What this covers
Run a credible research project inside ChatGPT — Deep Research mode, Projects for context, and the spot-check loop that keeps the output honest.
Key tools and concepts:
- ChatGPT: OpenAI’s conversational AI assistant — the product that brought the GPT models to a mass audience.
Who this is for
ChatGPT Plus / Team / Enterprise users who want to use Deep Research without getting fooled by polished but shallow output.
When to reach for it
When you have a research question that needs 30-90 minutes of web reading and a 5-15 page brief at the end.
Step by step
-
Create a Project to anchor context. ChatGPT sidebar →
+ New project. Name it after the deliverable, e.g.2026 SaaS APAC pricing research. Project header →...→Instructions, paste:Your role: senior research analyst, specializing in <your domain>. Research goal (the core question this project answers): <one line, e.g. "Median and distribution of annual contract value for B2B SaaS in APAC, 2026"> Audience: <e.g. "me + 2 investors, all product-literate but not APAC experts"> Deliverable: <e.g. "8-section brief, 600 words per section, each section cites sources"> Before every answer: 1. Restate my sub-question (≤1 sentence) 2. List which project sources you used this turn STOP and ask me when: - Information is insufficient to conclude (do not hallucinate numbers / quotes) - My new question conflicts with the project goal -
Upload seed material. Project right rail →
Files→Add files:- Existing decks / old briefs / any prior research (PDF / MD / DOCX)
- 3-5 authoritative URLs you already have (save as
seed_sources.md, one URL per line + 1 sentence on why it’s authoritative) - A
glossary.mddefining the domain terms you want AI to use (so it doesn’t fall back on generic explanations)
-
Refine the question — make ChatGPT challenge you first. Open the first chat in the project:
Research goal is in instructions. Based on seed_sources and glossary, before I run Deep Research, do exactly these 3 things: 1. Decompose the research goal into 5-8 sub-questions, each independently researchable, ≤20 words 2. For each sub-question, list: - Which seed source partially answers it - What source type is still needed (industry report / 10-K / academic / primary interview) 3. Challenge me: what's ambiguous in the goal / what angles am I likely missing / what hidden assumptions could invalidate the conclusion Do NOT answer the research goal itself. Only output the 3 blocks above.Manually drop fuzzy sub-questions, add the angles you missed, then proceed.
-
Run Deep Research. Same project, new chat. Model selector →
o3orGPT-5 (deep research)(exact name per current ChatGPT). ToggleDeep Researchmode on. Send:Research goal (from project instructions): <paste the 1-line goal> Refined sub-questions: 1. <sub-question 1> 2. <sub-question 2> ... Deliverable spec: - 8 sections, each section title = one sub-question - ≤600 words per section - 3-6 inline citations per section - End with a `## References` block: link + publication date + source type (report / 10-K / blog / primary) - Every number must have a source; if no primary source exists, explicitly write "not found" - Mark each finding as: "established" (≥2 independent sources agree) / "contested" (sources conflict) / "uncertain" (only 1 low-authority source) Time scope: only 2024+ sources as primary evidence.Once it starts, the run takes 5-30 minutes. Don’t send other messages in this chat (it cancels Deep Research).
-
Mark weak sections + re-research. When the output lands, scan and mark a section “weak” if:
- Citations are all blogs / Medium / Reddit, no primary data
- Numbers given without primary-source links (hedge words like “approximately”, “estimated”)
- The sub-question is dodged or answered too generically
Pick the 2 weakest, open a new chat (same project, Deep Research still on):
Below is section <N> from the prior Deep Research run: <paste the section> Redo this section. Rules: 1. Replace all citations — only use these source types: industry reports (Gartner / Forrester / IDC / commissioned research) / public-company filings / primary data (expert interviews, community surveys, official API data) 2. No Medium / Reddit / personal blogs as primary evidence 3. Every number needs a clickable original source link (not "as reported by X") -
Citation spot-check: Deep Research’s only non-skippable safety step. Pick 5 inline citations at random, for each:
- Click the link, confirm the page exists (not a dead link / paywalled)
- Cmd+F the cited number or phrase in the original page, confirm it actually appears
- Flag “weak” if: original says the opposite / cited from a secondhand citation (the source quotes an earlier source) / source is >2 years stale
For every flagged section, redo via step 5.
-
Executive summary — with explicit confidence labels:
Based on all 8 sections + the redone sections, write a 1-page executive summary: - Length ≤250 words - Structure: - "Established" paragraph (multi-source corroborated, ≥2 conclusions) - "Contested" paragraph (sources conflict — name the conflict, don't pick a side) - "Uncertain" paragraph (missing data or single source — list what's still needed) - "Next steps" paragraph (≤3 actionable items) Tag each conclusion in brackets with confidence: [established / contested / uncertain]. No "significant" / "substantial" without magnitude. -
Archive + reuse. Bundle the entire project (instructions, seed sources, refined sub-questions, Deep Research output, citation audit, executive summary):
chatgpt_research_2026_05_21_saas_pricing/ ├── instructions.md ├── seed_sources.md ├── refined_subquestions.md ├── deep_research_v1.md ├── deep_research_v2_redone_sections.md ├── citation_audit.md └── executive_summary.mdNext similar research: duplicate the project → change instructions “Research goal” → swap seed sources. Steps 1-3 collapse into 5 minutes.
Recommended workflow
Project setup → question refinement → Deep Research run → weak-section re-research → manual citation spot-check → executive summary with contested/uncertain flags.
Common mistakes
- Treating the Deep Research draft as the final
- Not using Projects — losing context across runs
- Skipping the “challenge my question” step
- Not flagging “contested vs settled”
FAQ
- Deep Research vs regular chat?: Deep Research browses the web for 5-30 min and writes a long brief. Regular chat is faster but no live web.
- How many runs can I do?: Plus plan: limited per month — check your account. Reserve Deep Research for substantive briefs.