Skip to content
Codextroop Prompt Library

Market Research Report

Deep Research Brief (DeeperSearch)

AI Tool:
GrokWorks with Grok
Category: Reports

A real multi-angle research pass with a decision target - turn on DeeperSearch / Think. Not a first-page summary.

Prompt Quality - 100/100 · Strong

The Prompt

grok-deep-research-brief.md

Run a proper research pass on this - I want you to actually dig, not answer from general knowledge with a couple of quick searches bolted on. Use DeeperSearch if it is available.

RESEARCH QUESTION:
{{question}}

WHAT I ALREADY KNOW:
{{known}}

WHAT WOULD CHANGE MY MIND OR DECISION:
{{decision_needle}}

INSTRUCTIONS:
1. Identify 3 - 5 genuinely different angles this question needs (e.g. technical feasibility, cost, precedent / track record, expert consensus, recent developments) - not 3 - 5 searches of the same angle worded differently.
2. Search each angle separately and go deep enough to find something beyond the first-page summary - actual sources, not just aggregator articles about the sources.
3. Where sources disagree, say so explicitly and give your read on which is more credible and why, rather than presenting a false average of conflicting claims.
4. Directly address the "what would change my mind" item - tell me plainly whether what you found moves that needle or not.
5. Give a bottom line: if you had to decide this yourself right now with what you found, what would you do.

OUTPUT FORMAT:
- Angles investigated (list)
- Findings per angle (with source credibility noted where sources conflict)
- Direct answer to "what would change my mind"
- Your bottom-line call if you were deciding

CONSTRAINTS:
- Prioritize original sources and primary data over aggregator summaries.
- If after real digging the answer is genuinely unclear or contested, say that honestly rather than manufacturing false confidence.

Negative prompt

Included when you copy the prompt

Do not answer from general knowledge with a couple of bolted-on searches. Do not average conflicting sources into a false consensus. Do not skip the decision-needle item. Do not invent citations. If a source is missing, say so.

The PromptUTF-8

Fill in your details below - this prompt updates as you type, then copy it.

Fill in your details

0 of 3 filled. Preview updates as you type - then copy the finished prompt (negative prompt included when present).

What you need to know or decide.

Your current understanding - so it skips the basics.

The specific fact that would actually change your call.

grok-deep-research-brief.md

Prompt preview

Run a proper research pass on this - I want you to actually dig, not answer from general knowledge with a couple of quick searches bolted on. Use DeeperSearch if it is available.

RESEARCH QUESTION:
{{question}}

WHAT I ALREADY KNOW:
{{known}}

WHAT WOULD CHANGE MY MIND OR DECISION:
{{decision_needle}}

INSTRUCTIONS:
1. Identify 3 - 5 genuinely different angles this question needs (e.g. technical feasibility, cost, precedent / track record, expert consensus, recent developments) - not 3 - 5 searches of the same angle worded differently.
2. Search each angle separately and go deep enough to find something beyond the first-page summary - actual sources, not just aggregator articles about the sources.
3. Where sources disagree, say so explicitly and give your read on which is more credible and why, rather than presenting a false average of conflicting claims.
4. Directly address the "what would change my mind" item - tell me plainly whether what you found moves that needle or not.
5. Give a bottom line: if you had to decide this yourself right now with what you found, what would you do.

OUTPUT FORMAT:
- Angles investigated (list)
- Findings per angle (with source credibility noted where sources conflict)
- Direct answer to "what would change my mind"
- Your bottom-line call if you were deciding

CONSTRAINTS:
- Prioritize original sources and primary data over aggregator summaries.
- If after real digging the answer is genuinely unclear or contested, say that honestly rather than manufacturing false confidence.
0 / 3 filledPreview
grokdeepersearchresearchsourcesbrief

Examples

Example input

Question: Should we switch our EU analytics from a US SaaS to an EU-hosted alternative this quarter? Already know: We are GDPR-bound; current vendor is cheaper; legal wants a DPA story. Needle: Whether any EU-hosted option can match our current event volume and warehouse export without a 3-month rewrite.

What you should get

3 - 5 distinct angles, findings with source-conflict notes, a direct yes/no on whether the needle moved, and a bottom-line call - not a Wikipedia-style overview.

People also copied

Rumor / Claim Verification

When to use this prompt

Use it when

  • You need help with researchers.
  • You need help with founders.
  • You need help with operators.
  • You need help with writers.

Skip it when

  • You only need a one-line ask with no constraints, format, or context.
  • You cannot supply the fill-in details the template asks for.
  • A different category or tool-specific workflow fits the job better.

How to get a good result

Done when

3 - 5 distinct angles, findings with source-conflict notes, a direct yes/no on whether the needle moved, and a bottom-line call - not a Wikipedia-style overview.

Common mistakes

  • Leaving question, known, decision_needle vague - the model then invents placeholders.
  • Asking Grok for a shorter answer after a long structured prompt - that throws away the format rules.
  • Stopping when the first draft appears instead of checking it against the success criteria below.

FAQ

What do I fill in before I copy this prompt?
Replace these placeholders with your details: question, known, decision_needle. Required fields are marked on the page. The prompt text updates as you type.
Which AI tools is this prompt written for?
It works best with Grok. Copy the filled prompt and paste it into the tool you already use.
What should a good result look like?
3 - 5 distinct angles, findings with source-conflict notes, a direct yes/no on whether the needle moved, and a bottom-line call - not a Wikipedia-style overview.
What are common mistakes with this prompt?
Leaving question, known, decision_needle vague - the model then invents placeholders. Asking Grok for a shorter answer after a long structured prompt - that throws away the format rules. Stopping when the first draft appears instead of checking it against the success criteria below.

Best for

  • Researchers
  • Founders
  • Operators
  • Writers

More prompts like this — same topic and style.

More like this

Rumor / Claim Verification
Research a Topic
Compare Research Sources
Real-Time X Sentiment Read