> ## Documentation Index > Fetch the complete documentation index at: https://dhiwisepvtltd-5e04e6a4.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # Prompting for Solve > Prompting guide for Rocket.new Solve: frame strategic questions, scope research, and structure inputs so you get sharper business and market reports. Solve is only as good as the question you ask it. This guide teaches you how to frame strategic research questions that produce structured, actionable reports - not generic overviews. ## Anatomy of a good Solve prompt Great Solve prompts share four qualities: 1. **Specific scope** - a defined topic, market, or set of companies 2. **Clear deliverable** - what format or structure you want in the answer 3. **Context** - enough background for Rocket to understand your situation 4. **Constraints** - time frames, geographies, segments, or other boundaries **Good prompt** ```plaintext wrap theme={null} Compare Notion, Coda, and Slite for small team collaboration. For each, analyze: pricing tiers and limits, core features (docs, databases, wikis), integrations with Slack and Google Workspace, and user sentiment on G2 and Reddit. Which one is best for a 10-person startup that needs docs + lightweight project tracking? Include a recommendation with trade-offs. ``` **Why it works:** Names specific products, lists specific comparison criteria, defines the user's context (10-person startup), and asks for a clear deliverable (recommendation with trade-offs). **Weak prompt** ```plaintext wrap theme={null} Tell me about collaboration tools. ``` **Why it's weak:** No specific products, no comparison criteria, no context about who's asking or why. The result will be a broad, generic overview that doesn't help you make a decision. ## Prompt frameworks for common analysis types ### Market sizing Include the product category, geography, time frame, and what breakdown you want. **Good prompt** ```plaintext wrap theme={null} What is the total addressable market for AI-powered code review tools in North America? Break down by segment: enterprise (1000+ employees), mid-market (100-999), and SMB (under 100). Include 2024 market size and projected CAGR through 2028. Identify the top 5 vendors and their estimated revenue. ``` **Weak prompt** ```plaintext wrap theme={null} How big is the AI code review market? ``` ### Competitive analysis Name the competitors, specify what dimensions to compare, and describe your vantage point. **Good prompt** ```plaintext wrap theme={null} I'm building a scheduling tool for healthcare clinics. Compare Calendly, Acuity Scheduling, and Zocdoc on: features specific to healthcare (HIPAA compliance, intake forms, insurance verification), pricing for small practices (1-5 providers), patient experience (booking flow, reminders, telehealth), and integration with EHR systems. What gaps could a healthcare-focused alternative fill? ``` **Weak prompt** ```plaintext wrap theme={null} Compare scheduling tools. ``` ### Pricing strategy Describe your product, target audience, and the decision you're trying to make. **Good prompt** ```plaintext wrap theme={null} I'm launching a B2B SaaS product for HR teams at companies with 50-500 employees. My competitors charge $5-15/user/month. Should I use per-seat pricing, flat-rate tiers, or usage-based pricing? Analyze each model's pros and cons for my segment. Include examples of successful B2B HR tools and their pricing approaches. Recommend a pricing structure for launch with reasoning. ``` **Weak prompt** ```plaintext wrap theme={null} What should I charge for my SaaS product? ``` ### Due diligence / investment analysis Provide the company or space, the investment stage, and what you need to evaluate. **Good prompt** ```plaintext wrap theme={null} Build an investment thesis for a Series A company in the vertical SaaS space for restaurants. Cover: market size and growth, competitive landscape (Toast, Square, Clover), typical unit economics at this stage, key risks and mitigants, what metrics a Series A company should demonstrate, and comparable recent funding rounds in the space. ``` **Weak prompt** ```plaintext wrap theme={null} Should I invest in restaurant tech? ``` ## Follow-up patterns The first Solve prompt gets you a broad report. Follow-ups help you drill deeper. Here are effective follow-up patterns: ### Drill down on a section ```plaintext wrap theme={null} Expand on the pricing analysis. Show me a detailed breakdown of each competitor's pricing tiers, what's included at each level, and how they've changed their pricing over the past 2 years. ``` ### Challenge the findings ```plaintext wrap theme={null} This analysis assumes our target customer is a startup. What changes if we target mid-market companies (200-1000 employees) instead? How does the competitive landscape shift? ``` ### Make it actionable ```plaintext wrap theme={null} Based on this report, give me a prioritized action plan. What are the 5 things I should do in the next 30 days to validate this opportunity? ``` ### Cross-reference ```plaintext wrap theme={null} In my previous analysis of [Topic A], you identified [Finding X]. How does that relate to the findings in this report? Are there contradictions or reinforcements? ``` ## Common mistakes > **Too broad - no boundaries** > > "Analyze the fintech market" is too broad to produce useful results. Narrow it: which segment of fintech? Which geography? What decision are you trying to make? > **No context about your situation** > > Solve produces better recommendations when it knows who you are. "I'm a 5-person startup building for small retailers" gives Solve the context to tailor its analysis - not just describe the market in the abstract. > **Asking for everything in one prompt** > > "Give me a full market analysis, competitive teardown, pricing strategy, and go-to-market plan" will produce a shallow overview of each. Break it into focused questions and use follow-ups to go deep. > **Not specifying the deliverable** > > If you want a comparison table, say so. If you want a recommendation, ask for one. If you want data points with sources, request them. Solve will match the output to your request. For additional guidance on question framing, report evaluation, and iterative research strategies, see [Solve best practices](https://dhiwisepvtltd-5e04e6a4.mintlify.site/solve/best-practices.md). ## Quick reference: prompt template Use this template as a starting point for any Solve prompt: ```plaintext wrap theme={null} [What I want to know]: [specific question about a defined topic] Context: [who I am, what I'm building, what decision I'm trying to make] Scope: [specific companies, markets, geographies, or time frames] Deliverable: [comparison table, recommendation with trade-offs, prioritized list, action plan] Constraints: [budget, team size, timeline, or other limiting factors] ``` ## What's next? > **[Configure Intelligence](https://dhiwisepvtltd-5e04e6a4.mintlify.site/learn/guides/configure-track.md)** > > Set up the Intelligence wizard for useful monitoring signals. > **[Prompting for Build](https://dhiwisepvtltd-5e04e6a4.mintlify.site/learn/guides/for-build.md)** > > Describe apps and features for better Build results. > **[Solve overview](https://dhiwisepvtltd-5e04e6a4.mintlify.site/solve/overview.md)** > > Learn how Solve works under the hood. > **[Solve best practices](https://dhiwisepvtltd-5e04e6a4.mintlify.site/solve/best-practices.md)** > > More tips for getting the most from Solve. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.