Build your own AI Search Tool in ChatGPT: Prompts & Reports
Turn ChatGPT into your personal AI search tool and get reports to see how visible your brand is relative to competitors.
AI search is a critical piece of today's marketing strategy. It's why the demand for AI search tools is exploding. Brands want to see how visible they are in LLMs.
The problem with most AI search tools is that they are lying to you. This is a great post from Tim Soulo detailing the why.
It's so hard to replicate the results your users get when they search via AI assistants. Let's demonstrate with ChatGPT.
Let's say we start with
"What is the best CRM for small businesses? Give me just the brand names, no descriptions", run that prompt. Here’s my results.
Now, run the following prompt to turn memory off
"What is the best CRM for small businesses? Give me just the brand names, no descriptions. “Ignore memory and answer only based on your training + browsing.”
Here’s what I get:
Slightly different, let’s run without browsing
"What is the best CRM for small businesses? Give me just the brand names, no descriptions. “Ignore memory and answer only based on your training. Do NOT browse; answer from your internal knowledge only.”
The list I get is below, slightly different again.
HubSpot
Zoho CRM
Pipedrive
Freshsales
Salesforce Essentials
Insightly
Copper
Keap
Nimble
Monday CRM
Lastly, let’s run the first prompt through the API (with web search turned on) as that is what most AI search tools use:
All slightly different. Is that bad? No, you still get a sense of your visibility. But, I do feel overtime memory will mean our results in our own ChatGPT instance will be much different than what you’ll get via the API.
But, you can get a lot of information from ChatGPT without even needing an AI search tool if you know how to prompt. In this tutorial we’re going to create 3 reports for you get a sense of what topics ChatGPT associates with your brand, and how visible your brand is for those ‘topics’ in ChatGPT without you needing an AI search tool.
Report 1: Your Brand’s Visibility – Shows where your brand shows up most strongly in AI (Top-3, Mid, Low, or Missing) for key topics.
Report 2: Market Hotlist –The top 3 brands for each topic, so you can see which areas are dominated by the same players and which ones are still wide open.
Report 3: Competitor Contrast – Compares you to a rival, highlighting what you both share, what only you own, and what only they own.
ChatGPT Brand Visibility Tool
Let’s get started with building your brand visibility reports:
Step 1) Set your scope
Before you run anything, you need two short lists:
Brands (the players): Start with your company, then add 3–5 obvious competitors.
Topics (the playing field): The things you want to be known for. Think: product features, categories, problems you solve, who you serve, or qualities like “easy to use,” “affordable,” “integrates with everything.”
Brands = who’s competing, topics = what they’re competing on.
You can do this manually or you can do it via a prompt. Remember, topics in AI search are the new keywords and they’ll be many more of them.
Here’s a simple prompt you can run in any AI assistant e.g. I ran this in Perplexity to get those details for you.
PROMPT:
You are a marketing analyst.
Focus Brand: [INSERT YOUR BRAND]
Task:
1) Confirm my brand with its website + 1-line description.
2) Suggest 5–7 direct competitors (companies a buyer would compare us to).
3) Suggest 8 topics we should track, mixing:
- Features / capabilities
- Use cases / problems solved
- Buyer types
- Attributes (pricing, ease of use, integrations)
Output format (markdown):
## Brands
- [Brand] — description
- [Competitor 1] — description
- [Competitor 2] — description
...
## Topics (8)
- Topic 1
- Topic 2
- ...
- Topic 8
FINAL COPY-PASTE LISTS
BrandsToTrack: ["FocusBrand", "Competitor 1", ...]
Topics: ["topic 1", "topic 2", ..., "topic 8"]
You’ll get lists like the following:
BrandsToTrack: ["HubSpot", "Salesforce", "Zoho CRM", "ActiveCampaign", "Pipedrive", "Adobe Marketo Engage", "Pardot"]
TopicsToTrack: ["Marketing automation capabilities", "Sales pipeline management and forecasting", "CRM functionality and contact management", "Integration ecosystem and third-party connections", "Pricing and scalability for different business sizes", "Ease of use and implementation complexity", "All-in-one platform vs specialized tools", "Target market positioning and buyer personas"]
Note: I’d like run this a couple of times in ChatGPT/Perplexity and manually pick the topics you feel are most important.
Step 2) Run two prompts to map brands and topics
Now that you’ve got your lists, it’s time to run two simple prompts.
Brand → Topics – Ask AI: “What are the top things (topics) people associate with [Brand]?”
This shows what each brand is known for today.Topic → Brands – Ask AI: “What brands do you most connect with [Topic]?”
This shows who “owns” each topic in the market.
These two views give you what you need to build three simple reports:
Your Brand’s Visibility: where your brand shows up strongest (and where it’s missing).
Market Hotlist: The top 3 brands for each topic, so you can see which areas are dominated by the same players and which ones are still wide open.
Competitor Contrast: what you share with a rival and what’s unique to each.
Run these prompts once, and you’ll have everything you need to create reports that make brand visibility in AI easy to track and act on.
Let’s create the Brands → Topics list:
PROMPT
You are an impartial marketing analyst.
Task: For each brand in the list below, list the 10 things you most strongly associate with that brand.
Focus on product features, customer types, problems solved, and qualities (like ease of use, pricing, integrations).
Brands:
["HubSpot", "Salesforce", "Zoho CRM", "ActiveCampaign", "Pipedrive", "Adobe Marketo Engage", "Pardot"]
Output format (JSON only):
{
"BrandAssociations": {
"HubSpot": ["1. ...","2. ...","3. ...","...","10. ..."],
"Salesforce": [...],
"Zoho CRM": [...],
"ActiveCampaign": [...],
"Pipedrive": [...],
"Adobe Marketo Engage": [...],
"Pardot": [...]
}
}
You’ll get a list like this for each Brand:
You can store this list in memory using the following - “Store this list in memory under 'BrandAssociations'“
Let’s create the Topic → Brands list:
PROMPT
You are an impartial marketing analyst.
Task: For each topic in the list below, list the 10 brands you most strongly associate with that topic.
Focus on real brands that buyers would consider in this market.
Topics:
["Marketing automation capabilities",
"Sales pipeline management and forecasting",
"CRM functionality and contact management",
"Integration ecosystem and third-party connections",
"Pricing and scalability for different business sizes",
"Ease of use and implementation complexity",
"All-in-one platform vs specialized tools",
"Target market positioning and buyer personas"]
Output format (JSON only):
{
"TopicAssociations": {
"Marketing automation capabilities": ["1. ...","2. ...","...","10. ..."],
"Sales pipeline management and forecasting": [...],
"CRM functionality and contact management": [...],
"Integration ecosystem and third-party connections": [...],
"Pricing and scalability for different business sizes": [...],
"Ease of use and implementation complexity": [...],
"All-in-one platform vs specialized tools": [...],
"Target market positioning and buyer personas": [...]
}
}
You’ll get a list like this for each Topic:
You can store this list in memory using the following - “Store this list in memory under 'TopicAssociations'“
Step 3) Turn the outputs into 3 simple reports
With your two lists from Step 2, you now have everything you need to build your brand visibility reports.
Report 1: Your Brand’s Visibility
What it shows: Where your brand shows up for each topic in comparison to the competitor list you generated in Step 1.
Why it’s useful: You can quickly see where you’re strong, where you’re weak, and which gaps to fix first. Think of it like a “heatmap” of your visibility in AI.
PROMPT:
You are an analyst.
Task:
From the TopicAssociations list stored in memory, build a rank tracker.
Steps:
1. Use this brand list: ["HubSpot", "Salesforce", "Zoho CRM", "ActiveCampaign", "Pipedrive", "Adobe Marketo Engage", "Pardot"]
2. For each topic, create a table row.
3. Columns = all brands in the list.
4. Cell = the rank (1–10) for that brand on the topic, or "—" if missing.
Output format:
Markdown table:
Topic | Brand A | Brand B | Brand C | Brand D | ...
After the table, output a short summary table:
Brand | #Topics Appearing | #Top3 Rankings | #Missing (no rank)
Here’s an example of the report you’ll get:
Report 2: Market Hotlist
What it shows: The top 3 brands for each topic, so you can see which areas are dominated by the same players and which ones are still wide open.
Why it’s useful: Helps you understand where you must compete to stay visible (table stakes) versus where you can break through by owning a less-contested topic (white space). Think of it like a “leaderboard” for the most important topics in your market.
PROMPT:
You are an analyst.
Task:
From the TopicAssociations list stored in memory, show which topics are most strongly dominated by certain brands, and where the opportunities are.
Steps:
1. For each topic, list the Top 3 brands with their exact rank positions.
2. Create a table: Topic | #1 Brand | #2 Brand | #3 Brand
3. After the table, add two lists:
- Dominated Topics: Topics where the same brand holds #1 in multiple places (list the brand + topics).
- Opportunity Topics: Topics where the Top 3 are different brands (no clear leader). These are good white-space opportunities.
Here’s an example of the report you’ll get:
You’ll get a summary like:
Summary:
HubSpot is the clear dominator, owning 4/8 topics at #1.
Salesforce follows with 3 #1 positions, mainly in pipeline and data/CRM-related themes.
Zoho CRM wins pricing/scalability outright.
White-space opportunity is in multi-brand split topics (Automation, Pricing, Ease of use), where no single vendor owns the narrative.
Report 3: Competitor Contrast
What it shows: A side-by-side view of you vs. one competitor and what you both share, what only you own, and what only they own.
Why it’s useful: Helps you sharpen your differentiators and decide whether to fight for their ground or double down on your own. Use it to shape messaging and AI engine optimisation.
PROMPT:
You are an analyst.
Task:
Compare two brands from the list below. Show what they share, what only Brand A owns, and what only Brand B owns.
Steps:
1. Use the BrandAssociations list stored in memory
2. Compare Brand A vs Brand B.
3. Output three lists:
- Shared (up to 5): things both brands are strongly linked to
- Brand A Only (up to 5): things only Brand A is linked to
- Brand B Only (up to 5): things only Brand B is linked to
Output format:
Shared: [ ... ]
Brand A Only: [ ... ]
Brand B Only: [ ... ]
Inputs next:
Brand A = HubSpot
Brand B = Salesforce
Here’s an example of the report you’ll get:
And there you go!
We’re in the early days of AI Engine Optimisation. ChatGPT is an incredible tool to start understanding how visible your brand is for relative topics.
Until Next Time,
Happy AI’fying
Kieran
How would this response rank against using a tool like semrush which also has ai search functionalities?