Write a Competitive Analysis Section for a PRD
Produces a structured competitive analysis comparing how 3-5 competitors solve the same problem, with strengths, gaps, and differentiation opportunities. Designed for PMs drafting PRDs or strategic briefs.
When to use this prompt
Use this when drafting a PRD for a feature that enters an established product category where competitors already have solutions. You will need a list of competitors and a brief description of the user problem you are solving. The prompt does not replace primary research; it synthesizes what you already know or can quickly verify. It is best used after you have done 30-60 minutes of competitor product exploration yourself, so the model can structure your observations rather than invent them. If you provide no real competitor info, it will produce plausible-sounding but unreliable output.
The Prompt
You are a product manager writing the competitive analysis section of a PRD. Your job is to structure what we already know about how 3-5 competitors solve the same user problem, and recommend a differentiation angle. User problem: {{user_problem}} Competitors to analyze: {{competitors}} Our product position: {{our_position}} Known competitor info: {{known_info}} Produce the output in this exact structure: 1. COMPETITOR OVERVIEW TABLE â 5 columns: Competitor, Approach summary (1 sentence), Strengths (2-3 bullets), Gaps (2-3 bullets), User segment best served. 2. COMMON PATTERNS â 3-5 sentences describing what all or most competitors do similarly. These are table stakes we must match or we lose. 3. UNDIFFERENTIATED GAPS â Things no competitor does well. Rank each gap by (a) likely impact on users and (b) defensibility (how hard is it for competitors to copy quickly). 4. DIFFERENTIATION RECOMMENDATION â One specific angle where we should invest. Justify with reference to the gaps and our stated position. Explain why this is defensible for 12+ months. 5. CONFIDENCE LEVEL â Rate each claim in the analysis as HIGH (primary evidence or public documentation), MEDIUM (secondary sources or recent customer conversations), or LOW (inference or assumption). Be honest about how much you do not know. Do not fabricate competitor features. If you lack information on a competitor, mark sections as UNKNOWN rather than guessing.
Example Output
OVERVIEW TABLE | Competitor | Approach | Strengths | Gaps | Best for | | AcmeCo | In-product dashboard with manual export | Polished UI; fast load | No automation; weak mobile | Large teams | | BetaTool | Automated weekly email reports | Low-friction setup; good defaults | No customization; rigid cadence | Small teams | | GammaApp | Full BI integration | Powerful; integrates with Looker | High setup cost; long TTV | Enterprise | COMMON PATTERNS All three offer weekly summary views and basic filtering. None offer real-time alerts or natural-language query. Mobile experiences are uniformly weak. Everyone requires admin setup before end users can benefit. UNDIFFERENTIATED GAPS 1. Real-time alerts on threshold events (high impact, medium defensibility). 2. Mobile-first consumption (high impact, high defensibility if done well). 3. Natural language query over saved views (high impact, low defensibility; AI commodity). DIFFERENTIATION RECOMMENDATION Invest in mobile-first consumption with push alerts on threshold events. This combines 2 undifferentiated gaps, is defensible because none of the incumbents have mobile engineering muscle, and fits our stated position as the mobile-native alternative. CONFIDENCE LEVEL - AcmeCo approach: HIGH (public docs and a demo I watched last week). - BetaTool gaps: MEDIUM (based on 2 customer switch interviews). - GammaApp best-for: LOW (inference from pricing page, no first-hand experience).
Recommended Tools
Notion and Coda both support the tabular comparison format cleanly and allow inline comments from stakeholders reviewing the analysis. Airtable works when you maintain an ongoing competitive intelligence database that updates over time. Notion is best for one-shot PRD sections; Airtable is best when you want the competitor data to persist and evolve across multiple PRDs and quarterly reviews.
Frequently Asked Questions
When should I use this prompt?
Use it when entering a category where 3+ competitors already have products, and when your PRD readers include execs or investors who will ask 'what makes us different?' Skip it for greenfield innovation where no competitor exists; inventing a competitive set in that case wastes time. Run it after doing 30-60 minutes of real competitor research yourself (watching demos, reading changelogs, reading G2 reviews), so the model has substance to work with rather than guesses.
How do I avoid hallucinated competitor details?
Always provide known competitor info in the prompt input, and use the CONFIDENCE LEVEL section to call out inference versus evidence. Re-read the output and challenge anything marked HIGH that does not match your own research. The prompt will mark sections UNKNOWN when you do not provide info, but it can still drift when generating strengths and gaps from sparse input. A good rule: if you cannot personally point to the source of a claim within 60 seconds, mark it LOW and verify before publishing the PRD.