AI Prompt Library / Ideation and Validation
AI Prompts for Startup Ideation and Validation
A curated collection of expert-level prompts and techniques for the start of your startup journey: generating ideas that fit your skills, stress-testing them against the market, understanding your future customers, and planning what to build first.
In short
This collection gives you 25 core prompting techniques explained with examples, then 18 copy-paste templates: five for ideation and validation (idea generator, market viability analysis, user persona generator, pain point miner, value proposition crafting), five for MVP and product development (naming, user stories, walking skeleton outline, onboarding flow, AI error handling), five for prompt engineering workflows (universal template generator, few-shot curator, chain-of-thought deconstructor, workflow automator, refinement coach), and three worked examples that combine several techniques at once. Replace the [bracketed placeholders] with your own context, paste the result into any general-purpose AI assistant, and treat the output as a draft to test with real people.
How to use these templates
Every prompt below follows the same anatomy: a role for the AI to adopt, the task you want done, your specific context in [square brackets], and the exact format of the answer. That structure is deliberate. It layers several of the techniques explained further down this page, especially role-based prompting, instructional prompting, and formatting constraints.
Copy a template, swap every placeholder for your real details, and send it. If the first answer misses, do not start over: run the Prompt Refinement Coach template from the workflows section, feeding back what you wanted versus what you got.
Top 25 prompting techniques explained
Master these core techniques to unlock the full potential of any AI model. They are the foundation underneath every template in this library.
1. Zero-Shot Prompting
Asking a task without prior examples.
"Summarize this article."
2. Few-Shot Prompting
Providing 2-5 examples to guide the AI's output format and style.
"Q: What is the capital of France? A: Paris. Q: What is the capital of Japan? A: [AI completes]"
3. Chain-of-Thought (CoT)
Instructing the AI to think step-by-step to break down complex problems.
"Calculate the ROI, showing your work step-by-step."
4. Role-Based Prompting
Assigning an expert persona to the AI to guide its tone and knowledge.
"Act as a senior financial analyst and critique this plan."
5. RAG (Retrieval-Augmented Generation)
Providing specific documents for the AI to use as a source of truth.
"Using the attached manual, explain how to reset the device."
6. Tree of Thoughts (ToT)
Asking the AI to explore multiple reasoning paths and choose the best one.
"Brainstorm 3 marketing strategies, evaluate their pros/cons, then recommend one."
7. Instructional Prompting
Giving clear, numbered instructions for a multi-part task.
"First, summarize. Second, extract themes. Third, draft 3 posts."
8. Formatting Constraints
Specifying the exact output format (e.g., table, JSON, bullet points).
"Provide the answer as a Markdown table."
9. Negative Prompting
Telling the AI what to avoid or exclude from its response.
"Describe remote work benefits, but do not mention video conferencing."
10. Comparative Analysis
Asking the AI to compare and contrast multiple items on specific criteria.
"Compare Product A and B on price, features, and reviews."
11. Template-Based Prompting
Creating a fill-in-the-blanks template for the AI to complete.
"Complete this summary: Topic: [Topic], Decisions: [AI fills this in]."
12. Meta-Prompting
Asking the AI to help you create a better prompt for a specific task.
"Critique this prompt and suggest three improvements."
13. Self-Consistency
Asking for multiple answers to the same question and taking the consensus.
"What's Australia's capital? Answer 3 times."
14. Persona Adoption
Guiding the AI's personality and communication style.
"Respond to this complaint in a friendly, empathetic tone."
15. Context Stuffing
Providing a large amount of background text for the AI to reference.
"Given the company history below, draft a mission statement."
16. Step-Back Prompting
Asking the AI to abstract away from details to find a higher-level solution.
"Instead of fixing this specific bug, what's a general principle to prevent this type of error?"
17. Output Priming
Starting the AI's response for it to guide the format and content.
"Here is a summary of the report:
1. Key Finding A..."
18. Flipped Interaction
Instructing the AI to ask you questions to gather the information it needs.
"Ask me 5 questions you need to know to write a comprehensive business plan."
19. Guardrail Prompting
Setting explicit do-not-do rules to ensure safety and compliance.
"You must not provide medical advice. If asked, refer the user to a doctor."
20. Alternative Generation
Requesting multiple different versions or approaches to a single problem.
"Write three completely different headlines for this blog post."
21. Analogy Prompting
Asking the AI to explain a complex topic using a simple analogy.
"Explain blockchain technology like you're explaining it to a fifth grader using a Lego analogy."
22. Persona Fusion
Combining multiple expert roles for a more nuanced perspective.
"Act as a CFO with the communication skills of a marketing director."
23. Emotional Context Setting
Informing the AI about the emotional state of the user or audience.
"The customer is very frustrated. Draft a response that is highly empathetic and de-escalating."
24. Counterfactual Prompting
Asking the AI to reason about what would have happened if a condition were different.
"What would have been the likely outcome if we had launched in Q2 instead of Q4?"
25. Structured Input (Delimiters)
Using clear markers like ### or XML tags to separate different parts of your input.
"###Instruction###
Summarize the text.
###Text###
[Article text]"
Ideation and validation prompts
Five copy-paste templates for the earliest decisions: what to build, whether the market cares, who exactly it serves, and how to say it.
1. The AI Startup Idea Generator+
Purpose: Generate viable, AI-first business ideas based on your unique skills and market trends.
When to use: At the very beginning of your journey when you need inspiration and want to explore opportunities aligned with your strengths.
Act as a startup incubator strategist with expertise in identifying high-growth AI opportunities. My background is in [Your Field/Industry, e.g., 'digital marketing for e-commerce brands']. My key skills are [Skill 1, e.g., 'SEO'], [Skill 2, e.g., 'data analysis'], and [Skill 3, e.g., 'client relationship management']. I am passionate about [Your Passion, e.g., 'helping small businesses compete'].
Based on my profile and current AI trends, generate 5 unique AI startup ideas. For each idea, provide:
1. Idea Name: a catchy, descriptive name.
2. One-Liner: a single sentence explaining the business.
3. Problem Solved: the specific "hair-on-fire" problem it addresses.
4. Core AI Function: the key AI technology that drives the solution.
5. Monetization Model: a realistic pricing strategy.
2. The Market Viability Analysis Prompt+
Purpose: Get a quick, AI-driven analysis of a startup idea's market potential.
When to use: When you have a promising idea and need a fast reality check before investing significant time.
Act as a market research analyst for a venture capital firm. I am evaluating a startup idea: [Your one-sentence startup idea, e.g., 'An AI tool that automates social media content creation for local restaurants'].
Provide a concise market viability report covering:
1. Market Size & Trend: is this a large and growing market?
2. Competitive Landscape: who are the top 3 direct and indirect competitors?
3. Key Differentiator: what is the most significant potential advantage?
4. Customer Urgency: how severe is the problem this idea solves?
5. Potential Roadblocks: what are the 2-3 biggest challenges?
Structure the output as a professional memo.
3. The High-Fidelity User Persona Generator+
Purpose: Move beyond generic profiles and create detailed, empathetic personas that guide product development and marketing.
When to use: After selecting a promising idea and needing to deeply understand your target customer.
Act as an expert UX researcher. My startup, [Startup Name], solves [The Problem] for [Target Market].
Based on this, create two distinct user personas: a 'Primary Persona' (ideal early adopter) and a 'Secondary Persona' (future growth market).
For each persona, generate a detailed profile including:
- Name & photo idea
- Demographics & role
- Goals & frustrations
- Watering holes (where they get info)
- A day-in-the-life narrative
4. The AI-Powered Pain Point Miner+
Purpose: Extract genuine, high-urgency customer problems from real-world conversations on forums and social media.
When to use: Anytime you need to connect with the authentic voice of your customer.
Act as a senior customer insight analyst. I have pasted user comments from [Source, e.g., 'a LinkedIn thread about project management software'].
Analyze these raw comments and perform the following synthesis:
1. Top 5 pain points
2. Verbatim "golden" quotes (3-5 emotional quotes)
3. Unmet needs (3 inferred needs)
4. Urgency score (1-10 for each pain point)
5. Executive summary (a 3-sentence takeaway)
Here are the comments:
[Paste raw comments here]
5. The Value Proposition Crafting Prompt+
Purpose: Translate your solution into a compelling, customer-centric message that drives action.
When to use: After validating a problem and before building your landing page or MVP.
Act as a brand strategist in the style of Donald Miller's "Building a StoryBrand." My startup helps [Target Customer] solve [Validated Problem].
Generate these messaging components:
1. The Villain: the external, internal, and philosophical problems.
2. The Hero's Goal: what the customer wants to achieve.
3. Our Role as the Guide: how we help.
4. The Plan: a simple 3-step plan for using our product.
5. The Core Value Proposition: 3 variations of a compelling value proposition.
MVP and product development prompts
Once an idea survives contact with the market, these five templates turn it into something buildable.
1. The Startup Naming and Brand Identity Prompt+
Purpose: Brainstorm memorable names and establish a foundational brand identity for your new venture.
When to use: Once you have a validated idea and are ready to start building your brand presence.
Act as a world-class branding strategist. My startup is [one-sentence startup idea]. Our target audience is [Target Customer]. Our core values are [Value 1], [Value 2], and [Value 3].
Generate the following branding assets:
1. 10 startup names: short, memorable, and likely to have an available .com domain.
2. 3 taglines: for the top 3 name options.
3. Brand Personality: describe the brand's voice in 3-5 keywords.
4. Mission Statement: a concise, inspiring mission statement.
2. The User Story and Acceptance Criteria Prompt+
Purpose: Translate a feature idea into a clear, actionable user story with specific acceptance criteria for development.
When to use: When planning your MVP features to ensure development is focused on user value.
Act as an expert Agile Product Owner. I need to define a user story for a key feature of my MVP.
Feature idea: [Briefly describe the feature, e.g., 'An AI that summarizes long articles for users'].
User persona: [Briefly describe the user, e.g., 'A busy market researcher'].
Create the following:
1. User Story: write a classic user story in the format "As a [persona], I want to [action], so that [benefit]."
2. Acceptance Criteria: list 5-7 specific, testable criteria in the format "Given [context], when I [action], then I [expected outcome]."
3. The MVP Walking Skeleton Outline Prompt+
Purpose: Outline the absolute minimum technical components required to make your MVP function end-to-end.
When to use: For technical or semi-technical founders planning their MVP architecture.
Act as a senior software architect specializing in lean MVP development. My product, [Product Name], needs a "walking skeleton" technical outline. The core user action is [describe the key user action].
Outline the absolute minimum components required for an end-to-end functional MVP:
1. Frontend: what are the 3-4 essential UI screens or components?
2. Backend Logic: what is the single most important workflow?
3. Data Model: what are the 2-3 essential data tables needed?
4. The User Onboarding Flow Design Prompt+
Purpose: Map out the ideal user experience from first contact to achieving the "aha!" moment.
When to use: Before designing your MVP's user interface to ensure high activation rates.
Act as an expert UX/UI designer specializing in user onboarding. My product is [Product Description]. The primary "aha!" moment for the user is when they [describe the core value moment].
Design a frictionless user onboarding journey in 5-7 steps. For each step, describe:
1. The user action
2. The interface
3. The goal
5. The AI Feature Error Handling Prompt+
Purpose: Proactively design graceful failure states for your AI feature, which builds user trust.
When to use: During MVP design to account for the imperfections of AI technology.
Act as a senior product designer with expertise in AI applications. My core AI feature is [describe the AI feature].
Brainstorm 3-5 potential failure modes for this AI feature. For each, design a user-friendly way to handle the error, including the error message copy and the user recovery path.
Example failure modes:
- The AI API is down.
- The AI returns a low-confidence result.
Prompt engineering and workflow prompts
Meta-level templates: use AI to improve the way you work with AI.
1. The Universal Business Prompt Template Generator+
Purpose: Have the AI generate a fill-in-the-blanks prompt template for any recurring business task.
When to use: When starting any new or complex AI task to ensure your prompt is structured for success.
Act as a prompt engineering coach. I want to [describe your business task, e.g., 'write a client proposal'].
Generate a fill-in-the-blanks prompt template for me to use, based on a best-practice framework. The template should include sections for Role, Task, Context, Format, Constraints, and Examples, with guiding questions for each section.
2. The Few-Shot Example Curator+
Purpose: Create high-quality, diverse examples for few-shot prompting.
When to use: When you need to teach the AI a specific style or format but don't have good existing examples.
Act as an expert AI training data creator. I need to create a few-shot prompt to teach an AI how to [your goal, e.g., 'write empathetic customer service apologies']. My brand voice is [describe your voice, e.g., 'professional, accountable, but warm'].
Generate three distinct, high-quality examples I can use in my prompt. They should cover:
1. A standard scenario
2. A complex scenario
3. An edge case
3. The Chain-of-Thought Deconstructor+
Purpose: Break down a complex business problem into a logical, step-by-step reasoning plan.
When to use: When facing a multi-faceted decision, such as a market entry strategy.
Act as a senior strategy consultant. My business is facing a complex challenge: [describe the problem, e.g., 'Deciding whether to acquire a small competitor'].
Deconstruct this problem into a step-by-step 'Chain-of-Thought' analysis plan. Provide 5-7 logical steps an analyst would take to arrive at a well-reasoned recommendation.
4. The AI Workflow Automator+
Purpose: Design an automated workflow connecting multiple apps and AI steps.
When to use: When you want to automate a process like lead nurturing or content repurposing.
Act as a workflow automation expert. I want to automate the process: [describe the process, e.g., 'When a new lead fills out a form, analyze their company, draft a personalized outreach email, and add them to my CRM']. The tools I use are [list your tools].
Design a step-by-step workflow automation. For each step, specify the Trigger, the Action, and the Data passed.
5. The Prompt Refinement Coach+
Purpose: Get specific, actionable feedback on your own prompts to improve their performance.
When to use: When you're getting inconsistent or low-quality results and don't know why.
Act as a world-class prompt engineering coach. My original prompt was:
"[Paste your original prompt]"
The AI's output was:
"[Paste the disappointing output]"
My goal was to get:
"[Describe your desired outcome]"
Please critique my original prompt and provide a revised, improved version. Explain exactly why your version is better.
Combining techniques for expert-level results
The true power of prompt engineering comes from layering techniques to solve complex problems. Here are three worked examples of combining core principles for consulting-grade outputs.
Example 1: Strategic report generation
Combines Role-Based prompting to set expertise, Chain-of-Thought for systematic analysis, and Formatting Constraints for a professional output.
Act as a senior market analyst for the renewable energy sector.
Analyze the viability of a new solar panel startup. Work step-by-step:
1. Assess the current market size and growth projections.
2. Analyze the top 3 competitors and their market share.
3. Evaluate the potential impact of new government regulations.
4. Provide a final recommendation with a risk assessment.
Format the entire analysis as an executive memo.
Example 2: Empathetic customer complaint response
Layers Persona Adoption for tone, Few-Shot Prompting for structure and style, and Negative Prompting for safety.
Adopt a highly empathetic but professional customer service persona.
Respond to the customer complaint below, following the style of these examples:
[Example 1: successful apology email]
[Example 2: another successful apology email]
Do not make any promises about financial compensation.
Customer complaint: [Paste customer email here]
Example 3: Innovative brainstorming session
Uses Tree of Thoughts to explore options, Alternative Generation to ensure diversity of ideas, and self-critique to refine the final output.
I need innovative names for a new AI-powered coffee brand. Your task is to:
1. Brainstorm three distinct thematic directions (e.g., 'futuristic,' 'organic,' 'productivity-focused').
2. For each theme, generate 5 potential names.
3. Critique your own list of 15 names, identifying the top 3 overall.
4. For the top name, explain why it is the most effective.
Put these prompts in context
BLOG GUIDE
ICP Template for Startups: the 7-field Ideal Customer Profile framework
BOOK
60-Day Founder: A Builder's Guide to Production AI
ACADEMY
60-Day Founder: from raw idea to investor-ready proof
Back to the hub: AI Prompt Library for Solo Founders, or continue with The Solopreneur's AI Arsenal summary.
Frequently asked questions
What are the best AI prompts for validating a startup idea?+
This library walks validation in five steps: the Market Viability Analysis prompt gives you a fast reality check on market size, competitors, and urgency; the AI-Powered Pain Point Miner extracts real customer problems from raw forum and social media comments; the High-Fidelity User Persona Generator turns the target market into two concrete personas; the Value Proposition Crafting prompt shapes the message; and the AI Startup Idea Generator helps if you have not chosen an idea yet.
How many prompts and techniques are in this library?+
The library contains 25 core prompting techniques explained with examples, 15 copy-paste templates organized into three phases (ideation and validation, MVP and product development, and prompt engineering workflows), and 3 worked examples that combine multiple techniques in a single prompt.
What is the difference between zero-shot and few-shot prompting?+
Zero-shot prompting asks the AI to do a task without providing prior examples, such as asking it to summarize an article. Few-shot prompting provides 2 to 5 examples first so the AI can match the format and style you want, which is useful when you need consistent output like branded emails or structured tables.
Can I use these prompts with any AI model?+
Yes. Every template here is plain text: a role, a task, context placeholders in square brackets, and formatting instructions. Paste any of them into a general-purpose AI assistant, replace the bracketed placeholders with your own details, and adapt the wording to your situation.