How to use ChatGPT, Gemini or Perplexity to map any market with speed and clarity

Table of contents

• Before prompting
• Step 1: Defining the scope of your competitive landscape
• Step 2: Defining the structure and categorization of your mapping
• Step 3: Sourcing companies for your competitive landscape
• Step 4: Enriching your map
• Practical example

Definition

A competitive landscape is a visual map that groups company logos into clear categories so readers can understand how the market is structured at a glance. You can find examples here.

About this guide

The aim of this guide is to show through practical steps how to create a competitive landscape with the help of LLMs. The guide is intended for anyone who needs to build such maps, from VC and PE analysts to founders, marketing analysts or corporate strategists.

Before Prompting

Before jumping into prompting it helps to define your goal broadly and have an idea of what you want to achieve with your mapping. Here are a few examples of what you might be looking for:

• Market category mapping
  • Type of map: Organizes a market into broad functional categories that describe the main areas of the space.
  • Goal: Understand the overall structure of the market and how different categories relate to each other.
• Value chain structure mapping
  • Type of map: Shows the ecosystem with adjacent, upstream and downstream players.
  • Goal: Understand how the value chain is organized.
• Positioning mapping
  • Type of map: Groups companies by customer segment such as SMB, mid market or enterprise.
  • Goal: Understand which customer segments are well served and which are underserved.
• Geographic mapping
  • Type of map: Organizes companies by country or region.
  • Goal: Understand maturity levels or regional strengths as well as potential opportunities.
• Go to market mapping
  • Type of map: Categorizes players by distribution style such as partner led, PLG or sales driven.
  • Goal: Understand which GTM models dominate the market and why.
• Pricing model mapping
  • Type of map: Groups companies by pricing approach such as usage based or seat based.
  • Goal: Understand how monetization shapes the market.
• Maturity or stage mapping
  • Type of map: Separates incumbents, challengers and emerging startups.
  • Goal: Understand trends in company growth and financing.

Once you’re clear about your goal it’s time to move to the next phase and define the scope of your mapping.

Step 1: Defining the scope of your mapping

Defining the scope of a market and therefore the scope of your mapping can be difficult because, when you start this exercise, you might not have a clear understanding of the market or its structure. If this happens, a practical way to clarify the scope is to use LLMs and go through a set of questions and answers. Here is a prompt you can use to do that:

Prompt #1

Comments:

• You can copy and paste one of the goals I listed in the previous section.
• Depending on the market I want a landscape of, I often reduce the number of questions that the LLM should ask me.

Step 2: Defining the Market Categorization

Now that you have a clear description of your market and a well defined scope, you can start structuring it into categories that will help you organize the companies operating in this space.

You can use the following prompt to do so:

Prompt #2:

Now, can you turn that description into the structure of a competitive landscape? Please do the following:

• Extract the major categories that define this market.
• Write a short description for each category so a reader understands what it covers and why it matters.
• Keep the output concise and clear.
• Do not list companies. Focus only on categories.

Comments:

• Once you have your categories, take the time to review them and adjust them based on your needs. LLMs usually provide a good starting structure, but I often need to refine it a bit. 
• For example I often add categories that might be missing or split broad ones into more focused sections. 
• If you’re unhappy about a categorization, don’t hesitate to run the prompt again, you might get better results. It’s the nature of LLMs to be non deterministic, so you can get different answers based on the same prompt.

Step 3: Sourcing companies for your landscape

Now that you have a clear market description and scope as well as its main categories, it’s time to fill it up with companies.

From my experience LLMs are a helpful starting point for this part but they are far from a silver bullet. They can populate the first set of companies, but you will usually need to do additional research if you want to add more depth and substance.

Anyway, here is a prompt that will help you populate your landscape with the first companies (adapt it to your own needs):

Prompt #3

Additional comments:

• Results tend to be way better when you use the deep search mode of ChatGPT, Gemini or Perplexity.
• From my experience, providing an initial list of relevant companies usually leads to way better outputs, especially for niche markets.
• Running the prompt on several LLM providers such as ChatGPT, Gemini and Perplexity often surfaces different startups. I personally always do it.
• One important warning is that these models can return incorrect information because they rely on web search and training data. So it is worth verifying each company manually and its information. I regularly get startups/companies that don’t exist anymore or that were purely invented (it’s really crazy sometimes). This risk is not so high during the first two steps (market scope and categorization) but becomes real at this stage. So I wouldn’t blindly trust LLMs here.

Manual digging

I will not go into detail on how to manually search and find startups for your landscape, but here are the three main strategies I use depending on the type of competitive landscape I need:

• Sponsored results on Google searches. Run Google searches on market terms and check which companies appear in sponsored ads. When you use simple queries that match how the market speaks, you often see ads from companies active in that space. This works well for geo specific markets or for terms that end users actively search because these queries reveal where companies try to capture customer intent.
• Software directories. Search major product directories such as G2 or GetApp for established companies and Product Hunt for startups (etc.). It is nothing new, but it still works well today.
• Venture databases. Check venture oriented databases such as Tracxn, Dealroom, Specter or Crunchbase, which often surface companies that match the stage or funding profile you are looking for.

Comments:

• I have not automated this part yet, although it might be possible to run agents on it in the future. 
• For now, this manual work is still what takes a landscape from roughly half complete to fully complete in my case.

Step 4: Enriching your mapping

This step is not as critical as the previous ones and you can skip it, but I like to include it because it gives me a more dynamic view of the market rather than a static picture.

I usually enrich my competitive landscape by doing three things:

• Tag each company to segment and filter them with more nuance. This can include geo tags to filter by region, business model tags, market sub segment tags, customer type tags, investor tags or company stage tags. If you already have a clear tag list, you can ask LLMs to apply it.
• Add key company events to the map. I like to list the major events related to each company (such as funding rounds or acquisitions) because it gives me a more dynamic sense of how each company is evolving as well as the market (for example, are there many funding rounds the past year or not).
• Ask LLMs to describe each company based on the information you need. LLMs search and summarise much faster than humans, so I often use a standard prompt for every company and ask for specific details. I then add this information to the description field of each company.

Practical Example

In order to illustrate the real results you can get from this process, here is the different prompt chained together in order to create a mapping about emerging Sport Tech startups and which I use without prior knowledge about the market.

Prompt #1

I’m an analyst in a VC firm and I want to create a competitive mapping of emerging B2B and B2B2C Sport Tech startups in Europe and North America. My goal is to understand the space from a venture capitalist perspective (e.g: How the market is structured, what are the established and crowded categories, the emerging ones etc.) .

Before we start, I need your help to clearly define the market I want to analyze and the scope of its mapping. Please follow these instructions:

• Ask questions that guide me toward defining the market.
• Ask one question at a time.
• Do not ask more than ten questions.

At the end, provide a clear description of the market I want to explore in no more than two paragraphs.

Prompt #2

Now, can you turn that description into the structure of a competitive landscape? Please do the following:

• Extract the major categories that define this market.
• Write a short description for each category so a reader understands what it covers.
• Keep the output concise and clear.
• Do not list companies. Focus only on categories.

Prompt #3

Now find the companies that fit each category and return the results in a clear list. For every company you include, provide:

• One sentence describing what the company does
• The company website
• The company stage and location
• The level of funding or the latest round if available

Only include companies that are relevant to the market and the category. Avoid duplicates.

---

You can see the results on the different main LLMs here:

• Results on Perplexity
• Results on ChatGPT 5.1
• Results on Gemini 3