Research · Market & customer research
The Best Prompt for Market Research
The best prompt for market research makes the model show where each number comes from. This prompt asks for a scoped market definition, need-based segments, a top-down and bottom-up size range, competitors and pricing, and it tags every figure as sourced, provided by you, estimated or unverified. It ends with the assumptions your decision depends on and a low-cost way to test each one with real customers.
Desk Market Research Brief Prompt
Produces a desk-research brief on a market, with segments, a range-based size estimate, competitors and pricing, and every figure labeled as sourced, provided or estimated.
- Works with
- Perplexity, ChatGPT, Claude, Gemini
- Difficulty
- Intermediate
- Complexity
- Multi-part reply
- Variables
- 5
- Version
- 1.0
Customize this prompt
Fill in what you know. Anything left blank stays as a [bracketed placeholder]. Nothing you type leaves your browser.
5 variables · fields marked * matter most
# Role You are a market research analyst preparing a desk-research brief. You separate what is sourced from what is estimated, and you would sooner report a gap than fill it with a plausible number. # Brief - Product or service: [Product or service] - Target customer: [Target customer] - Geography: [Geography] - Decision this research must inform: [Decision the research must inform] - What I already know (data, competitors, prices): [What you already know] # Evidence labels Tag every figure and every factual claim about a company with one of: [SOURCED] - you retrieved it in this session; give publisher, date of the figure and a link. [PROVIDED] - it comes from my notes above. [ESTIMATE] - your own reasoning; show the inputs and arithmetic. [UNVERIFIED] - recalled without a source you can show; treat as a lead to check. If you cannot search the web in this session, say so in the first line and do not use [SOURCED] at all. # Sections 1. Market definition - what is in and out of scope, the unit of demand (customers, transactions, seats and so on), and adjacent markets I should not confuse it with. 2. Customer segments - three to five segments by need or buying situation, not demographics alone. For each: the problem, current workaround, who pays, and what triggers a purchase. 3. Market size - a top-down and a bottom-up estimate, each as a range. List every input in a table with its label. If the two approaches disagree widely, explain why instead of averaging them. 4. Competition - direct competitors, substitutes, and the option of doing nothing. Name only companies you can source or that I provided; otherwise describe the type of competitor. 5. Pricing and channels - observed price points and pricing models, and how customers in this market find and buy. 6. Trends and constraints - demand drivers, regulation, and anything that could shrink the market, each tied to evidence. 7. Implications for the decision - what the evidence supports, what it does not, and how confident you are (high, medium, low) with a reason. 8. Assumptions to test - the five assumptions the decision depends on most, and a cheap primary-research method for each (interviews, survey, landing-page test, pre-orders). # Constraints - No figure without a label, a date and a unit. No growth rates or market sizes from unnamed "industry reports". - Prefer primary sources: statistical agencies, regulators, company filings, trade bodies. - Do not present a single-number market size. - Keep it under 900 words excluding tables. # Output format Eight numbered headings. Tables for market-size inputs and competitors. End with a source list containing only the links you used.
Use it when you are
- Sizing an opportunity before writing a business plan or pitch
- Checking whether a new product line or region is worth deeper investigation
- Preparing a first-pass market overview for a client or leadership team
- Listing the assumptions to test before commissioning paid research
Why this prompt works
Four labels make the evidence quality visible
A market brief full of confident numbers is hard to audit. Tagging each figure as sourced, provided, estimated or unverified shows at a glance which parts rest on a document you can open and which rest on reasoning or recall, so you know where to spend your checking time.
The model must declare whether it can search
Whether a model has live web access changes what its figures are worth. Requiring a first-line statement, and banning the sourced label when search is unavailable, prevents recalled numbers from being dressed up as citations. It also tells you immediately whether to rerun the prompt in a tool that can search.
Two sizing methods and a range
Analysts cross-check a top-down estimate against a bottom-up one because each fails in a different way. Asking for both as ranges, with an input table and an explanation of any disagreement, exposes the assumptions doing the most work and avoids a single figure that implies precision the evidence cannot support.
Segments are defined by need and trigger
Demographic segments describe who people are, not why they buy. Asking for the problem, the current workaround, the payer and the purchase trigger gives you segments you can recruit for interviews and write messaging against, and it feeds directly into the bottom-up size estimate.
It ends at primary research
Desk research can narrow a question but cannot confirm that customers will pay. Closing with the five assumptions the decision rests on, each paired with an inexpensive test, turns the brief into a plan for the work that must happen outside the chat window.
How to use it
- Use a model or mode with web search switched on. This prompt depends on current, linkable sources.
- Write the decision as something you will do or not do. It shapes section 7 and the assumptions list.
- Paste anything you already know, including competitor names and prices, so the model does not spend the brief rediscovering them.
- Go straight to the market-size input table. Open every [SOURCED] link and check the figure, its date and its geography.
- Challenge the estimate: ask the model to recalculate with your corrected inputs, or to show which single input changes the range most.
- Take the assumptions list into interviews or a small test before relying on the size estimate.
AI model compatibility
| Model | Status | Notes |
|---|---|---|
| Perplexity | Expected to work | Suited to this prompt because answers come with linked sources. Open each link behind a [SOURCED] figure and confirm that the page states that number, for that geography and year. |
| ChatGPT | Expected to work | Turn on web search before running the prompt. Without it, expect the first line to say so and the figures to be labeled [ESTIMATE] or [UNVERIFIED]. |
| Claude | Expected to work | Use a version with web search enabled if you want sourced figures. Without search, the prompt is still useful for the market definition, segments and assumptions to test. |
| Gemini | Expected to work | Keep the evidence labels section intact and check that linked sources are the original publisher, not a page that repeats the figure second-hand. |
“Expected to work” means the prompt uses plain, model-independent structure and has been reviewed for that — not that it has been run on that model. A model is marked Tested only when a test record is published below.
Worked example
Scenario: A founder weighing a filter-refill subscription for independent cafes before committing to a pilot.
# Role You are a market research analyst preparing a desk-research brief. You separate what is sourced from what is estimated, and you would sooner report a gap than fill it with a plausible number. # Brief - Product or service: Subscription refill service for commercial coffee machine water filters - Target customer: Independent cafes with one to three sites - Geography: United Kingdom - Decision this research must inform: Whether to spend three months building and running a pilot in one city - What I already know (data, competitors, prices): Spoke to six cafe owners; four replace filters late or not at all. Two equipment suppliers sell filters as one-off purchases. No sales data yet. # Evidence labels Tag every figure and every factual claim about a company with one of: [SOURCED] - you retrieved it in this session; give publisher, date of the figure and a link. [PROVIDED] - it comes from my notes above. [ESTIMATE] - your own reasoning; show the inputs and arithmetic. [UNVERIFIED] - recalled without a source you can show; treat as a lead to check. If you cannot search the web in this session, say so in the first line and do not use [SOURCED] at all. # Sections 1. Market definition - what is in and out of scope, the unit of demand (customers, transactions, seats and so on), and adjacent markets I should not confuse it with. 2. Customer segments - three to five segments by need or buying situation, not demographics alone. For each: the problem, current workaround, who pays, and what triggers a purchase. 3. Market size - a top-down and a bottom-up estimate, each as a range. List every input in a table with its label. If the two approaches disagree widely, explain why instead of averaging them. 4. Competition - direct competitors, substitutes, and the option of doing nothing. Name only companies you can source or that I provided; otherwise describe the type of competitor. 5. Pricing and channels - observed price points and pricing models, and how customers in this market find and buy. 6. Trends and constraints - demand drivers, regulation, and anything that could shrink the market, each tied to evidence. 7. Implications for the decision - what the evidence supports, what it does not, and how confident you are (high, medium, low) with a reason. 8. Assumptions to test - the five assumptions the decision depends on most, and a cheap primary-research method for each (interviews, survey, landing-page test, pre-orders). # Constraints - No figure without a label, a date and a unit. No growth rates or market sizes from unnamed "industry reports". - Prefer primary sources: statistical agencies, regulators, company filings, trade bodies. - Do not present a single-number market size. - Keep it under 900 words excluding tables. # Output format Eight numbered headings. Tables for market-size inputs and competitors. End with a source list containing only the links you used.
What a good response should contain
- A market definition that separates independents from chains and from offices with coffee machines
- Segments based on buying situation, such as owner-operators who maintain their own machine versus those on a service contract
- A bottom-up size range built from a number of cafes and a replacement frequency, with each input labeled and the cafe count either sourced with a link or marked as needing one
- A competition section that includes equipment service contracts and simply not replacing filters as substitutes
- Five assumptions to test, such as willingness to pay monthly, each with an interview or pre-order test
This is an illustrative scenario showing how to fill in the prompt and what to look for. It is not a recorded model output. Recorded outputs will be published here once this prompt has a test record.
Testing and editorial review
Status: Untested. Reviewed for structure and clarity by the editorial team. No recorded model test yet.
No test record has been published for this prompt yet. When it is tested, this section will show the model and version, the date, the inputs, the actual output and notes against our seven-point rubric. We do not publish numerical scores.
| Author | TheBestPromptFor Editorial Team — The editorial team writes, reviews and maintains the prompts and guides on TheBestPromptFor. Prompts are drafted with AI assistance and reviewed against published editorial standards. |
|---|---|
| Published | November 15, 2026 |
| Last reviewed | November 15, 2026 |
| Last tested | Not yet tested |
| Version history |
|
Limitations and things to check
- Without web search the model cannot retrieve current figures, and anything it recalls may be outdated or wrong.
- Linked sources can be secondary pages that repeat a number without its original context. Trace important figures to the original publisher.
- Paid analyst reports and private-company data are usually out of reach, so competitor revenue and share will often be missing.
- Niche and local markets may have little published data, leaving the size estimate dependent on your own inputs.
- Desk research does not show whether customers will buy. Only primary research and sales attempts do.
Alternative prompts
Business
Competitor Analysis Prompt (From Your Own Research)
Organizes the competitor information you have collected into a comparison, a positioning read and a list of what you still need to find out.
- ChatGPT
- Claude
- Gemini
Marketing
Evidence-Based Customer Persona Prompt
Builds personas from the interviews, reviews or survey answers you paste in, tagging every line as evidence, inference or assumption.
- ChatGPT
- Claude
- Gemini
Research
User Interview Discussion Guide Prompt
Writes a timed discussion guide for customer or user interviews: screener, opening script, behavior-based questions with probes, and a table tying each question to a learning goal.
- ChatGPT
- Claude
- Gemini
Business
Evidence-Based SWOT Analysis Prompt
Builds a SWOT around a specific decision, ties each point to evidence you supplied, and converts the grid into strategic options.
- ChatGPT
- Claude
- Gemini
Related resources
Guide
How to Improve AI Responses
A diagnostic approach to weak AI answers: name the fault, trace it to the gap in your prompt, then follow up precisely or rewrite and start again.
Technical guide
How to Test AI Prompts
A seven-criterion rubric and a nine-field test record for evaluating prompts, plus why this site describes results in words and publishes no scores.
Frequently asked questions
Can I trust a market size figure from an AI model?
Trust the method, then check the inputs. A range built from inputs you can open and verify is useful. A single number with no source is not, however confident it sounds. That is why the prompt bans unlabeled figures and asks for the arithmetic behind every estimate.
Which model should I use for this?
One that searches the web and links its sources, since the brief is only as good as the evidence behind it. A model without search can still help you define the market, structure segments and list assumptions, but its figures should be treated as leads to verify.
How is this different from a competitor analysis?
This prompt looks at the whole market: who buys, how many of them there are and what shapes demand. Competitors are one section. If you already know the market and need a detailed comparison of specific rivals, use the competitor analysis prompt instead.
Sources and references
- Prompt engineering guide — OpenAI
- Prompt engineering overview — Anthropic
Was this prompt useful?
No account needed. Do not include personal information.