Deep Research Prompt Generator
Create an efficient prompt for deep research from any topic of your choosing.
Use This When
General business and marketing workflows.
Inputs Needed
Model/tool, objective, inputs, constraints, output format, examples, evaluation criteria.
Expected Output
Reusable prompt template with variables, instructions, examples, output format, validation tests.
The Workflow Prompt
You are a prompt engineer and AI workflow architect. Objective: Deep Research Prompt Generator Context: Create an efficient prompt for deep research from any topic of your choosing. Original task: <Role> You are a highly specialized AI assistant expert in creating conversational generation prompts into a structured deep research prompt. </Role><Context> Users want to create quality deep research that tackles all the features, aspects and themes of the topic. Your purpose is to create high-quality, comprehensive research prompts that explore all key features, aspects, and themes of a given topic. You will convert a user's initial conversational request into a detailed, well-organized prompt suitable for deep research. </Context> <Task> Your primary task is to receive a user's research topic and transform it into a structured, deep research prompt.1. First, ask the user to provide the specific topic they want to research. 2 Once the topic is provided, analyze it for key details and identify potential areas for deeper exploration. 3 Based on your analysis, ask the user for additional, specific details that would help broaden and deepen the research. 4. If the user provides more details, integrate them. If not, proceed with the transformation using only the initial information provided. Do not ask for details a second time. 5. Your final output must be an improved, structured prompt that the user can directly use for deep research.</Task><Important_Considerations>Do not add or hallucinate any details from user's original promp/topic..The LLM will not do the research. The output will be a PROMPT that will be used by users to create a deep research. Maintain the core idea and purpose of the user's original request. </Important_Considerations> Inputs I may provide: Model/tool, objective, inputs, constraints, output format, examples, evaluation criteria. Operating instructions: - First, restate the objective in one clear sentence. - If critical information is missing, ask up to 5 focused questions. If there is enough information to proceed, make practical assumptions and label them. - Use a Detailed response style. - Be specific to the business, audience, channel, and constraints provided. - Avoid generic AI advice. Give concrete recommendations, examples, templates, copy, or steps I can use. - When current facts, competitors, laws, prices, policies, or market claims matter, use current research and cite sources. - Do not expose hidden chain-of-thought. Provide a concise rationale or decision summary instead. - End with a short QA checklist that helps me verify the output. Required output: Reusable prompt template with variables, instructions, examples, output format, validation tests. Caution: Avoid generic output; require concrete examples, assumptions, and next steps.
QA Follow-Up Checklist
After the AI returns its output, verify against:
- Output is specific to the provided business/context.
- Assumptions are clearly labeled.
- No unsupported claims without source checks.
- Next actions are clear and usable.
Follow-Up Prompt
Now turn the result for 'Deep Research Prompt Generator' into a client-ready version: tighten wording, remove fluff, add missing assumptions, and provide the next 3 actions.
Avoid / Cautions
Avoid generic output; require concrete examples, assumptions, and next steps.
How Different Verticals Use This Workflow
Restaurant & Hospitality
A multi-unit restaurant operator deciding whether to expand into Calgary uses the prompt to build a research brief on Calgary's downtown F&B landscape, competitor density by cuisine, average rent per sqft, and labor availability. Output: a 12-page report with 5 named competitors, rent data from CBRE, and a recommendation to delay 6 months. Saves an estimated $480K from avoiding a bad lease.
Retail & E-commerce
A DTC supplement founder deciding whether to launch in EU uses the prompt to build a research brief on EU Novel Foods regulations for adaptogens, top 5 EU competitors, distribution options (warehouse vs 3PL), and VAT implications. Output: a 9-page report flagging that 2 of their 3 hero SKUs aren't EU-compliant. Avoids a $200K product reformulation surprise.
Professional Services & B2B
A 40-person consulting firm deciding whether to add a private equity practice uses the prompt to build a research brief on PE consulting market size, average ACV by firm tier, top 8 competitors, and talent acquisition costs. Output: a 14-page report with realistic 3-year revenue model. Firm decides to acquihire a 4-person PE consultancy instead of building.
Beauty & Personal Care
A clean beauty founder deciding whether to launch a peptide line uses the prompt to build a research brief on the peptide skincare market, regulatory landscape (Health Canada and FDA), top 6 competitors at their price point, and ingredient sourcing. Output: 11-page report identifying a Korean peptide supplier 30% cheaper than US options. Saves $140K in year-one COGS.
Local & Trade Services
A residential GC deciding whether to acquire a 4-person painting subcontractor uses the prompt to build a research brief on Toronto painting market size, average labor margin, customer acquisition cost in trades, and integration risks. Output: 8-page report flagging the seller's customer concentration (62% from one builder). Walks away from the deal.
Frequently Asked
What inputs actually move the needle for a deep research prompt vs a Google search?
Three things: the decision the research will inform (not the topic), the audience for the research output (you alone? a board?), and what would make it useful vs not. 'Research the EV charging market' is a search. 'Research the EV charging market to decide whether to invest $2M in a charging network business by Q3, for a board of 4 angel investors, useful = a defensible recommendation with risks named' is a deep research prompt. The decision shapes the depth, sourcing, and structure.
Should I use ChatGPT 5.5 Thinking, Claude Opus 4.7, or Perplexity for deep research?
ChatGPT 5.5 Thinking with web search enabled or Perplexity Sonar Pro for the actual research pass — they pull current sources. Claude Opus 4.7 to structure the research prompt before you run it and to synthesize the findings after. Don't use any of them as the final source on contested facts (regulatory changes, financial figures, recent acquisitions) — they hallucinate confidently. Always verify the 3 numbers you'd actually quote in a board deck.
What does a great deep research prompt output look like?
A multi-section prompt with: the research question stated as a decision, 5-10 specific sub-questions, the desired output structure (executive summary, key findings with evidence, risks, recommendations), source quality criteria (only cite from named sources, no Quora or Medium), and a confidence-rating instruction (model rates its confidence per finding). The best ones include 'flag where the evidence contradicts the conventional wisdom in this space' as a forcing function for nuance.
When is this the wrong tool to reach for?
Skip it for research that requires interviews, observation, or proprietary data. AI deep research synthesizes public information well; it doesn't talk to 30 customers or run a survey. If your real question is 'do our customers actually want this,' AI research can pre-load context, but the answer lives in conversations. Also skip for hyper-current topics (last 7 days) — even web-enabled models lag by hours and may miss the thing your decision hinges on.