acceptable, state what and how. - Follow-up actions need an owner placeholder. </constraints> <format> A table (Objective, Status, Evidence) followed by Lessons Learned and Follow-Up Actions sections. </format>

Compares a project's original objectives against what actually happened, with actionable lessons learned.

๐Ÿ’ก

Pro tip: Include issues honestly, even ones that reflect badly on the plan. The lessons learned section is only useful if it's specific.

Process Improvement Recommendation Memo

30/30

You are a senior business analyst making the case for a process change to a decision-maker. <context> Analysis is complete and the findings need to become a persuasive one-page recommendation memo for a decision-maker who is short on time. </context> <inputs> - Process analyzed: [PLACEHOLDER] - Key finding(s): [PLACEHOLDER] - Recommended improvement: [PLACEHOLDER] - Expected benefit: [PLACEHOLDER] - Resources or approval needed to implement: [PLACEHOLDER] </inputs> <task> Write a one-page memo making the case for the recommended improvement, addressed to the decision-maker. </task> <constraints> - Open with the recommendation, not with background. - Tie the expected benefit to a number or clear outcome wherever the input allows it. - Close with the specific approval or resource being requested. </constraints> <format> A memo (To/From/Date/Re header, then body) capped at roughly 400 words. </format>

Turns a finished analysis into a persuasive, lead-with-the-ask memo for a busy decision-maker.

๐Ÿ’ก

Pro tip: Put a number on the expected benefit if you have even a rough one. Memos that open with a recommendation and a number get read fastest.

Frequently Asked Questions

Copy a prompt, replace the bracketed inputs with your project's real details, and paste it into Claude. Ask follow-up questions to refine the draft before you share it.
No. Rough notes or a placeholder project name are enough. Claude produces a realistic first draft that you then edit with your actual numbers and decisions.
Yes, every prompt on this page is free to copy and use with any Claude plan.
Yes, treat them as starting templates. Add or remove constraints so the output matches your team's format and tone.
Any current Claude model handles them well. Longer documents like a full BRD or traceability matrix benefit from a model with a larger context window if you're pasting in a lot of source material.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.