Research

Build a source-aware research plan

Plan what to investigate, where to look, and how to separate evidence from inference

5 min setupTested with:ChatGPTClaudeGeminiReviewed: 2026-08-27

Add your context

Your text stays in this browser. AILesson Prompts does not send it to a model or server.

Research topicDecision this research should supportScope and constraints

Your prompt

Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt

Create a research plan for this topic:
[topic]

The research must support this decision:
[decision]

Scope and constraints:
[limits]

Produce:
1. The decision-relevant questions, ordered by importance
2. A source hierarchy that prioritizes primary and current sources
3. Suggested search queries for each question
4. What evidence would confirm or weaken each possible conclusion
5. A table for recording claim, source, publication date, evidence, and confidence

Do not answer the research questions yet. Flag time-sensitive claims that will require current verification.

Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.

From input to outcome

A worked example

See how concrete context turns this recipe into a usable result

Actual input

Research topic: Whether a 35-person remote design agency should replace its current video-meeting tool with an AI meeting assistant that records, transcribes, and summarizes client calls.
Decision this research should support: Recommend whether to run a paid pilot, and define the minimum consent, privacy, security, and workflow requirements that a pilot product must meet.
Scope and constraints: Cover clients and staff in the EU and UK. Compare at most three products. Use no more than two working days for desk research. Prioritize sources published or updated within the last 12 months. Do not evaluate transcription accuracy through hands-on testing yet.

Example output

1. Decision-relevant questions

What consent and notification controls exist for recording EU and UK participants?
How does each provider process, retain, delete, and train on recordings and transcripts?
Which security controls and contractual commitments can be verified?
Can administrators enforce the required settings across the agency?
Does the workflow support client calls without unacceptable friction or hidden cost?

2. Source hierarchy

Current product documentation, privacy notices, data-processing agreements, subprocessors, security pages, and pricing pages.
EU and UK regulator guidance and applicable official legal texts.
Current trust-center audit reports or certifications.
Reputable independent analysis for context only; reviews and forums may suggest questions but should not establish compliance.

3. Suggested searches

Consent: [product] recording consent EU UK notification controls
Data use: [product] privacy DPA retention deletion AI training transcripts
Security: [product] trust center SOC 2 ISO 27001 encryption subprocessors
Administration: [product] admin disable recording retention policy SSO
Workflow and cost: [product] pricing meeting bot client consent integrations 2026

4. Evidence tests

A pilot is strengthened by documented participant notification, configurable retention and deletion, a suitable DPA, verifiable security controls, centralized administration, and transparent pricing. It is weakened by ambiguous training use, uncontrollable bot entry, missing regional terms, unverified security claims, or costs that cannot be bounded. Transcription quality remains unresolved because hands-on testing is excluded.

5. Evidence log

Claim	Source	Date	Evidence	Confidence
Requirement or product claim	Direct URL or document	Published or accessed date	Relevant passage or setting	High/Medium/Low

Current verification required: pricing, feature availability, subprocessors, retention controls, certifications, privacy terms, and regulator guidance.