Why use a structured prompt instead of just asking for a comparison?

A free-form ask produces uneven coverage and confident-sounding guesses. A structured prompt fixes the dimensions, demands the same fields for every competitor, and requires source citations, so the output is comparable and easier to fact-check.

Will the LLM's competitor facts be accurate?

Not automatically. Models can hallucinate pricing and features, especially for niche or recent products. That is why the prompt requires citations and a confidence flag — treat any uncited claim as a lead to verify, not a fact.

Does this tool do the research itself?

No. It builds the prompt only. You run that prompt in your own LLM or a research-enabled assistant. The builder runs entirely in your browser with no API key.

Can I add my own comparison dimensions?

Yes. Use the custom dimensions field to add anything specific to your market — onboarding time, integrations, compliance certifications — and they are woven into the prompt alongside the presets.

What output format works best?

A comparison table is best for side-by-side scanning and slide decks. Per-competitor briefs suit deeper qualitative analysis. The narrative summary is useful when you need a written recommendation rather than raw data.

Competitor Analysis Prompt Builder

Competitor analysis prompt builder

Asking an LLM “compare us to our competitors” gives you a vague, uneven answer. This builder turns your inputs — your product, a list of rivals, and the dimensions you care about — into a structured competitor-analysis prompt that forces the model to evaluate every competitor on the same axes, return a clean comparison format, cite its sources, and flag anything it is unsure about. The result is research you can actually act on and verify.

How it works

You describe your product, paste competitor names one per line, and select the comparison dimensions (pricing, features, positioning, weaknesses) plus any custom axes. You also choose an output format. The builder assembles a prompt that sets the model’s role as a market analyst, lists the competitors and dimensions explicitly, requires identical fields for each competitor, demands a source or a “no reliable source” note per claim, and asks for a final summary with a positioning recommendation. Everything runs in your browser; you copy the prompt and run it wherever you like.

Tips and notes

Prefer research-enabled models. Citations are far more reliable when the assistant can browse or search rather than recall from training data.
Verify uncited claims. Treat any fact without a source as a lead to check, especially pricing and feature availability.
Keep the competitor list tight. Three to six rivals produces sharper, more comparable output than a long list.
Add custom dimensions. Market-specific axes like compliance or integrations often matter more than generic feature counts.