Launch price: $49 one-time for all 38 specialists. Get it →
Claude prompts
10 Claude Prompts for Business, Strategy and Small Business
Business plans, idea validation, pricing, market research, project plans, SOPs, HR and contract review.
Updated October 202610 promptsFree to copy, no signup
These Claude prompts help you plan, test and run a business with honest analysis rather than encouragement. Several of them interview you before writing, label assumptions separately from facts and end with a test or decision point, so the output is something you can act on and defend.
How to use these prompts
Answer the interview questions with real numbers, even rough ones, and say when you do not know.
Ask Claude to label which statements are assumptions so you know what to verify.
Keep each plan or analysis in its own chat or Claude Project so the context stays focused.
Take legal, tax and financial questions from these prompts to a qualified professional before acting.
Use it when: Use it to draft a business plan grounded in your actual business, with Claude asking the right questions before it writes a word.
Prompt 01
Help me write a business plan for [BUSINESS NAME OR IDEA]. The plan is for [PURPOSE: BANK LOAN, INVESTORS, A GRANT, OR MY OWN CLARITY].
Before writing anything, interview me. Ask one question at a time and wait for my answer. Cover the problem and customer, the product or service, how I'll reach customers, pricing and costs, competition, the team, traction so far, the funding I need and what it's for, and the biggest risks. Ask follow-ups when an answer is vague. Twelve to fifteen questions is plenty.
Then write the plan with these sections: executive summary, problem and solution, target market, competition and differentiation, marketing and sales, operations, team, financial summary (revenue model, key assumptions, a simple 12-month projection, break-even point), funding request, risks and mitigations, and milestones.
Ground rules:
- Use only my answers. Where a number is an assumption, label it and show the reasoning.
- Where I don't know something, write a clearly marked gap saying what I need to find out.
- Write for the audience I named: lenders focus on cash flow and repayment, investors on market size and growth.
At the end, list the three weakest parts of the plan and how to strengthen each one.
Why it works: The interview gathers your real numbers and context before drafting, so the plan describes your business instead of a generic one. Labelling assumptions and marking unknowns as gaps shows a lender or investor exactly what is evidence and what still needs work.
Tip: Answer with real numbers even when they are rough; a plan built on honest estimates survives tough questions far better than a polished guess.
2. Brutally honest business idea validation prompt
Claude chat
Use it when: Use it before you invest time or money in an idea to stress-test demand, competition and risk, and find the cheapest way to test it.
Prompt 02
Give me a brutally honest assessment of a business idea. Please skip the cheerleading: I'd rather find the flaws now than after I've spent my savings.
<idea>[WHAT IT IS, WHO IT'S FOR, HOW IT MAKES MONEY]</idea>
<me>[YOUR RELEVANT SKILLS, BUDGET, TIME AVAILABLE, ANY UNFAIR ADVANTAGE]</me>
<evidence>[ANY SIGNS OF DEMAND: CONVERSATIONS, SEARCHES, PRE-ORDERS, WAITLIST]</evidence>
Assess it in this order:
1. Steelman: the strongest honest case for the idea.
2. Demand: who has this problem badly enough to pay, how they solve it today, and how strong my evidence really is. Separate what I've observed from what I'm assuming.
3. Competition and alternatives, including doing nothing or using a spreadsheet.
4. Economics: rough unit economics and what would have to be true for this to work, with assumptions labelled.
5. The top five risks, ranked by how likely each is to kill the business.
6. The cheapest test: an experiment I can run in two weeks for little money that would prove or disprove real demand, with a clear pass or fail threshold.
7. Verdict: pursue, pivot or pass, with your confidence level and the evidence that would change your mind.
Where you lack market data, say so plainly instead of estimating figures.
Why it works: Steelmanning the idea before attacking it gives a fair assessment instead of reflexive negativity or cheerleading. Separating observed demand from assumed demand and ending with a cheap test that has a pass or fail threshold turns opinion into something you can check with real customers.
Tip: Run the cheapest test before building anything; people saying an idea sounds great is very different from people paying for it.
3. Pricing strategy prompt for your product
Claude chat
Use it when: Use it to work out how to price a product or service, with packages, a value metric and a safe way to test price changes.
Prompt 03
Help me set pricing for [PRODUCT OR SERVICE].
<context>
What it does and for whom: [DETAILS]
Current pricing and results, if any: [PRICE, CONVERSION RATE, CHURN, FEEDBACK]
Costs: [COST TO DELIVER PER CUSTOMER OR UNIT, FIXED COSTS]
Alternatives and their prices: [COMPETITORS, OR WHAT CUSTOMERS DO INSTEAD]
Customer segments and how value differs between them: [DETAILS]
Goal: [E.G. MORE REVENUE PER CUSTOMER, MORE SIGNUPS, MOVE UPMARKET]
</context>
Work through:
1. The value created for each segment in their terms (time saved, money made, risk avoided) and a rough sense of willingness to pay. Label anything that is a guess.
2. The pricing model that fits (per seat, usage-based, flat tiers, per project, retainer) and the value metric that grows as the customer gets more out of it.
3. A recommended structure of two or three tiers or packages, what goes in each, and why most customers should land on the middle one.
4. A price floor from my costs and a ceiling from the alternatives, and where I should sit between them.
5. Risks with existing customers and how to grandfather or communicate any change.
6. How to test it: customer conversations, a price test on new signups, or a staged rollout, and what to measure.
If my costs or segments are unclear, ask me before recommending numbers.
Why it works: Starting from the value each segment gets, then bounding the price between your costs and the alternatives, grounds the recommendation in logic you can check. Asking for a test plan and a way to handle existing customers keeps a price change from becoming an expensive surprise.
Tip: Ask five recent customers what they would use if you disappeared tomorrow; their answers set your real price ceiling.
4. Market research plan prompt
Claude chat
Use it when: Use it to separate what you actually know about a market from what you are assuming, and get a research plan to close the gaps that matter.
Prompt 04
Help me plan market research for [MARKET, PRODUCT OR DECISION].
The decision this research should inform: [E.G. WHETHER TO LAUNCH IN A NEW CITY, WHICH SEGMENT TO TARGET, WHETHER TO RAISE PRICES]
What I already know or believe: [PASTE NOTES, DATA, CUSTOMER FEEDBACK, REPORTS YOU HAVE READ]
Budget and time: [BUDGET, WEEKS AVAILABLE]
1. Sort everything I've given you into three groups: known (backed by data or direct observation, with the source), believed (my assumptions or second-hand claims), and unknown (questions that matter but have no answer yet). Be strict: anything without a source goes under believed.
2. Rank the beliefs and unknowns by how much the decision would change if they turned out wrong.
3. For the top five, propose the best research method (customer interviews, a survey, search data, competitor analysis, public reports, a small test), what exactly to ask or measure, and why that method suits the question.
4. Draft a short plan: order of work, rough time and cost, and what result would change the decision.
5. Write five interview questions about past behaviour, not hypotheticals, for the most important unknown.
Leave gaps as gaps: list any market sizes or statistics I need as things to find, rather than quoting figures you can't source.
Why it works: Sorting your information into known, believed and unknown exposes how much of your picture rests on assumptions. Ranking unknowns by how much they would change the decision focuses limited research budget on the questions that matter, and interview questions about past behaviour produce more reliable answers than hypotheticals.
Tip: Ask people what they did the last time they had the problem, not what they would do; past behaviour is far more reliable.
5. Realistic ways to make money with your skills
Claude chat
Use it when: Use it to find realistic ways to earn extra income that fit your skills, time and budget, with honest expectations and risks.
Prompt 05
Help me find realistic ways to earn extra income. I want honest options, not get-rich-quick ideas.
<me>
Skills and experience: [WHAT YOU'RE GOOD AT, INCLUDING HOBBIES]
Time available: [HOURS PER WEEK, AND WHEN]
Money I can invest: [AMOUNT, CAN BE ZERO]
Location and restrictions: [COUNTRY, LIMITS IN YOUR EMPLOYMENT CONTRACT, VISA RULES]
What I enjoy and what I'd hate doing: [DETAILS]
Income goal: [AMOUNT PER MONTH, AND BY WHEN]
</me>
1. Suggest 6 to 8 options that fit my skills and constraints, mixing service work (freelancing, consulting, tutoring), products (digital or physical) and local or part-time work.
2. For each: what the work actually involves, how people usually find their first paying customer, startup cost, time to first income, a realistic earnings range for a beginner, and the main risks. Mark earnings as rough and explain what they depend on.
3. Rank the options by fit for me and explain the ranking.
4. For the top two, give a four-week plan to land the first paying customer.
5. Warn me about scams and traps common in this space, such as courses promising income, upfront fees and MLMs, and flag tax or contract questions to check with a professional.
If my goal is unrealistic for the time I have, tell me plainly.
Why it works: Feeding in your skills, hours, budget and restrictions filters out options that sound good but do not fit your life. Asking for startup cost, time to first income, rough earnings with their drivers and common scams for each option gives you a realistic comparison instead of hype.
Tip: Selling a skill you already have usually earns sooner than building a product; start with a service, then package what repeats.
6. Low-budget small business marketing plan prompt
Claude chat
Use it when: Use it to get a practical, local-first marketing plan for a small business with very little money and limited time.
Prompt 06
Create a marketing plan for my small business. My budget is tiny, so focus on what works with time and effort rather than ad spend.
<business>
What we sell and to whom: [DETAILS]
Location and service area: [CITY, REGION, OR ONLINE]
Monthly marketing budget: [AMOUNT]
Hours per week I can spend: [HOURS]
How customers find us today: [WORD OF MOUTH, GOOGLE, WALK-INS, SOCIAL MEDIA]
What I've tried and what happened: [DETAILS]
</business>
1. Quick wins for the first two weeks: the free or cheap basics many small businesses neglect, such as a complete Google Business Profile, asking happy customers for reviews, a clear web page with a phone number and call to action, and relevant directory listings. Tailor the list to my type of business.
2. Two ongoing channels that suit my customers and my hours, each with a weekly routine.
3. A referral or repeat-customer idea that uses the customers I already have.
4. One partnership idea with a nearby or complementary business.
5. Whether to spend the small budget at all, on what, and what to try only after the free channels are working.
6. A simple tracking sheet recording where each new customer came from, reviewed weekly.
Make every action concrete enough to start this week. Ask me questions if my type of business changes what will work.
Why it works: Leading with free basics before paid channels matches the order that usually pays off for a small local business. Limiting ongoing work to two channels that fit your hours, plus a simple record of where each customer came from, keeps the plan doable and shows what is working.
Tip: Ask every new customer how they found you and write it down; after a month you will know exactly where to double down.
7. Project plan prompt with milestones and risks
Claude chat
Use it when: Use it to turn a project goal into a plan with phases, milestones, owners, dependencies, a critical path and a risk register.
Prompt 07
Act as an experienced project manager and turn this into a project plan.
<project>
Goal and definition of done: [WHAT SUCCESS LOOKS LIKE]
Deadline and fixed dates: [DATES]
Team and roles: [NAMES OR ROLES, AND THEIR AVAILABILITY]
Budget: [AMOUNT, OR NOT APPLICABLE]
Constraints and dependencies: [OTHER TEAMS, VENDORS, APPROVALS]
Background: [PASTE THE BRIEF, MEETING NOTES OR SCOPE DOCUMENT]
</project>
1. Clarify scope: what's in, what's out, and the assumptions. Ask me about anything ambiguous before going further.
2. Break the work into phases and milestones. Each milestone needs a deliverable someone can check, rather than an activity like "work on design".
3. A task table: task, owner, estimate, dependency, start and end dates. Flag anyone who is overloaded.
4. The critical path: the tasks that delay the whole project if they slip.
5. A risk register: the top risks with likelihood, impact, an early warning sign, a mitigation and an owner.
6. Communication: who gets updates, how often and in what format, plus the decision points where a stakeholder must sign off.
Keep estimates realistic and add a buffer to the critical path, saying how much and why. If the deadline isn't achievable with this team and scope, tell me and suggest what to cut or move.
Why it works: Requiring milestones that are checkable deliverables rather than activities makes progress visible to everyone. The critical path, an overload check and a risk register with early warning signs show where the schedule will really slip, and asking Claude to say when the deadline is unrealistic avoids a plan built on hope.
Tip: Paste the task table into your spreadsheet or PM tool, then ask Claude for a weekly status update template that reports against it.
8. SOP writer prompt for any process
Claude chat
Use it when: Use it to turn the way you do a task into a clear standard operating procedure that a new person can follow without asking you.
Prompt 08
Help me write a standard operating procedure (SOP) for [PROCESS NAME].
Who will follow it: [ROLE AND EXPERIENCE LEVEL, E.G. A NEW PART-TIME ASSISTANT]
How I do it now: [DESCRIBE IT LOOSELY, PASTE NOTES, OR PASTE A TRANSCRIPT OF YOU TALKING IT THROUGH]
Tools involved: [SOFTWARE, EQUIPMENT, ACCESS NEEDED]
First, ask me about anything a newcomer would get stuck on: decisions they'd need to make, exceptions, what "done" looks like, and what can go wrong. Put all your questions in one numbered list and wait for my answers.
Then write the SOP with:
1. Purpose: one or two lines on why the process exists.
2. Scope and trigger: when to run it and when it doesn't apply.
3. Before you start: the access, tools and inputs needed.
4. Numbered steps, one action each, starting with a verb. Write decisions as "If X, then Y", and mark where a screenshot would help.
5. Quality check: how to confirm it was done correctly.
6. Common problems and their fixes.
7. Who to ask, and when to escalate.
8. Process owner and last-reviewed date.
Write at a level the person I described can follow without me in the room.
Why it works: Asking about decisions, exceptions and what done looks like before drafting captures the knowledge experts leave out because it feels obvious. One verb-led action per step and explicit if-then decisions let a newcomer follow the SOP without interrupting you.
Tip: Record yourself doing the task once while narrating, transcribe it and paste it in; you will capture steps you would never think to write down.
9. Job description prompt for HR and hiring
Claude chat
Use it when: Use it to write a specific, inclusive job description that attracts the right candidates and sets clear expectations from day one.
Prompt 09
Write a job description for a [JOB TITLE] at [COMPANY].
<details>
What this person will actually do in a typical week: [DETAILS]
What success looks like after 6 and 12 months: [OUTCOMES]
Team, reporting line and key collaborators: [DETAILS]
Must-have requirements: [ONLY THE TRULY ESSENTIAL ONES]
Nice-to-haves: [LIST]
Salary range, location, remote policy and benefits: [DETAILS]
About the company and how the team works: [DETAILS]
</details>
Guidelines:
1. Open with why the role exists and the impact it will have, in two or three sentences.
2. Describe responsibilities as real activities and outcomes, replacing vague phrases like "drive synergies".
3. Keep must-haves to five or six genuine requirements. Ask me to justify any degree or years-of-experience requirement, because unnecessary ones shrink the pool of good candidates.
4. Use inclusive, plain language: no gender-coded words like "rockstar" or "ninja", no age-coded phrases like "digital native", and no requirements unrelated to the work.
5. Include the salary range and a short note on the hiring process and timeline.
6. Add an accommodations statement and encourage people to apply even if they don't meet every nice-to-have.
After the draft, list the phrases you changed or would flag for bias, and check that the job title matches what candidates actually search for.
Why it works: Describing the role through weekly activities and outcomes at 6 and 12 months helps candidates judge fit honestly. Capping must-haves, questioning degree and years requirements and checking for coded language widen the pool of qualified applicants without lowering the bar.
Tip: Ask someone currently doing a similar job to read it; if they would not recognise the role, rewrite the responsibilities.
10. Contract review summary prompt in plain English
Claude chat
Use it when: Use it to understand a contract before you sign: key terms, obligations, risks, and the questions to take to a lawyer.
Prompt 10
Help me understand this contract in plain English before I sign it. I know this isn't legal advice; I want to understand what I'm agreeing to and what to ask a lawyer.
<contract>
[PASTE THE FULL CONTRACT TEXT]
</contract>
My role in this contract: [E.G. FREELANCER, TENANT, BUYER, EMPLOYEE]
What matters most to me: [E.G. PAYMENT TERMS, ENDING IT EARLY, OWNING MY WORK]
Jurisdiction, if known: [COUNTRY OR STATE]
Please give me:
1. A one-paragraph summary of what this contract does.
2. A key terms table: the term, what it means in plain English, and the clause number.
3. My obligations and deadlines, and the other party's.
4. Money: fees, payment timing, penalties, renewals and price changes.
5. Risks and unusual terms: anything one-sided, vague or worth negotiating, such as liability, indemnity, IP ownership, non-compete, auto-renewal, termination and dispute resolution. Quote the clause and explain why it could matter to me.
6. Anything that seems to be missing for a contract of this type.
7. Questions to ask the other party or a lawyer, in priority order.
Quote the contract when you make a point. Say clearly when something depends on local law or needs a lawyer to interpret, and flag ambiguous wording as ambiguous rather than guessing what it means.
Why it works: Clause numbers and quoted wording for every point let you verify the summary against the contract instead of trusting a paraphrase. Telling Claude what matters to you and to flag ambiguity and local-law questions turns the output into a focused list to take to a lawyer.
Tip: Remove names, signatures and account numbers before pasting if the review does not need them.
Getting more out of every prompt
Ask for the critic before the cheerleader
Claude tends to be agreeable unless you ask otherwise. Prompts like the brutally honest validation one explicitly ask for risks, a steelman and a verdict. Use that framing whenever you need an honest second opinion on a plan.
Feed it customer evidence
Paste interview notes, reviews, support tickets or sales call summaries into the strategy, pricing and research prompts. Real customer language makes the analysis specific and shows where your assumptions differ from what buyers say.
Turn plans into documents
Once a plan is solid, ask Claude to format it as a document or spreadsheet. Where file creation is available in Claude, it can produce Word, Excel or PowerPoint files you can share with partners or lenders.
Frequently asked questions
Can Claude write a business plan?
Yes, Claude can draft a complete business plan, but it is only as good as the information you give it. The most reliable approach is to let it interview you first, then write the plan with assumptions clearly labelled. Lenders and investors will check your numbers, so verify every figure yourself.
Can Claude review a contract?
Claude can summarise a contract in plain English, highlight unusual or one-sided terms and suggest questions to ask. It does not replace a lawyer and cannot account for every local law or your full situation. Use its summary to prepare for a conversation with a legal professional, not to decide alone.
Can Claude help me make money?
Claude can help you evaluate side-income ideas, plan a service business, write proposals and test demand cheaply. It cannot guarantee income, and any prompt or product promising easy money should be treated with caution. Realistic results depend on your skills, time and market.
Is it safe to paste business documents into Claude?
Review your organisation's policies and the privacy settings for your Claude plan before sharing confidential material. Remove personal details, account numbers and anything you are not permitted to share. Many tasks work just as well with names and figures anonymised.