# 30 Ready-to-Use AI Prompts for Solo SaaS Founders
## From Idea to $10k MRR — The Complete Prompt Operating System

**Companion to *The Solo SaaS Playbook*** · By Abhishek Verma · PlanMySaaS Press · 2026

---

## How to Use This Pack

**Step 1:** Find the prompt matching your current stage (Idea → Build → Launch → Grow).
**Step 2:** Copy the entire prompt block. Replace `[bracketed]` variables with your specifics.
**Step 3:** Paste into the recommended AI tool. Read the output, approve what's right, redirect what's not.

**These prompts work with ANY AI model:** ChatGPT, Claude, Gemini, Copilot, Llama, Mistral, or any other LLM. Just copy, fill the brackets, and paste.

**Which type of AI tool works best:**
| Task | Best Tool Type | Examples |
|------|---------------|----------|
| Strategy, planning, writing | Any chat-based AI | ChatGPT, Claude, Gemini, Perplexity |
| Code generation | Any AI code editor | Cursor, Windsurf, GitHub Copilot, Cody |
| Research with live data | AI with web search | ChatGPT, Perplexity, Gemini |
| Data analysis | AI with code execution | ChatGPT Code Interpreter, Claude Artifacts |

---

## STAGE 1: IDEA VALIDATION (Prompts 1–6)

### P1 · Validate Your Idea in 60 Seconds

**Use with:** Any AI · **Output:** 1-page assessment

> You are a brutally honest SaaS advisor who has evaluated 500+ startup ideas. I need you to evaluate mine.
>
> **My idea:** `[Describe your SaaS idea in 2-3 sentences]`
> **Target user:** `[Who is this for? Be specific — role, company size, industry]`
> **How they solve this today:** `[What do they currently use? Spreadsheets? A competitor? Nothing?]`
>
> Evaluate this idea against these 5 criteria. Score each 1-10:
> 1. **Pain severity** — Is this a painkiller or a vitamin?
> 2. **Willingness to pay** — Would this audience pay $29+/mo for this?
> 3. **Market size** — Is this a $1M or $100M opportunity?
> 4. **Solo-buildability** — Can one developer ship an MVP in 4 weeks?
> 5. **Defensibility** — What stops someone from cloning this in a weekend?
>
> Then give me:
> - A 1-sentence verdict (BUILD / PIVOT / KILL)
> - The single biggest risk
> - One thing I should validate before writing any code

### P2 · Generate Your ICP (Ideal Customer Profile)

**Use with:** Any AI · **Output:** Printable ICP card

> Create a detailed Ideal Customer Profile for my SaaS product.
>
> **Product:** `[What it does in one sentence]`
> **Problem it solves:** `[The specific pain point]`
> **Price point:** `[$X/mo]`
>
> Generate a structured ICP card with:
> - **Job title & seniority** (exact titles they'd have on LinkedIn)
> - **Company size** (employee count range)
> - **Industry verticals** (top 3, ranked by fit)
> - **Annual revenue range** of target companies
> - **Tech stack they likely use** (5-7 tools)
> - **Where they hang out online** (specific subreddits, Twitter accounts they follow, Slack communities, newsletters they read)
> - **Their daily frustration** (the moment in their day when they feel the pain)
> - **Budget authority** (can they buy $29/mo software without approval?)
> - **Search queries they'd type** (5 exact Google searches)
> - **Objections they'll raise** (top 3, with counter-arguments)

### P3 · Competitor Teardown Matrix

**Use with:** Any AI (web search enabled preferred) · **Output:** Comparison table

> Research and analyze the competitive landscape for `[your product category]`.
>
> Find 5 competitors (3 direct, 2 indirect). For each, create a row in a table with:
> - Name + URL
> - Pricing (exact tiers)
> - Core features (top 5)
> - Weakest point (from G2/Capterra 1-star reviews)
> - Strongest point (from G2/Capterra 5-star reviews)
> - Tech stack (check BuiltWith)
> - Estimated MRR (from public data, social proof, or team size)
> - Their positioning in one sentence
>
> Then answer: What gap exists that none of them fill? That's my opportunity.

### P4 · The "Why Now" Thesis

**Use with:** Any AI · **Output:** 1-paragraph pitch

> Write a compelling "Why Now" thesis for `[product]`.
>
> Explain why this product is possible, necessary, or profitable in 2026 in a way it wasn't in 2023. Frame it as ONE of:
> - **Technology gap** (a new API, model, or platform that didn't exist)
> - **Behavior gap** (a shift in how people work, buy, or communicate)
> - **Regulation gap** (a new law, compliance requirement, or industry standard)
>
> Cite at least one real, verifiable data point or trend. Keep it to one paragraph that I could use in a pitch deck or landing page.

### P5 · Name Your SaaS (10 Options)

**Use with:** Any AI · **Output:** Ranked list of 10 names

> Generate 10 name candidates for my SaaS product.
>
> **What it does:** `[one sentence]`
> **Target audience:** `[who]`
> **Tone:** `[professional / playful / technical / premium]`
>
> For each name:
> - The name itself
> - Is the .com domain available? (suggest checking)
> - One-sentence tagline that pairs with it
> - Why it works (etymology, sound, memorability)
>
> Constraints: Must be 1-2 words. Must be easy to spell over the phone. Must not conflict with an existing well-known brand. Avoid generic names like "AppFlow" or "DataHub."

### P6 · Map the Cost of Inaction

**Use with:** Any AI · **Output:** 4-row table

> For `[your ICP — specific role and company]`, calculate the cost of NOT solving `[your problem]` across 4 dimensions:
>
> | Dimension | Weekly Cost | How I Calculated It |
> |-----------|-------------|---------------------|
> | **Money** (direct spend) | $X | ... |
> | **Time** (hours wasted) | X hrs | ... |
> | **Reputation** (trust/brand damage) | Qualitative | ... |
> | **Opportunity** (revenue they're missing) | $X | ... |
>
> Be specific. Use real salary data, tool costs, and industry benchmarks. This table will go on my landing page to justify my pricing.

---

## STAGE 2: ARCHITECTURE & PLANNING (Prompts 7–12)

### P7 · The 9-Decision Tech Stack

**Use with:** Any AI · **Output:** Decision matrix

> I'm a solo developer building `[describe your SaaS]`. Help me make the 9 critical technology decisions.
>
> For each decision, give me: your recommendation, a one-line reason, and one alternative.
>
> 1. **Frontend framework**
> 2. **Hosting/deployment**
> 3. **Database**
> 4. **ORM**
> 5. **Authentication**
> 6. **Payments**
> 7. **Email service**
> 8. **AI/LLM provider** (if applicable)
> 9. **Monitoring/error tracking**
>
> Constraints: Optimize for speed-to-ship, not theoretical scalability. Prefer managed services (zero DevOps). Every choice must have a free tier. I'm shipping in 4 weeks.

### P8 · Generate Your Prisma Database Schema

**Use with:** Any AI code editor · **Output:** Complete schema.prisma file

> Generate a production-ready Prisma schema for `[your SaaS product]`.
>
> Requirements:
> - Multi-tenant architecture (User → Workspace → Resources)
> - NextAuth.js compatible (Account, Session, VerificationToken models)
> - Stripe billing fields on the Workspace model
> - My core domain model: `[describe your main resource — e.g., "Projects with tasks" or "Campaigns with emails"]`
>
> For each model, include:
> - All fields with correct types and decorators
> - Relations with onDelete cascade where appropriate
> - Indexes for common query patterns
> - A 1-line comment explaining WHY each model exists
>
> Use PostgreSQL as the datasource. Include both `url` and `directUrl` for Supabase connection pooling.

### P9 · Design Your API Routes

**Use with:** Any AI · **Output:** Route map + code skeleton

> Design the REST API route structure for `[your SaaS]` using Next.js App Router API routes.
>
> My core resources are: `[list your main entities]`
>
> For each resource, generate:
> - The file path (e.g., `/app/api/projects/route.ts`)
> - HTTP methods (GET, POST, PATCH, DELETE)
> - Request body schema (TypeScript type)
> - Response format
> - Auth requirement (public / authenticated / admin)
> - One-line description
>
> Also include routes for:
> - Stripe webhook handler
> - Auth endpoints
> - Any utility endpoints (search, export, etc.)

### P10 · AI Router Pattern

**Use with:** Any AI code editor · **Output:** TypeScript file

> Create an `ai/router.ts` file for my Next.js SaaS app with this pattern:
>
> - Takes a typed prompt config: `{ model: 'claude' | 'openai', systemPrompt: string, userMessage: string, temperature?: number, maxTokens?: number }`
> - Returns typed JSON response (parsed, not raw string)
> - Claude as primary, OpenAI as fallback (configurable via env var)
> - Automatic retry with exponential backoff (3 attempts)
> - Token usage logging
> - Error handling that never crashes the app
>
> Include a usage example showing how to call it from a Next.js API route.

### P11 · Cost Projection: 0 to 10,000 Users

**Use with:** Any AI · **Output:** Cost table

> Project my monthly infrastructure costs at different user scales.
>
> My stack: `[list your services — e.g., Vercel, Supabase, Stripe, Resend, OpenAI]`
>
> Create a table showing costs at: 0, 100, 500, 1,000, 5,000, 10,000 users.
> For each scale, show cost per service and total.
> Flag the "breakpoint" where each service jumps to a paid tier.
> Calculate cost-per-user at each scale.
>
> Also answer: At what user count does my infrastructure cost exceed $500/mo? What's the biggest cost driver?

### P12 · 4-Phase Development Plan

**Use with:** Any AI · **Output:** 12-week roadmap

> Create a 12-week, 4-phase development plan for `[your SaaS MVP]`.
>
> My feature list: `[paste your feature list]`
>
> For each phase (3 weeks each):
> - Phase name and goal
> - Which features ship in this phase
> - Daily task breakdown for each week
> - "Done" criteria — what must work by end of phase
> - What is FORBIDDEN in this phase (scope creep prevention)
>
> Phase 1 must end with a deployable app that does ONE core thing. Phase 4 must end with a product ready for paying customers.

---

## STAGE 3: BUILD & SHIP (Prompts 13–18)

### P13 · Build a Feature (Cursor Prompt)

**Use with:** Any AI code editor · **Output:** Working code

> Build `[feature name]` for my Next.js SaaS app.
>
> **Spec:**
> `[Paste your feature spec — what it does, inputs, outputs, edge cases]`
>
> **Constraints:**
> - Follow existing project conventions (check CLAUDE.md or .cursorrules)
> - Use existing UI components from `/components/ui/`
> - All database queries must filter by `workspaceId` (multi-tenant)
> - Include loading states and error handling
> - Mobile responsive
>
> **After building:**
> - Write 2 unit tests covering happy path and one failure mode
> - Write a meaningful commit message following conventional commits

### P14 · Write End-to-End Tests

**Use with:** Any AI code editor · **Output:** Test files

> Write Playwright end-to-end tests for the 3 critical flows that must NEVER break:
>
> 1. **Signup → Activation:** User signs up, verifies email, lands on dashboard
> 2. **Core Action:** User `[describe your core action — e.g., creates a project, sends a campaign]`
> 3. **Payment:** User upgrades to paid plan via Stripe Checkout
>
> For each flow:
> - Use `@playwright/test` syntax
> - Include proper selectors (prefer data-testid attributes)
> - Handle loading states and async operations
> - Take a screenshot on failure
>
> Also generate a GitHub Actions CI workflow that runs these tests on every PR.

### P15 · Landing Page Copy (Full Page)

**Tool:** Claude · **Output:** Complete page content

> Write the complete landing page copy for `[product name]`.
>
> **ICP:** `[who it's for]`
> **Core benefit:** `[what it does for them]`
> **Price:** `[$X/mo]`
> **Tone:** `[professional but human / bold and direct / warm and approachable]`
>
> Generate these sections in order:
> 1. **Hero:** Headline (under 10 words) + subheadline (under 25 words) + CTA button text
> 2. **Social proof bar:** 3 metrics (e.g., "500+ teams", "4.9/5 rating", "2M+ tasks completed")
> 3. **Problem section:** "You're probably..." (3 pain points)
> 4. **Solution section:** "Here's how [Product] fixes this" (3 features with benefits)
> 5. **How it works:** 3 steps with icons
> 6. **Testimonials:** 3 fictional but realistic quotes from ICP personas
> 7. **Pricing:** Feature comparison table for 2-3 tiers
> 8. **FAQ:** 6 questions and answers
> 9. **Final CTA:** Urgency-driven closing with button
>
> Write 3 alternative headlines so I can A/B test.

### P16 · Launch Day Choreography

**Tool:** Claude · **Output:** Hour-by-hour schedule

> Plan my launch day for `[product]`. I'm launching on Product Hunt.
>
> Create an hour-by-hour schedule from 12:01 AM PT to 11:59 PM PT covering:
> - Product Hunt posting (optimal time, title, tagline, first comment)
> - Hacker News "Show HN" post (title + timing)
> - Twitter/X thread (full 5-tweet thread, written out)
> - LinkedIn post (full text)
> - Reddit posts (which subreddits + post text for each)
> - Email to waitlist (subject + body)
> - DMs to 20 supporters asking for upvotes (template)
>
> Include: what to reply when someone comments, how to handle negative feedback, and when to go to sleep.

### P17 · SEO Content Article

**Tool:** Claude · **Output:** 1500-word blog post

> Write a 1500-word SEO article for my SaaS blog.
>
> **Primary keyword:** `[keyword]`
> **Secondary keywords:** `[2-3 related keywords]`
> **Target audience:** `[ICP]`
> **Goal:** Rank on page 1 for `[keyword]` and drive signups
>
> Structure:
> - H1 title (include primary keyword naturally)
> - Meta description (under 155 characters)
> - Introduction (hook + problem + promise)
> - 3-4 H2 sections with actionable content
> - Each section: explanation + real example + practical takeaway
> - Conclusion with CTA to try my product
> - Internal link suggestions to other pages on my site
>
> Tone: Expert but accessible. No fluff. Every paragraph must teach something.

### P18 · Build-in-Public Weekly Update

**Tool:** Claude · **Output:** Twitter thread + newsletter

> Draft my weekly build-in-public update.
>
> **What I shipped this week:** `[list features/fixes]`
> **Key metric:** `[e.g., "Hit 50 users" or "$500 MRR"]`
> **Biggest lesson:** `[what you learned]`
> **Next week's focus:** `[what you're building next]`
>
> Generate:
> 1. A 3-tweet thread (specific, not promotional, includes one screenshot description)
> 2. A 200-word newsletter update for subscribers
> 3. A LinkedIn post version
>
> Tone: Honest, specific, humble. Show the numbers, even if they're small.

---

## STAGE 4: GROW & RETAIN (Prompts 19–24)

### P19 · Onboarding Email Sequence (5 Emails)

**Tool:** Claude · **Output:** 5 email templates

> Write a 5-email onboarding sequence for new `[product]` users.
>
> **Activation metric:** `[the ONE thing they must do — e.g., "create their first project"]`
> **Timeline:** Emails sent on Day 0, Day 1, Day 3, Day 7, Day 14
>
> For each email: subject line, body (under 150 words), one clear CTA.
>
> Email 1 (Day 0): Welcome + quickstart (get them to activation in 5 minutes)
> Email 2 (Day 1): Feature highlight (the one feature that makes them go "wow")
> Email 3 (Day 3): Social proof (how others are using it)
> Email 4 (Day 7): Check-in (are they stuck? Offer help)
> Email 5 (Day 14): Upgrade nudge (if on free plan) or feedback ask (if paid)

### P20 · Churn Analysis & Exit Interview

**Tool:** Claude · **Output:** Email + analysis framework

> A user just cancelled their `[product]` subscription.
>
> Draft a churn exit email that asks 3 questions:
> 1. What were you hoping `[product]` would do for you?
> 2. What are you switching to?
> 3. Would `[specific feature you're building]` have changed your mind?
>
> Tone: Warm, not defensive. Make them feel like they're helping you, not being interrogated.
>
> Also create a Google Sheets tracker template with columns: Date, User, Reason Category (price/features/competitor/no longer needed/other), Verbatim Quote, Follow-Up Action.

### P21 · Pricing Page A/B Test

**Tool:** Claude · **Output:** 2 pricing page variants

> Design two pricing page variants for `[product]` to A/B test.
>
> **Current price:** `[$X/mo]`
> **Current conversion rate:** `[X%]` (or "unknown, first attempt")
>
> Variant A: Value-anchored (show the cost of NOT using your product, then your price feels cheap)
> Variant B: Competitor-anchored (show competitor prices side by side, you're the better deal)
>
> For each variant: headline, pricing table layout, feature list, CTA text, and the psychological principle being used.

### P22 · Weekly KPI Dashboard Prompt

**Use with:** Any AI with code execution · **Output:** Analysis

> Analyze my SaaS metrics for this week.
>
> `[Paste your data: signups, activations, revenue, churn, support tickets]`
>
> Calculate:
> - Week-over-week growth rate for each metric
> - Activation rate (signups → activated)
> - Revenue per user
> - Quick ratio (new MRR + expansion MRR) / (churned MRR + contraction MRR)
>
> Flag anything that's trending in the wrong direction. Suggest one action item for next week.

### P23 · Feature Request Prioritization

**Tool:** Claude · **Output:** Prioritized backlog

> I have these feature requests from users: `[paste list of requests]`
>
> Prioritize them using the ICE framework:
> - **Impact** (1-10): How much will this move the needle on retention/revenue?
> - **Confidence** (1-10): How sure are we that users actually want this?
> - **Ease** (1-10): How easy is this to build solo in under a week?
>
> Score = Impact × Confidence × Ease
>
> Output a ranked table. For the top 3: write a one-paragraph spec I can hand to Cursor.

### P24 · Quarterly Roadmap Update

**Tool:** Claude · **Output:** Public roadmap

> Draft my quarterly public roadmap update.
>
> **Last quarter's wins:** `[what shipped]`
> **Key metric changes:** `[before → after]`
> **User feedback themes:** `[top 3 requests]`
>
> Generate:
> 1. **Recently Shipped** — last 90 days, with one-line descriptions
> 2. **Now Building** — current quarter's 3 initiatives
> 3. **Up Next** — following quarter's candidates
> 4. **Not Doing (and why)** — 2-3 things you're deliberately NOT building
>
> Format as a clean markdown page I can publish on my website.

---

## STAGE 5: SCALE & EVOLVE (Prompts 25–30)

### P25 · Annual Review: Pivot or Persist?

**Tool:** Claude · **Output:** 1-page scorecard

> Given my 12 months of data, help me decide: persist, pivot the ICP, or pivot the product.
>
> **Month 1 metrics:** `[MRR, users, churn]`
> **Month 12 metrics:** `[MRR, users, churn]`
> **Best-performing ICP segment:** `[who retained best]`
> **Top churn reason:** `[most common exit reason]`
>
> Analyze: Is the retention curve improving? Is there a segment that loves the product? Am I growing fast enough to hit $10k MRR in the next 6 months?
>
> Output a scorecard with a clear PERSIST / PIVOT ICP / PIVOT PRODUCT recommendation.

### P26 · Hire Your First Contractor

**Tool:** Claude · **Output:** Job description + evaluation

> I need to hire a part-time contractor to help with `[area — e.g., frontend, customer support, content]`.
>
> Generate:
> 1. A job description for Upwork/Contra (under 300 words)
> 2. A 3-question screening test I can send before interviewing
> 3. A 1-hour paid trial task to evaluate their actual skill
> 4. A scoring rubric (1-5) for evaluating submissions
>
> Budget: `[$X/hr]`. Location preference: `[any / timezone-specific]`.

### P27 · Prepare for Your First Investor Conversation

**Tool:** Claude · **Output:** Pitch prep doc

> I might have an informal conversation with an angel investor about `[product]`.
>
> Prepare me with:
> 1. **30-second elevator pitch** (problem → solution → traction → ask)
> 2. **Key metrics to share** (and what each one signals)
> 3. **The 5 questions they WILL ask** (and my best answers)
> 4. **Red flags to avoid** (things that kill investor interest instantly)
> 5. **My ask** (if I were to raise, what would be reasonable at my stage?)
>
> Current traction: `[MRR, users, growth rate]`

### P28 · Create Your CLAUDE.md / .cursorrules File

**Tool:** Claude · **Output:** Project config file

> Generate a CLAUDE.md file for my SaaS codebase that any AI coding assistant can use as context.
>
> **Product:** `[name and what it does]`
> **Stack:** `[framework, DB, auth, payments]`
> **Project structure:** `[paste your folder structure]`
> **Coding conventions:**
> - `[e.g., All components under 300 lines]`
> - `[e.g., Multi-tenant: all queries filter by workspaceId]`
> - `[e.g., Use server actions, not API routes, for mutations]`
>
> The file should include: project overview, tech stack, folder structure, naming conventions, database rules, common patterns (with code examples), and things to NEVER do.

### P29 · Automate Your Weekly Reporting

**Tool:** Claude · **Output:** Script or Zapier workflow

> Design an automated weekly reporting system for my solo SaaS.
>
> Every Monday at 9 AM, I want an email/Slack message with:
> - New signups this week (from `[DB/Stripe]`)
> - MRR change (from Stripe)
> - Churn events (from Stripe)
> - Top 3 feature requests (from `[support tool]`)
> - Uptime % (from `[monitoring tool]`)
>
> Suggest the simplest implementation: either a Next.js cron job, a Zapier/Make workflow, or a standalone script. Include the code or step-by-step setup.

### P30 · Version Control Your Business

**Tool:** Claude · **Output:** Business changelog

> Create a "business changelog" template that I update monthly.
>
> For each month, track:
> - **Product:** What shipped, what was cut, what broke
> - **Revenue:** MRR, new customers, churned customers, net revenue retention
> - **Marketing:** Best-performing channel, content published, backlinks earned
> - **Operations:** Hours worked, biggest time sink, what I automated
> - **Lessons:** One thing I'd do differently, one thing that worked surprisingly well
>
> Generate the template and fill in Month 1 as an example using this data: `[paste your current metrics]`

---

## Pro Tips for Getting Better AI Outputs

1. **Be specific with inputs.** "A project management tool" → bad. "A Kanban-style project tracker for 2-5 person design agencies who currently use Trello but find it too generic" → good.

2. **Always specify the output format.** "Give me a table" or "Output as a Prisma schema" forces structured, usable results.

3. **Use the "Act as" prefix** for specialized tasks. "Act as a senior Stripe integration engineer" produces more precise code than a generic request.

4. **Chain prompts.** Use P2's ICP output as input for P3's competitor research. Use P7's stack choice as input for P8's schema. Each prompt builds on the last.

5. **Save what works.** When a prompt produces something you ship, save the exact prompt + output to a `prompts/` folder in your repo. It becomes a reusable company asset.

---

*The Solo SaaS Playbook* — by Abhishek Verma — PlanMySaaS Press · 2026
Read more at <https://www.planmysaas.com/book>

© 2026 Abhishek Verma. Personal use only. Do not redistribute.
