Before I clone a funnel, pitch a client, or launch a product—I run this prompt.
Because if you’re not doing deep research on a company before you try to out-market, out-price, or out-position them… you’re guessing.

Most people rely on surface-level Google searches or weak AI prompts like “Tell me about Company X.”
That’s not research. That’s noise.
This post gives you a surgical prompt framework that turns ChatGPT into your private analyst—digging through company data, uncovering strategy gaps, and handing you insights that are actually useful.
If you’ve ever Googled ‘how to research competitors with ChatGPT’—this is the guide you actually need.
Whether you’re building a competing product, training a custom GPT, or just want to understand how a business wins…
This prompt will get you the full picture.
Let’s open it up.
What Most People Do (And Why It Sucks)
Most people treat ChatGPT like a talking Google search bar because they don’t know how to research competitors with ChatGPT the right way.
They type something like “Tell me about Tesla” and expect deep, actionable insights.
What they get is a mashed-up Wikipedia summary… bland, outdated, and useless.
Even worse—when the goal is serious (like building a competitor, writing a sales pitch, or training a GPT)…
That kind of prompt isn’t just lazy. It’s dangerous.
Bad research leads to:
- Weak strategies built on fluff
- Funnels that miss the mark
- Campaigns that feel like guesswork
Most prompts are vague, unstructured, and missing the most important thing: → The why behind the research.
ChatGPT isn’t magic. It needs direction. It needs structure.
And if you give it the right setup, it becomes a strategic weapon.
That’s where this Prompt Surgeon framework comes in.
Whether you’re launching a course, funnel, or custom GPT, this is how to research competitors with ChatGPT at a strategic level.
The Prompt Breakdown – Why Each Part Matters
This isn’t just a prompt.
It’s a modular intelligence-gathering system you run inside ChatGPT.
Below is the full structure—each line does heavy lifting.
Let’s break it down and show how to make it yours.
🔧 Prompt Header: The Setup
“I want you to carry out deep research on the following company using the information I have given you. My goal for using this information is to <insert reason here>.”
Why it matters:
This primes ChatGPT with intent—which sets the tone for everything that follows.
You’re telling it this is not a casual ask. It’s a mission with a purpose.
Tips:
- Be blunt about the goal. Don’t say “understand the brand.” Say “I want to reverse engineer their funnel and beat them.”
- This context calibrates what ChatGPT looks for and how it responds.
🧭 Context Clarifier
“For context, <this is a company I own – a competitors company – a company I have a general interest in – a potential clients company – a clients company>”
Why it matters:
This one line can radically change the depth and lens of the research.
- If it’s your company → ChatGPT will focus on optimisation and self-assessment.
- If it’s a competitor → It becomes a teardown.
- If it’s a client → It will look for gaps, offers, and positioning angles.
Pro move: You can also say things like “This is a competitor I want to dominate in SEO” or “This is a SaaS company I want to model my funnel after.”
You can also cross-check SEO strategies with tools like Ahrefs or SimilarWeb
🗂️ Core Inputs
1. Name of the business or company
(Full legal name, and any variations it might be referred to as.)
Why: Avoids confusion with similarly named brands. Gets accurate data.
2. Website URL
(Main domain, and any other relevant sites or landing pages you want to include.)
Why: ChatGPT will scan site content if possible, and it uses this to cross-check tone, offers, structure, etc.
3. The kind of info I am looking to extract includes:
(delete/edit/change these as you see fit)
- Company history and background
- Products or services
- Market position and competitors
- Target audience
- Pricing strategy
- Customer reviews and public sentiment
- Social media presence and strategy
- Key team members or founder story
- Tech stack (if relevant)
- SEO / content strategy
- Paid ads or funnels
- Affiliates, partners, or referral model
- Any red flags or controversies
- Anything else?
Why:
This turns your prompt into a research brief.
It’s no longer “Tell me about this company.”
It’s “Here’s what I want. Go dig.”
Power move: Add or remove categories based on your objective.
For example:
- If building a funnel → Focus on offers, pricing, content, ads
- If pitching them → Focus on gaps, team, and what’s missing
🧨 Research Outcome
“Once the information is gathered I plan to <Train a GPT on it – Build a marketing funnel – Create a competing product – Create a teardown blog post – Use it to reverse-engineer their playbook>”
Why it matters:
This tells ChatGPT what to emphasise, what to ignore, and how to present the intel.
You’re giving the AI purpose, which always sharpens the output.
🧠 Output Style
“Tone of the output I want is <Dry and factual – Tactical and analytical – Like a spy report / teardown – Like a brand strategist – Like a startup mentor>”
Why it matters:
Tone shapes how information is framed and interpreted.
Same facts—different angle = different strategic value.
Examples:
- “Like a spy report” = gritty, red flags, leaks, threats
- “Like a startup mentor” = advice, gaps, growth areas
- “Tactical and analytical” = raw data + interpretation
This framework shows you exactly how to research competitors with ChatGPT—step by step
Real-World Use Cases – How to Deploy This Prompt Like a Pro
This isn’t a one-trick prompt.
It’s a research weapon that adapts to your goal.
Here’s how different pros would use it—along with the exact inputs they’d tweak. Remember that you must include the core inputs in your prompt such as the company name, URLs and the types of information you want (as given in the Core Inputs section above)
🔧 1. The Solo Founder Building a Competing Offer
“I want to carry out deep research on the following company. My goal is to reverse-engineer their entire offer stack and pricing model to build something better.”
Context: This is a competitor’s company
Tone: Tactical and analytical
Planned outcome: Build a competing product + funnel
This use case focuses on:
- Product breakdown
- Price psychology
- Funnel steps
- Copy positioning
- Gaps to exploit
🔧 2. The Agency Doing a Client Teardown
“I want to carry out deep research on the following company. My goal is to identify weaknesses and opportunities so I can pitch them a high-value marketing strategy.”
Context: This is a potential client’s company
Tone: Like a brand strategist
Planned outcome: Client pitch or proposal
Focus areas:
- SEO and ad strategy
- Brand tone + positioning
- Public perception (reviews, socials)
- Competitor comparison
- Funnel gaps
🔧 3. The Course Creator Studying a Top Seller
“I want to carry out deep research on the following company. My goal is to model their course structure, pricing, and sales page strategy for my own info product.”
Context: Competitor’s company
Tone: Like a startup mentor
Planned outcome: Model and optimise a course offer
Focus areas:
- Offer breakdown
- Course structure
- Pricing tiers
- Copy style
- Affiliate strategy
🔧 4. The Operator Training a Custom GPT
“I want to carry out deep research on the following company. My goal is to feed this data into a custom GPT that understands how they operate.”
Context: A client’s company
Tone: Dry and factual
Planned outcome: Train a client-facing AI assistant
Focus areas:
- Company background
- Team + founder info
- Products, services, and FAQs
- Tone of voice from site and social
- Internal processes (where possible)
💡 Pro Tip:
You can also run multiple versions of this prompt on the same company—each with a different angle or tone. That’s how you extract insights from all sides.
Want to Go Deeper? Precision Engineer It with ChatGPT Pro
You’ve seen the base prompt.
But if you’re serious about turning ChatGPT into a research-grade intelligence tool, here’s how to precision engineer your workflow—with browsing, memory, and strategic follow-ups baked in.
🕸️ 1. Feed It URLs—Then Layer Follow-Ups
You don’t need to copy-paste site content. Just give ChatGPT the URL and prompt it like this:
“Browse this homepage and summarise it like a conversion copywriter reverse-engineering the offer.”
“Analyse this About page and pull out founder positioning, tone of voice, and hidden trust signals.”
“Compare this pricing page to 3 competitors—what’s the strategic angle?”
You’re not reading pages. You’re running tactical AI audits.
🕵️ 2. Run SEO and Funnel Teardowns
Prompt ChatGPT to act like a specialist:
“You are an SEO strategist. Audit this company’s blog, organic structure, and keyword focus.”
“Now act as a funnel hacker. Break down their lead magnet, tripwire, and upsell flow.”
Each lens reveals different strategic levers.
Stack them to build a full-spectrum breakdown.
🤖 3. Train a Custom GPT on Their Strategy
Use the research to build your own competitor clone GPT.
System prompt example:
“You are an AI trained on <CompanyName>. You know how they think, market, and operate. Your job is to act as a brand consultant for anyone wanting to understand or outperform them.”
Boom—you now have an AI that thinks like your competitor.
You can simulate decisions. Model product tweaks. Even A/B test offers against their voice.
📊 4. Pull External Data from the Wild
With browsing enabled, ChatGPT becomes a sentiment analyst + data miner.
Ask it to:
- Scrape Trustpilot, Reddit, or Glassdoor for public sentiment
- Analyse YouTube interviews or podcast transcripts
- Extract ad copy from Facebook Ad Library or Google Ads
- Review LinkedIn profiles for team structure and hiring patterns
Then prompt:
“Based on all this, what patterns or positioning angles stand out?”
“What does this company understand about their audience that others don’t?”
This is high-level competitor forensics.
🧠 5. Create a Strategic Summary
Once you’ve pulled the data, wrap with:
“Summarise this company like a brand strategist preparing a pitch. What should I learn, copy, or avoid?”
You now have actionable intel, not just info.
This is what Precision Engineered looks like.
The Final Prompt Template
Copy it. Edit the < >. Run it inside ChatGPT Pro with ‘Deep Research’ selected. Watch what happens.
How to Research Competitors with ChatGPT: Full Prompt Breakdown
I want you to carry out deep research on the following company using the information I have given you. My goal for using this information is to <insert your reason here>.
For context, <this is a company I own – a competitor’s company – a company I have a general interest in – a potential client’s company – a client’s company>.
1. Name of the business or company:
<Insert full legal name and any variations it's known by>
2. Website URL(s):
<Insert main domain, and any other relevant URLs or landing pages>
3. The kind of information I am looking to extract includes:
(delete/edit/add to this list as needed)
- Company history and background
- Products or services
- Market position and competitors
- Target audience
- Pricing strategy
- Customer reviews and public sentiment
- Social media presence and strategy
- Key team members or founder story
- Tech stack (if relevant)
- SEO / content strategy
- Paid ads or funnel structure
- Affiliates, partners, or referral model
- Any red flags, controversies, or legal issues
- Anything else that would be strategically useful
4. Once the information is gathered, I plan to:
<Train a GPT on it – Build a marketing funnel – Create a competing product – Create a teardown blog post – Use it to reverse-engineer their playbook – Pitch them as a client>
5. The tone I want the output delivered in is:
<Dry and factual – Tactical and analytical – Like a spy report – Like a brand strategist – Like a startup mentor>
💡 Tip: Save this as a reusable prompt. Clone it every time you research a new company—and tweak the outputs based on what you’re building.
This prompt is designed for ChatGPT Pro users with browsing enabled using the ‘Deep Research’ option.
The Prompt Evolution Breakdown
From Vague to Precision Engineered—how we transform a throwaway prompt into a strategic weapon.
Most people prompt like tourists.
They type vague nonsense like:
“Tell me about Nike.”
…and expect to get something useful.
Below is what happens when you run that kind of prompt through the Prompt Evolution System from Prompt Surgeon.
🧪 LEVEL 1 – VAGUE
“Tell me about Nike.”
🧼 Output:
Generic summary. Repeats what you’d find on the first page of Google or Wikipedia.
✅ Feels like: “Cool facts.”
❌ Useless for strategy, action, or marketing.
🔧 LEVEL 2 – BASIC
“Can you give me a detailed overview of Nike as a company?”
🧼 Output:
Slightly more structured. Touches on products, history, market, revenue.
✅ Feels like: A blog intro.
❌ Still no depth, no lens, no strategic focus.
⚙️ LEVEL 3 – IMPROVED
“Can you analyse Nike’s business—specifically their positioning, pricing strategy, and customer audience?”
🧼 Output:
Now we’re getting insights. ChatGPT digs into specific components.
✅ Useful if you know how to interpret.
❌ Still lacking context, purpose, and tone.
🧠 LEVEL 4 – REFINED
“I want you to analyse Nike as if you’re a brand strategist helping a startup model their business after Nike. Focus on market positioning, tone of voice, product strategy, and brand story. Be analytical.”
🧼 Output:
ChatGPT now acts like a strategist. The response is filtered, insightful, and tailored.
✅ Clear angle of attack.
❌ Still missing structure and modularity.
🧠💀 LEVEL 5 – PRECISION ENGINEERED
The full modular prompt you just built.
With custom inputs. Browsing enabled. Strategic outcome defined.
Structured into:
- Purpose
- Context
- Research Categories
- Output Use Case
- Desired Tone
🧼 Output:
Detailed. Tactical. Categorised. Context-aware. Ready to use.
✅ Feels like: A teardown written by a consulting team.
🎯 THIS is what Prompt Surgeon does.
🧠 Final Word
“ChatGPT doesn’t find what you’re looking for.
It finds what you asked for.”
Structure your ask, and the AI becomes your intelligence team.
⚡ Ready to prompt like this every time?
Grab the Prompt Surgeon Playbook
and get the full system, teardowns, custom GPT access, and strategic walkthroughs.