Build a Competitor Analysis with AI

Traditional competitor analysis works like this: someone spends three hours opening tabs, copying pricing pages, reading G2 reviews, and eventually producing a spreadsheet that is outdated by the time it lands in a Slack channel. The insights are buried. The recommendations section is usually “we should monitor this closely.”

There is a better version. Gather the raw material yourself, paste it into a prompt, and get back a structured briefing with a positioning map, head-to-head comparisons, and actual strategic recommendations. The research still takes time. The synthesis takes thirty seconds.

Getting the Input Ready

The quality of the output depends almost entirely on the quality of the input. This prompt will not browse the web for you. It works with what you give it.

Here is what to collect for each competitor before you run the prompt:

• Their website homepage and pricing page (copy the text, or paste the URL if you are working in a browser-integrated tool)
• Any product descriptions or feature lists
• A handful of customer reviews from G2, Capterra, Trustpilot, or app stores
• Their positioning language: the headline on their homepage, how they describe themselves on LinkedIn
• Pricing, if it is publicly listed

For your own business, write two to four sentences: what you sell, who buys it, what it costs, and what you believe makes you different. Be direct. If your differentiator is vague in the input, the output will be vague too.

Three to five competitors is the right range for this prompt. Fewer and the positioning map has nothing to compare. More and the document becomes unwieldy.

The Prompt

With your business description and competitor research collected:

You are a senior market analyst. I will give you information about my business and a set of competitors. Produce a structured competitor analysis.

Output format:

## Competitive Analysis: [My Company] vs. Market

**Date:** [today's date]

### Overview Table
| Competitor | Target Customer | Core Value Prop | Pricing Model | Strengths | Weaknesses |
|---|---|---|---|---|---|
(Fill one row per competitor)

### Competitive Positioning Map
Describe (in text, since we can't embed images) where each competitor sits on two axes: Premium vs. Budget and Specialist vs. Generalist.

### Head-to-Head: How We Compare
For each competitor, write 3-5 bullet points comparing them directly to my business.

### Market Gaps & Opportunities
Based on this analysis, what underserved needs or positioning gaps exist that my business could own?

### Strategic Recommendations
3-5 concrete actions I should consider based on this analysis.

Rules:
- Only use information I provide; do not invent pricing, features, or claims
- Flag with [VERIFY] any inference you make that I should confirm
- Keep language direct and specific -- no generic consulting filler

---
MY BUSINESS:
[Describe your company: what you sell, who you serve, your pricing, your key differentiators]

---
COMPETITORS:
[For each competitor, paste: their name, website, pricing if known, key features or messaging, any reviews or feedback you've seen]

The [VERIFY] rule is the one that matters most. Claude will sometimes draw reasonable inferences from the data you provide. The flag makes those inferences visible so you can check them rather than accidentally presenting them as confirmed facts in a board meeting.

What the Output Looks Like

The analysis arrives as a structured document with four sections.

The Overview Table gives you one row per competitor: target customer, core value proposition, pricing model, a strengths column, a weaknesses column. It is designed to be skimmable. Print it. Put it on the wall. It is the map.

The Competitive Positioning Map describes in prose where each competitor sits on two axes: Premium versus Budget, and Specialist versus Generalist. Text-based positioning maps feel awkward at first, but the descriptions are often more precise than a two-by-two grid with a dot on it.

The Head-to-Head section is the most useful part for sales teams. Three to five bullet points per competitor, comparing them directly to your business. These are the points your sales reps are probably already making in calls; seeing them written out clearly helps you sharpen the language.

The Market Gaps section is where the prompt earns its time savings. After reading all the competitor data you pasted, Claude identifies where the market is underserved or where no one has planted a flag. These are starting points for conversation, not conclusions, but they are usually more specific than what comes out of a two-hour strategy meeting.

The Strategic Recommendations section gives you three to five concrete actions. Not “monitor competitors” actions. Specific moves: “Your two mid-market competitors both price on a per-seat model; a flat-rate tier would differentiate you with buyers who have unpredictable team sizes.”

Verifying Before You Share

After the analysis generates, do two things before forwarding it to anyone.

First, read every [VERIFY] flag. These are inferences Claude made from the data. Some will be obviously correct; verify them anyway if they are going into a presentation. A few may be wrong. Better to catch them here than in a meeting.

Second, spot-check the pricing data. Pricing pages change constantly. If a competitor’s price in your output does not match what is on their site today, update the input and run the prompt again. The whole thing takes two minutes to regenerate.

Keeping It Current

Competitive landscapes shift. A competitor gets acquired, reprices, or launches a new product tier, and last quarter’s analysis becomes misleading. Because this workflow is fast, maintaining a current analysis is actually practical.

Keep the input file somewhere accessible. Each quarter, update the competitor data, add any new entrants, and run the prompt again. The document regenerates in under a minute. Compare it to the previous quarter’s output and you have a clear picture of what changed and what held steady.

That is more competitive intelligence than most small teams produce in a year, and it costs about an hour per quarter of actual human time.