Claude Prompt Library

30 Claude Prompts for Business Cases

30 copy-paste prompts

Paste in your numbers and constraints and Claude returns cost-benefit tables, ranked options, payback math, and an executive summary you can bring straight into a decision meeting.

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.

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

Cost-Benefit Analysis

5 prompts

Build a cost-benefit table from raw line items

1/30

✨ What it does

Claude turns [PROJECT NAME] costs, including [LIST OF ONE TIME COSTS], into a year-by-year cost-benefit table with net figures. Check the recurring lines against last year's actuals in your books, then drop the table into the case.

You are a senior financial analyst who builds business cases for operations leadership. <context> I am proposing a project and I have a rough list of costs and expected benefits but no structured comparison yet. </context> <inputs> - Project name: [PROJECT NAME] - One time costs: [LIST OF ONE TIME COSTS] - Recurring annual costs: [LIST OF RECURRING COSTS] - Expected benefits, quantified where possible: [LIST OF BENEFITS] - Time horizon for the analysis: [NUMBER OF YEARS] </inputs> <task> Build a year by year cost-benefit table that separates one time costs, recurring costs, and benefits, then calculate net benefit per year and cumulative net benefit across the full horizon. </task> <constraints> Do not invent numbers I have not given you. If a benefit cannot be quantified, list it separately as a qualitative benefit rather than forcing a dollar figure. Keep commentary under 150 words, the table is the deliverable. </constraints> <format> Return a markdown table with columns for year, one time cost, recurring cost, benefit, and net, followed by a short paragraph stating the break even year. </format>

💡

Pro tip: Paste your numbers as plain text lines rather than a formatted table, Claude parses loose lists fine and you save the formatting step.

Compare in-house build versus buying a vendor solution

2/30

✨ What it does

Claude produces a fair, assumption-transparent cost comparison of building [TOOL PURPOSE] internally versus buying a vendor tool. Challenge the hour estimate with your engineers, then take the comparison to the decision meeting.

You are a procurement and finance advisor who evaluates build versus buy decisions. <context> My team is deciding whether to build a tool internally or license one from a vendor and I need a fair cost comparison. </context> <inputs> - What the tool needs to do: [TOOL PURPOSE] - Internal build estimate, hours and rates: [BUILD COST INPUTS] - Vendor quote and contract terms: [VENDOR QUOTE DETAILS] - Ongoing maintenance assumption for the internal option: [MAINTENANCE ASSUMPTION] - Timeframe to compare over: [NUMBER OF YEARS] </inputs> <task> Produce a side by side cost comparison of building versus buying across the given timeframe, including hidden costs like maintenance, opportunity cost of engineering time, and vendor lock-in risk. </task> <constraints> Be explicit about assumptions you had to make and flag any place the two options are not truly comparable. Avoid vague statements like the vendor option is more scalable, state why with a number or a specific mechanism. </constraints> <format> Return a two column comparison table followed by a three sentence recommendation. </format>

💡

Pro tip: List the vendor's contract terms including any auto-renewal or price escalation clauses, Claude will flag them as hidden costs if you include them.

Quantify the cost of doing nothing

3/30

✨ What it does

Claude calculates the annualized cost of leaving [PROCESS DESCRIPTION] unchanged, using [LIST OF PAIN POINTS] as the pain you already know. Put that do-nothing number at the top of the case, then present it before the project cost.

You are a business analyst who specializes in status quo cost modeling. <context> Leadership keeps asking why we need to change a process, and I need to show the real cost of leaving it as is. </context> <inputs> - Current process: [PROCESS DESCRIPTION] - Known pain points: [LIST OF PAIN POINTS] - Time or money currently lost, even roughly: [ROUGH LOSS ESTIMATE] - Frequency of the problem: [FREQUENCY] </inputs> <task> Calculate the annualized cost of continuing the current process unchanged, including labor time, error correction, and any lost revenue or customer impact, and present it as a do-nothing baseline to compare future options against. </task> <constraints> Show your math step by step so a skeptical CFO can follow it. Do not round in ways that inflate the number, if anything round down and note it as a conservative estimate. </constraints> <format> Return a short calculation walkthrough followed by a single bolded annualized cost figure. </format>

💡

Pro tip: Use this output as the baseline row in your cost-benefit table so every option is compared against a real number, not a vague sense that things are bad.

Stress test benefit assumptions with a skeptic's review

4/30

✨ What it does

Claude runs a skeptical review of [BENEFIT ASSUMPTIONS] and returns conservative alternatives you can defend. Replace the optimistic figures before the meeting, then present only the numbers you can still stand behind.

You are a finance director known for pushing back on optimistic project proposals. <context> I have drafted the benefit assumptions for a business case and I want them challenged before I present them. </context> <inputs> - Proposed benefits and how they were estimated: [BENEFIT ASSUMPTIONS] - Who provided the estimates: [SOURCE OF ESTIMATES] - Project timeline: [TIMELINE] </inputs> <task> Review each benefit assumption and identify which ones are likely overstated, which depend on conditions that may not hold, and which need a named owner to be credible. Suggest a more conservative version of each optimistic figure. </task> <constraints> Be direct and specific, do not soften criticism with hedging language. Reference the exact assumption you are challenging so I can trace it back to my draft. </constraints> <format> Return a numbered list, one entry per assumption, each with the original figure, the concern, and a suggested conservative figure. </format>

💡

Pro tip: Run this before you show the business case to anyone else, it catches the exact objections a tough stakeholder would raise.

Translate technical cost drivers into plain business language

5/30

✨ What it does

Claude translates [TECHNICAL LINE ITEMS] into plain sentences a non-technical approver can understand. Read them to someone outside your engineering team, then swap any leftover jargon before the packet goes out.

You are a technical program manager who translates engineering cost drivers for non technical executives. <context> My cost estimate includes technical line items that a business audience will not understand without translation. </context> <inputs> - Technical cost line items: [TECHNICAL LINE ITEMS] - Audience for this document: [AUDIENCE DESCRIPTION] - Total budget being requested: [TOTAL BUDGET] </inputs> <task> Rewrite each technical cost line item in plain business language that explains what it buys and why it matters to the outcome, without losing the underlying number. </task> <constraints> Avoid jargon and acronyms unless you define them on first use. Each rewritten line should be one sentence. Keep the original dollar figure attached to each line. </constraints> <format> Return a table with three columns, original technical line, dollar amount, and plain language explanation. </format>

💡

Pro tip: Feed it your engineering estimate spreadsheet lines verbatim, it works better with the exact wording than with a paraphrase.

Options Analysis

5 prompts

Generate a weighted scoring matrix for competing options

6/30

✨ What it does

Claude scores [LIST OF OPTIONS] against weighted criteria and returns a ranked, defensible recommendation. Adjust the weights if they do not match how your leadership actually decides, then bring the matrix to the review.

You are a decision analysis consultant who builds weighted scoring models for capital decisions. <context> I am choosing between several options for a project and need an objective way to compare them beyond gut feel. </context> <inputs> - Options being considered: [LIST OF OPTIONS] - Decision criteria that matter: [LIST OF CRITERIA] - Relative importance of each criterion, if known: [WEIGHTING NOTES] - Any known constraints, like budget cap or deadline: [CONSTRAINTS] </inputs> <task> Build a weighted scoring matrix scoring each option against each criterion on a 1 to 5 scale, apply the weights, and calculate a total score per option with a ranked recommendation. </task> <constraints> Explain the reasoning behind each score in one short phrase so the scoring is not a black box. If two options score within five percent of each other, say so explicitly rather than declaring a false winner. </constraints> <format> Return a markdown table with options as rows and criteria as columns, weighted totals in the final column, followed by a two sentence recommendation. </format>

💡

Pro tip: If you are not sure how to weight the criteria, ask Claude to propose weights first based on your stated priorities, then adjust before it scores the options.

Draft the classic three option business case structure

7/30

✨ What it does

Claude structures the decision around [PROBLEM STATEMENT] into do-minimum, do-some, and do-more options instead of a single ask. Fill each option with a real cost, then present all three so your reviewers can choose.

You are a management consultant who writes business cases using the do minimum, do some, do more framework. <context> I need to present a decision to leadership using three clearly differentiated options rather than a single recommendation. </context> <inputs> - Problem being solved: [PROBLEM STATEMENT] - Minimum viable action: [DO MINIMUM DESCRIPTION] - Moderate investment option: [DO SOME DESCRIPTION] - Ambitious option: [DO MORE DESCRIPTION] - Budget range available: [BUDGET RANGE] </inputs> <task> Structure the three options into a do minimum, do some, do more comparison, covering cost, expected outcome, risk level, and time to value for each, and recommend one with reasoning. </task> <constraints> Make sure the do minimum option is a genuine option, not a strawman designed to fail. Each option needs at least one real tradeoff, not just more money for more benefit. </constraints> <format> Return three labeled sections, one per option, each with the same four subheadings, followed by a recommendation paragraph. </format>

💡

Pro tip: Make the do minimum option genuinely viable in your inputs, executives spot a strawman low option immediately and it undermines the whole case.

Map second order consequences of each option

8/30

✨ What it does

Claude maps the downstream consequences of each shortlisted option beyond the immediate result. Mark which later effects you can actually live with, then add those tradeoffs to the case.

You are a strategy advisor who thinks in second order effects before a decision is finalized. <context> I have narrowed a decision to a short list of options and want to understand what happens after the immediate outcome, not just the first result. </context> <inputs> - Options under consideration: [SHORT LIST OF OPTIONS] - Immediate expected outcome of each: [EXPECTED IMMEDIATE OUTCOMES] - Teams or departments affected: [AFFECTED TEAMS] - Time frame to think ahead: [TIME FRAME] </inputs> <task> For each option, map the second order consequences, meaning what happens as a result of the first outcome, including effects on other teams, customer behavior, or future flexibility. </task> <constraints> Distinguish clearly between likely consequences and speculative ones by labeling each. Do not pad the list with consequences that would apply equally to every option, focus on what differentiates them. </constraints> <format> Return one section per option with a short bulleted list of second order consequences, each bullet labeled likely or speculative. </format>

💡

Pro tip: Ask a follow up question naming the option you are leaning toward and have Claude argue the strongest case against it before you commit.

Turn a single proposal into a real multi-option comparison

9/30

✨ What it does

Claude invents two credible alternative approaches so your single proposal can face a real comparison. Stress-test your favorite against those alternatives, then keep the comparison in the packet before review.

You are a business case reviewer whose job is to force single proposals into fair option comparisons. <context> I have only one proposal drafted and I know reviewers will ask what else was considered, so I need credible alternatives to compare it against. </context> <inputs> - The single proposal as currently written: [PROPOSAL SUMMARY] - Goal the proposal is meant to achieve: [GOAL] - Resources realistically available: [AVAILABLE RESOURCES] </inputs> <task> Generate two plausible alternative approaches to the same goal that were not in the original proposal, then compare all three, including the original, on cost, speed, and risk. </task> <constraints> The alternatives must be genuinely different approaches, not the same idea with a different budget. Be honest if the original proposal is not actually the best option once compared fairly. </constraints> <format> Return a short description of each alternative followed by a three column comparison table covering cost, speed, and risk. </format>

💡

Pro tip: Use this before you submit anything for approval, a proposal that survives an honest three way comparison is far more likely to survive committee questions.

Assess vendor options against a weighted requirements list

10/30

✨ What it does

Claude scores [VENDOR NAMES] against your must-have requirements and flags disqualifying gaps before price is compared. Drop any vendor that fails a must-have, then look at price only after those gaps are gone.

You are a vendor selection specialist who evaluates supplier proposals against business requirements. <context> I am comparing multiple vendor proposals for the same need and want a structured, defensible evaluation. </context> <inputs> - Vendors being compared: [VENDOR NAMES] - Key requirements, must-haves and nice-to-haves: [REQUIREMENTS LIST] - Pricing from each vendor: [PRICING BY VENDOR] - Any red flags noticed so far: [KNOWN RED FLAGS] </inputs> <task> Evaluate each vendor against the requirements list, flag which must-haves each one fails to meet, and produce a ranked recommendation that accounts for both price and requirement fit. </task> <constraints> A vendor that fails a must-have requirement cannot rank first regardless of price, state this rule explicitly in your reasoning. Keep pricing comparisons apples to apples, note any differences in contract length or included support. </constraints> <format> Return a requirements checklist table by vendor, followed by a ranked list with one sentence of reasoning per vendor. </format>

💡

Pro tip: List your must-haves and nice-to-haves in two clearly separate groups in the input, Claude enforces the must-have rule far more reliably that way.

Executive Summary

5 prompts

Compress a full business case into one page

11/30

✨ What it does

Claude compresses [BUSINESS CASE CONTENT] into one page an executive can read in under a minute. Cut anything that is not the ask or the number, then send that page the morning of your meeting.

You are an executive communications specialist who writes one page summaries for C-level readers. <context> I have a full business case document and the executive team only has time to read one page before the meeting. </context> <inputs> - Full business case content or key points: [BUSINESS CASE CONTENT] - Ask being made of leadership: [SPECIFIC ASK] - Decision deadline: [DEADLINE] </inputs> <task> Compress the business case into a one page executive summary covering the problem, the recommended option, the cost, the expected benefit, and the ask, in that order. </task> <constraints> Keep total length under 300 words. Do not use subheadings for every sentence, write in flowing short paragraphs. Lead with the recommendation, not the background. </constraints> <format> Return five short paragraphs, one for each required element, with no more than two sentences per paragraph. </format>

💡

Pro tip: Paste your entire draft business case as the input, do not pre-summarize it yourself, Claude compresses more accurately from the full source.

Write the summary before the analysis is finished

12/30

✨ What it does

Claude drafts an honest, caveat-labeled executive summary from your working hypothesis before the analysis is done. Use it to check you are building toward the right conclusion, then update it when the numbers land.

You are a strategy manager who writes executive summaries first, before the supporting analysis, to clarify thinking. <context> I am still building the analysis but want a draft executive summary now to make sure I am building toward the right conclusion. </context> <inputs> - Working hypothesis for the recommendation: [WORKING HYPOTHESIS] - Rough cost estimate so far: [ROUGH COST] - Rough benefit estimate so far: [ROUGH BENEFIT] - Biggest open question still unresolved: [OPEN QUESTION] </inputs> <task> Draft an executive summary based on the working hypothesis, clearly marking any figure that is still preliminary, and note the open question as a caveat rather than hiding it. </task> <constraints> Be explicit about what is confirmed versus assumed. Do not present a preliminary number with false precision, round it and label it as an estimate. </constraints> <format> Return a short summary of four sentences followed by a single caveat line starting with Note. </format>

💡

Pro tip: Use this draft as a north star while you finish the detailed analysis, then run it again once real numbers replace the estimates.

Rewrite a summary for a specific executive's known priorities

13/30

✨ What it does

Claude reframes [DRAFT SUMMARY TEXT] for a specific executive's known priorities without changing the underlying numbers. Send that tailored version to your reader, then keep the original numbers identical.

You are a chief of staff who tailors executive communications to the specific priorities of the person reading them. <context> I am sending the same business case to two different executives and I know they care about different things. </context> <inputs> - Draft executive summary: [DRAFT SUMMARY TEXT] - Executive receiving this version: [EXECUTIVE NAME AND ROLE] - Known priorities of this executive: [KNOWN PRIORITIES] </inputs> <task> Rewrite the draft summary to foreground the elements most relevant to this executive's known priorities, without changing any of the underlying facts or figures. </task> <constraints> Do not alter numbers, dates, or the recommendation itself, only reorder and reframe emphasis. Keep the length within ten percent of the original draft. </constraints> <format> Return the rewritten summary only, no explanation of what changed. </format>

💡

Pro tip: Name the exact priority, for example margin protection or headcount discipline, rather than a vague trait like detail oriented, the rewrite is sharper.

Write a summary that survives a five minute meeting

14/30

✨ What it does

Claude turns the case facts into spoken talking points for a five-minute slot, plus a rehearsed answer to the expected objection. Practice the objection answer once, then use the points in your meeting instead of reading the document.

You are an operations director who prepares executives to make a fast decision in a short meeting slot. <context> I have a five minute slot on a leadership agenda to get approval on this business case and need the summary to work as talking points, not just a document. </context> <inputs> - Business case core facts: [CORE FACTS] - Decision needed from the room: [DECISION NEEDED] - Likely pushback question: [LIKELY PUSHBACK] </inputs> <task> Write a talking points version of the executive summary structured so it can be delivered verbally in under two minutes, followed by a prepared response to the likely pushback question. </task> <constraints> Use short declarative sentences suitable for speaking aloud, not dense written prose. Do not exceed six talking points total. </constraints> <format> Return a bulleted list of talking points followed by a separate short paragraph labeled If asked, addressing the pushback. </format>

💡

Pro tip: Rehearse the talking points out loud once before the meeting, the prompt writes for speech but you still need to hear where it runs long.

Extract the single sentence that decides the case

15/30

✨ What it does

Claude extracts or writes the single sentence from [FULL BUSINESS CASE TEXT] that would actually convince the decision maker. If that sentence feels weak, tighten the case around it before you circulate the long version.

You are a board advisor who reduces complex proposals to their single decisive sentence. <context> I have written a long business case and want to know if it actually has one clear decisive argument or if it is just a pile of information. </context> <inputs> - Full business case text: [FULL BUSINESS CASE TEXT] - Decision this case is meant to drive: [DECISION NEEDED] - Reader who will judge whether it is convincing: [READER NAME AND ROLE] </inputs> <task> Read the business case and extract the single sentence that most decisively justifies the recommendation for [DECISION NEEDED]. If no single sentence in the document does this job well, write one that would, using only facts already present in the text. </task> <constraints> Do not introduce any fact, number, or claim that is not already in the source text. If the case is weak, say so plainly rather than manufacturing false confidence. </constraints> <format> Return the decisive sentence in quotes, followed by one sentence explaining why it is the strongest argument in the document. </format>

💡

Pro tip: If the returned sentence feels weak, that is a signal your business case lacks a real argument, not a prompting problem, go fix the analysis first.

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

Payback and ROI

5 prompts

Calculate simple payback period with a full worked walkthrough

16/30

✨ What it does

Claude walks through a year-by-year payback calculation from [INITIAL INVESTMENT], including partial years, and states the payback point. Check the cash-flow timing with your finance partner, then put the worked math in the appendix.

You are a corporate finance analyst who prepares payback period calculations for capital requests. <context> I need to show how long it will take for a project to pay back its initial investment and I want the math shown, not just a final number. </context> <inputs> - Initial investment amount: [INITIAL INVESTMENT] - Expected net cash inflow per year: [ANNUAL NET INFLOW] - Whether cash inflow is expected to grow over time: [GROWTH ASSUMPTION] </inputs> <task> Calculate the simple payback period, showing cumulative cash flow year by year until it turns positive, and state the exact payback point including partial years. </task> <constraints> Show the running cumulative total for every year, not just the final answer. If inflows are uneven or growing, do not average them, calculate year by year. </constraints> <format> Return a table with year and cumulative cash flow columns, followed by one bolded sentence stating the payback period. </format>

💡

Pro tip: If your cash inflows change over time, list each year separately in the input rather than giving one average figure, the calculation is only as accurate as that input.

Calculate ROI and compare it to a hurdle rate

17/30

✨ What it does

Claude calculates annualized ROI on [TOTAL INVESTMENT] against your hurdle rate and gives a pass, fail, or marginal verdict. Confirm the hurdle rate with finance, then only send the case forward if you accept that verdict.

You are a finance business partner who evaluates whether proposed projects clear the company's investment hurdle rate. <context> I need to know if this project's return on investment clears our required threshold before it goes to the investment committee. </context> <inputs> - Total investment: [TOTAL INVESTMENT] - Total expected return over the period: [TOTAL EXPECTED RETURN] - Time period the return is measured over: [TIME PERIOD] - Company hurdle rate: [HURDLE RATE] </inputs> <task> Calculate the return on investment as a percentage, annualize it if the period is not one year, and state clearly whether it clears the given hurdle rate. </task> <constraints> Show the formula used, not just the result. If the ROI is close to the hurdle rate, within five percentage points, flag it as marginal rather than a clean pass or fail. </constraints> <format> Return the formula with your numbers substituted in, the resulting percentage, and a one sentence verdict against the hurdle rate. </format>

💡

Pro tip: Always specify whether your expected return figure is gross or net of ongoing costs, that single detail changes the ROI result significantly.

Build a sensitivity analysis for optimistic, likely, and pessimistic cases

18/30

✨ What it does

Claude builds a three-scenario sensitivity table showing how payback and ROI move when one key variable changes. Pick the variable your leadership will challenge, then present the range instead of a single point.

You are a risk-aware financial planner who builds scenario ranges instead of single point estimates. <context> My business case currently has one set of numbers and I want to show how the outcome changes under different assumptions before I present it. </context> <inputs> - Base case assumptions: [BASE CASE ASSUMPTIONS] - Key variable most likely to swing the outcome: [KEY VARIABLE] - Realistic range for that variable, low to high: [VARIABLE RANGE] </inputs> <task> Build optimistic, likely, and pessimistic scenarios by varying the key variable across its realistic range, and recalculate the payback period and ROI under each scenario. </task> <constraints> Only vary the one key variable stated, hold all other assumptions constant across scenarios so the comparison is clean. Label which scenario is the base case. </constraints> <format> Return a three column table, one column per scenario, with rows for the variable value, payback period, and ROI. </format>

💡

Pro tip: Pick the single variable your stakeholders are most likely to challenge as unrealistic, that is the one worth stress testing, not the one you feel most confident about.

Explain net present value in plain terms for a non-finance audience

19/30

✨ What it does

Claude calculates NPV from [CASH FLOWS BY YEAR] with the discounting shown, then explains the result in plain language. Rehearse the plain explanation, then use it when your non-finance approver asks what the number means.

You are a finance educator who explains capital budgeting concepts to non finance managers making a real decision. <context> My business case needs a net present value calculation but the approver is not a finance person and will ask what it means. </context> <inputs> - Cash flows by year: [CASH FLOWS BY YEAR] - Discount rate to use: [DISCOUNT RATE] - Approver's background: [APPROVER BACKGROUND] </inputs> <task> Calculate the net present value using the given discount rate and cash flows, then explain in plain language what the resulting number means for this specific decision, avoiding finance jargon. </task> <constraints> Show the discounted value for each year before summing them. The plain language explanation should be no longer than three sentences and must avoid the phrase time value of money without explaining it first. </constraints> <format> Return a table of year, cash flow, and discounted value, followed by the total NPV, followed by the plain language explanation paragraph. </format>

💡

Pro tip: If you are unsure what discount rate to use, ask Claude to explain what a reasonable range would be for your industry before you commit to one number.

Reconcile conflicting ROI numbers from two different teams

20/30

✨ What it does

Claude pinpoints which assumptions make two teams' ROI numbers diverge, instead of averaging the disagreement away. Sit both of your teams on those assumption gaps, then present one reconciled figure.

You are an FP&A lead who reconciles conflicting financial estimates between teams before they reach leadership. <context> Two teams have produced different ROI estimates for the same proposal and I need to understand why before I present either one. </context> <inputs> - Team A's ROI estimate and how it was calculated: [TEAM A ESTIMATE AND METHOD] - Team B's ROI estimate and how it was calculated: [TEAM B ESTIMATE AND METHOD] - Proposal name both teams are estimating: [PROPOSAL NAME] - Deadline for reconciling the numbers: [RECONCILIATION DEADLINE] </inputs> <task> Compare the two methodologies line by line, identify exactly where the assumptions diverge, and recommend which assumptions are more defensible with reasoning. </task> <constraints> Do not simply average the two numbers, that hides the real disagreement. Name the specific assumption causing the gap, not a vague statement that the methods differ. </constraints> <format> Return a table listing each assumption, Team A's value, Team B's value, and which one you recommend using with a one line reason. </format>

💡

Pro tip: Ask both teams for their raw assumptions, not just the final percentage, the reconciliation only works if you can see the actual inputs behind each number.

Risk Assessment

5 prompts

Build a risk register with likelihood and impact scoring

21/30

✨ What it does

Claude turns [LIST OF RISKS] into a scored register with a specific mitigation for each item. Assign an owner to every high score, then add the register as your risk section.

You are a project risk manager who builds structured risk registers for approval documents. <context> My business case needs a formal risk section and I have a rough list of things that could go wrong but nothing scored or prioritized. </context> <inputs> - Rough list of possible risks: [LIST OF RISKS] - Project duration: [PROJECT DURATION] - Total budget at stake: [TOTAL BUDGET] </inputs> <task> Build a risk register scoring each risk on likelihood and impact using a low, medium, high scale, calculate an overall priority, and suggest one concrete mitigation per risk. </task> <constraints> Every mitigation must be a specific action, not a vague statement like monitor closely. Do not inflate every risk to high, use the full range of the scale honestly. </constraints> <format> Return a table with columns for risk, likelihood, impact, priority, and mitigation. </format>

💡

Pro tip: Push back if every risk comes back scored as high, ask Claude to re-rank them relative to each other so the register actually helps you prioritize.

Identify the risks nobody has mentioned yet

22/30

✨ What it does

Claude surfaces overlooked risk categories your team is likely to miss because you wrote the case together. Add the ones you cannot honestly dismiss, then keep the known risks as they are.

You are a devil's advocate reviewer whose job is to surface overlooked risks before a decision is finalized. <context> My team has reviewed this business case several times but I suspect we are all missing the same blind spots because we wrote it together. </context> <inputs> - Business case summary: [BUSINESS CASE SUMMARY] - Risks already identified: [KNOWN RISKS] - Team's general background or bias, if relevant: [TEAM BACKGROUND] </inputs> <task> Identify risks that are not in the known risks list, focusing on categories the team is likely to have overlooked given their stated background, such as regulatory, competitive, operational, or reputational angles. </task> <constraints> Do not repeat any risk already in the known list. Each new risk needs a one sentence explanation of why it is plausible given the specific details of this case, not a generic risk template entry. </constraints> <format> Return a bulleted list of new risks, each with a category label and a one sentence justification. </format>

💡

Pro tip: Be honest about your team's background and bias in the input, for example all engineers, that context is what lets Claude target the actual blind spot.

Draft a contingency plan for the top risk

23/30

✨ What it does

Claude writes a concrete, ordered contingency plan for [SPECIFIC RISK], with a measurable trigger. Confirm the trigger with your operations lead, then attach the plan before leadership asks for a fallback.

You are an operations risk planner who writes concrete contingency plans, not general risk statements. <context> One risk in my business case is severe enough that leadership will want to see a specific fallback plan before approving the project. </context> <inputs> - The specific risk: [SPECIFIC RISK] - Trigger condition that would activate the contingency: [TRIGGER CONDITION] - Resources available if the contingency is needed: [AVAILABLE RESOURCES] </inputs> <task> Draft a contingency plan that specifies the trigger condition, the immediate response steps, who would need to be involved, and the cost or time impact of activating the plan. </task> <constraints> The trigger condition must be measurable, not a vague feeling that things are going badly. List response steps in the actual order they would happen, not as a general list. </constraints> <format> Return four labeled sections: trigger, response steps as a numbered list, people involved, and cost or time impact. </format>

💡

Pro tip: Define the trigger condition as a specific number or date wherever possible, a measurable trigger is what makes a contingency plan credible instead of decorative.

Assess dependency risk on other teams or projects

24/30

✨ What it does

Claude ranks cross-team dependencies by delay risk and estimates the timeline hit if each one slips. Name the owner of each dependency, then show your leadership the slip impact in the case.

You are a program manager who specializes in mapping cross-team dependency risk. <context> My project depends on other teams or projects delivering on time, and I need to show leadership I have thought through what happens if they do not. </context> <inputs> - Other teams or projects this depends on: [DEPENDENCIES LIST] - Current confidence level in each dependency: [CONFIDENCE LEVEL] - Impact on this project if a dependency slips: [SLIP IMPACT] </inputs> <task> Assess each dependency for risk of delay, estimate the knock-on effect on this project's timeline if it slips, and suggest one way to reduce reliance on the riskiest dependency. </task> <constraints> Be specific about how many weeks or months a slip would cost this project, not just a qualitative statement of delay. Rank the dependencies from most to least risky. </constraints> <format> Return a ranked table with columns for dependency, confidence level, estimated delay impact, and mitigation suggestion. </format>

💡

Pro tip: Give an honest confidence level for each dependency rather than an optimistic one, the ranking is only useful if the inputs reflect real uncertainty.

Write the risk section the way a skeptical board member would want it

25/30

✨ What it does

Claude rewrites the risk section so every risk has a named owner and a specific consequence a board can question. Fill any missing owner before your board pack goes out, then keep that version in the appendix.

You are a board level risk officer who reviews business cases for material omissions before they reach the board. <context> This business case is going in front of a board that has rejected proposals before for glossing over risk, and I need the risk section to hold up under tough questioning. </context> <inputs> - Draft risk section: [DRAFT RISK SECTION] - Size of the investment: [INVESTMENT SIZE] - Any prior board concerns on similar proposals: [PRIOR BOARD CONCERNS] </inputs> <task> Review the draft risk section for anything that reads as understated or vague, and rewrite it so each risk statement includes a specific consequence and a named mitigation owner. </task> <constraints> Do not remove any risk that is already disclosed, only strengthen the language and specificity. Flag if the investment size seems disproportionate to the risk mitigation budget described. </constraints> <format> Return the rewritten risk section as a numbered list, followed by a short paragraph flagging any remaining gap. </format>

💡

Pro tip: Include any specific prior board pushback you remember from earlier proposals, Claude will target the rewrite at exactly the pattern that got rejected before.

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

Stakeholder Communication

5 prompts

Anticipate and pre-answer stakeholder objections

26/30

✨ What it does

Claude maps each attending stakeholder to their likely objection and drafts a fact-grounded response. Rehearse the two hardest answers, then walk into your meeting with those replies ready.

You are a change management consultant who prepares proposal owners for stakeholder pushback before it happens. <context> I am about to present this business case to a group of stakeholders who have blocked similar proposals in the past, and I want to walk in prepared. </context> <inputs> - Business case summary: [BUSINESS CASE SUMMARY] - Stakeholders attending and their known concerns: [STAKEHOLDERS AND CONCERNS] - Past objections to similar proposals: [PAST OBJECTIONS] </inputs> <task> List the objections each named stakeholder is most likely to raise based on their known concerns and past objections, then draft a short, direct response to each one grounded in the business case facts. </task> <constraints> Match each objection to the specific stakeholder likely to raise it, do not produce a generic objection list. Responses must reference actual figures from the case, not reassurance without substance. </constraints> <format> Return one entry per stakeholder with their likely objection and your prepared response underneath it. </format>

💡

Pro tip: Name the actual people attending, not just their roles, past pushback is usually personal and specific, not generic to the title.

Draft a one paragraph pitch for a hallway conversation

27/30

✨ What it does

Claude writes a thirty-second hallway pitch of the case that ends with a question inviting a real reaction. Try it on one ally first, then use the same lines in your next casual conversation.

You are a startup operator skilled at pitching ideas informally before they hit a formal process. <context> Before I submit this business case formally, I want to test it in casual conversations with a few key people, and I need a short version that works spoken. </context> <inputs> - Business case core idea: [CORE IDEA] - Ask being made: [THE ASK] - Person I am likely to run into first: [FIRST PERSON] </inputs> <task> Write a single paragraph, conversational pitch of the business case that could be said aloud in under thirty seconds, ending in a clear question that invites a reaction. </task> <constraints> Write it the way a person actually talks, not the way a document reads. No bullet points, no headers, just spoken sentences. Keep it under 80 words. </constraints> <format> Return the paragraph only, nothing else. </format>

💡

Pro tip: Read it out loud before using it, if any sentence trips your tongue it will trip in the actual conversation, ask for a rewrite of that line only.

Write a follow up email after a business case meeting

28/30

✨ What it does

Claude writes a concise follow-up email that confirms the meeting decision, owners, and next checkpoint date. Send it the same day, then put the checkpoint on your calendar.

You are an executive assistant who writes precise, action oriented follow up emails after decision meetings. <context> I just presented this business case in a meeting and need a follow up email that captures the decision and next steps accurately. </context> <inputs> - What was decided or requested in the meeting: [MEETING OUTCOME] - Open action items and who owns them: [ACTION ITEMS AND OWNERS] - Next checkpoint date: [NEXT CHECKPOINT DATE] </inputs> <task> Write a follow up email that confirms the decision made, lists action items with owners and deadlines, and states the next checkpoint date. </task> <constraints> Keep the email under 150 words. Do not restate the entire business case, this is a confirmation email, not a summary document. Use a neutral, professional tone with no exclamation points. </constraints> <format> Return the email with a subject line, followed by the body text. </format>

💡

Pro tip: Send this within a few hours of the meeting while attendees still remember the same version of what was decided, ambiguity grows the longer you wait.

Prepare answers for a finance committee's standard questions

29/30

✨ What it does

Claude drafts fact-grounded answers to the finance committee's usual questions and flags where this case breaks the usual pattern. Fill in your own numbers, then keep the answers in front of you during the review.

You are a finance controller who prepares proposal owners for standardized capital committee reviews. <context> My business case is going in front of a finance committee that always asks the same set of standard questions, and I want prepared answers grounded in my own numbers. </context> <inputs> - Business case key figures: [KEY FIGURES] - Standard questions this committee typically asks: [STANDARD QUESTIONS] - Anything unusual about this specific request: [UNUSUAL FACTORS] </inputs> <task> Answer each standard question directly using the specific figures from this business case, and flag any question where the unusual factors mean the normal answer pattern does not apply. </task> <constraints> Every answer must cite an actual number or fact from the inputs, not a generic finance committee answer. Keep each answer to two sentences maximum. </constraints> <format> Return each standard question followed immediately by its answer, in the order given. </format>

💡

Pro tip: Ask a colleague who has sat through this committee before for their actual standard questions, generic finance questions rarely match a specific committee's real habits.

Summarize the business case for a written status update

30/30

✨ What it does

Claude writes a recurring status update on the approved case and says so clearly when nothing has changed. Send it on your usual cadence, then do not skip a week just because there is no news.

You are a project coordinator who writes recurring status updates for stakeholders not in the room for every meeting. <context> This business case is now approved and in motion, and I need to report progress to stakeholders who were not part of the original decision meeting. </context> <inputs> - Original business case goal: [ORIGINAL GOAL] - Progress so far: [PROGRESS SO FAR] - Any changes to cost or timeline since approval: [CHANGES SINCE APPROVAL] - Reporting period: [REPORTING PERIOD] </inputs> <task> Write a short status update covering progress against the original goal, any changes to cost or timeline, and what is planned for the next reporting period. </task> <constraints> If there are no changes to cost or timeline, say so plainly rather than omitting the section, silence reads as evasive to stakeholders tracking approved spend. Keep the update under 200 words. </constraints> <format> Return three short sections labeled Progress, Changes, and Next period. </format>

💡

Pro tip: State explicitly when cost and timeline are unchanged rather than dropping those sections, stakeholders tracking spend read omission as something being hidden.

Free tool

Prompt Optimizer

Turn a rough idea into a structured, professional AI prompt.

Try it free →

Frequently Asked Questions

A business case prompt gives Claude your actual numbers, constraints, and audience up front, then asks for a specific artifact like a cost-benefit table or a payback calculation. A general prompt like write a business case for a new tool produces filler because Claude has nothing concrete to calculate or compare.
Yes, if you give it the raw cash flow numbers and ask it to show the math year by year. Claude handles simple payback, ROI, and net present value calculations reliably when the formula and inputs are explicit, but you should still spot check the arithmetic on anything going to a finance committee.
Three is the practical standard, often framed as do minimum, do some, and do more. A single proposal invites the question of what else was considered, and more than three options tends to slow down the actual decision without adding real clarity.
No. Ask Claude to reframe emphasis and language for a specific reader, but never to alter figures or the underlying recommendation. A business case that is inconsistent between versions sent to different people is a credibility risk, not a communication improvement.
Explicitly ask it to play the skeptic and challenge your own assumptions before you present them, and tell it to round conservatively when a benefit is uncertain. The prompts in the cost-benefit and risk sections above are built specifically to surface overstated assumptions before a reviewer does.

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.