AI for Research Analyst
Open-ended coding takes 4–8 hours per wave, building a 40–80-slide research deck consumes another 2–4 hours of purely mechanical formatting, and proposal methodology sections get rewritten from scratch every engagement despite the structure never changing. The job requires both deep analytical thinking and an enormous volume of repetitive text work — and the text work tends to crowd out the thinking. These guides show you how to move through verbatim coding, report narrative drafting, and proposal writing faster so your time goes where your expertise actually matters.
Try right now
Copy a prompt, paste into ChatGPT, Claude, or Gemini
Works with any free AI chatbot, no signup needed
A review of your draft survey questions flagging leading language, double-barreled constructions, acquiescence bias, and ambiguous wording — with suggested rewrites for each problem.
Review these survey questions for response bias, leading language, double-barreled questions, and acquiescence bias. Flag each issue and suggest a rewrite. Questions: [paste survey questions]
View full prompt →Tip: Run this before programming in Qualtrics — fixing questions after fielding starts is expensive. Ask follow-up: "Now check these for reading level — aim for 8th grade."
A vivid, narrative persona profile for a market segment — complete with a name, motivations, frustrations, decision triggers, and a day-in-the-life vignette.
Create a persona profile for this audience segment: [paste key attributes — demographics, attitudes, behaviors, needs]. Include: first name, age range, job type, top 3 motivations, top 2 frustrations, how they make decisions, and a 3-sentence day-in-the-life. Keep it vivid and specific.
View full prompt →Tip: The more specific your input data, the more differentiated the persona. If the output sounds generic, follow up with "Make this persona more distinct from a typical mainstream consumer — what makes this segment unusual?" Repeat for each segment and compare the narratives side by side.
A compelling 250–300 word persona description that makes your segment feel like a real person — with demographic grounding, a day-in-the-life narrative, attitudes toward the category, and clear str...
Write a vivid persona for a customer segment with these characteristics: [paste segment profile — demographics, attitudes, behaviors, key quotes]. Give the persona a name. Include: who they are, their mindset, a day in their life, their relationship with [product category], and what would change their behavior. Make them feel like a real person.
View full prompt →Tip: Include at least 2–3 actual verbatim quotes from your open-ends in the segment profile — the AI will weave them in to make the persona feel more authentic. Keep the persona name gender-neutral unless your data specifically supports otherwise.
A table coding each verbatim response to 1–2 themes, plus a frequency count of how often each theme appeared — ready to paste into your report.
Here are [number] open-ended responses to "[survey question]". Identify 5–7 recurring themes, then code each response to 1–2 themes. Output as a table: Response | Theme 1 | Theme 2. Add a theme frequency summary at the end.
View full prompt →Tip: Paste responses in batches of 50–75 for best results. If the themes feel too broad, ask the AI to "split [theme name] into more specific subcategories."
A table with each response coded into your predefined categories, with flags on anything that doesn't fit cleanly — ready to paste into your analysis file.
Code these [N] open-ended survey responses using these categories: [list categories]. Return a table: Response | Code | Notes (flag if unclear). Responses: [paste responses]
View full prompt →Tip: Paste responses in batches of 30–50 for best accuracy. If you have a lot of "Other" responses, ask the AI to suggest additional categories after the first batch.
A structured executive summary with a clear narrative arc — organized around the most actionable findings — ready for your client deck.
Draft a 3-page executive summary from these survey findings. Client question: [client objective]. Key data: [paste top-line results]. Audience: [e.g., senior marketing leadership]. Lead with the 3 most actionable insights.
View full prompt →Tip: Add a sentence about the client's business situation for sharper framing. If the draft buries the most important finding, say "Put [finding] in the opening — that's the headline."
A structured research proposal covering objectives, recommended methodology, sample design with justification, timeline, and key deliverables — ready to edit and send.
Write a research proposal for this client need: [describe the business question]. Industry: [industry]. Budget: [$X]. Timeline: [X weeks]. Include: objectives, methodology recommendation, sample design, project timeline, and deliverables.
View full prompt →Tip: If you don't have a budget figure yet, use "TBD" and ask for multiple methodology options at different price points. Add "keep it under 2 pages" for a concise proposal clients actually read.
A draft set of survey questions plus a bias review flagging any leading language, double-barreled questions, or missing response options.
Draft [number] survey questions to measure [research objective] for [target audience]. Then review each question and flag: leading language, double-barreled questions, or missing answer options like "not applicable" or "prefer not to say."
View full prompt →Tip: Include your methodology in the prompt (e.g., "5-point Likert scale" or "multiple select") so the response categories come back correctly formatted. Run a second pass asking it to suggest a logical question order to reduce priming effects.
3–5 sharp insight statements that move from data observation to business implication to recommended action — the kind of "so what" language that makes research presentations land.
Turn these research findings into insight statements for [client/team]: [paste findings]. Each insight should follow this structure: Observation (what the data shows) → So What (why it matters for the business) → Action (one concrete next step). Write 4-5 insights.
View full prompt →Tip: This works best after you've already identified your headline findings — paste the 4–5 most significant ones, not everything in your dataset. If an insight sounds generic, follow up with "Make the Action more specific to [client's business context]."
A structured draft questionnaire with a screener, warm-up questions, main battery, and demographics — organized correctly and using appropriate scale types — ready to review and program into Qualtr...
Write a [length]-minute online survey for [target audience] about [research topic]. Include: [number]-question screener ([qualification criteria]), [number] warm-up questions, [number] main battery questions about [key topics], and standard demographics. Use 5-point scales where appropriate.
View full prompt →Tip: Specify the survey length (e.g., "15-minute") and qualify who should get through the screener — the more specific you are about quotas, the better the screener logic. Review scale types before programming; AI defaults to Likert but some topics call for frequency or ranking scales.
A concise industry overview covering key players, recent trends, consumer behavior shifts, and must-know terminology — enough to speak confidently in a client meeting without hours of research.
Give me a 1-page briefing on the [industry/category] market in [geography]. Cover: major players and their positioning, key trends in the past 1-2 years, how consumer behavior has shifted, and terminology I need to know for a meeting with [type of client]. Keep it concise — I have 20 minutes to read this.
View full prompt →Tip: Verify any specific figures (market share percentages, specific dates) before citing them in client meetings — use the briefing for background context and terminology, not as a primary source. If the industry is niche, add "especially focus on [specific segment]" to prevent overly broad output.
A list of 10 challenging questions your client is likely to ask about your findings — with suggested responses for each — so you walk into the presentation ready for anything.
I'm presenting research findings to a [CMO/VP Marketing/brand team] at [type of company]. Key findings: [paste 3-5 main findings]. Play the role of a skeptical client and generate 10 tough questions they might ask. Include a suggested response for each.
View full prompt →Tip: Do this the night before, not the morning of. Use the questions to identify any gaps in your analysis that you should fill before presenting — if you can't answer a question, that's your to-do list.
A list of 12–15 questions your client is likely to ask during or after the presentation — with a short suggested answer for each — so you walk in prepared for the tough ones.
I'm presenting these research findings to a [client role, e.g., VP of Marketing] at a [company type]: [paste key findings]. Generate 12 questions they're likely to ask, and a 2-sentence suggested answer for each. Focus on questions that challenge methodology, sample size, or implications.
View full prompt →Tip: The most valuable questions come from imagining a skeptical client. After reviewing the AI's list, add 2–3 questions you personally dread — then ask the AI to help you answer those too.
A line-by-line review of your survey questions flagging leading language, double-barreled questions, loaded terms, and order effects — with specific rewrites for each problem found.
Review these survey questions for methodological problems: leading language, double-barreled questions, loaded words, acquiescence bias, and unclear scales. For each problem found, explain the issue and provide a corrected version. Questions: [paste survey questions]
View full prompt →Tip: Run this before you send the questionnaire to programming, not after. If the AI flags something you disagree with, ask "What's the specific bias risk here?" — understanding the reasoning helps you decide whether to accept the rewrite.
A structured overview of the key players in a market category — their positioning, pricing models, and recent moves — formatted as a table plus a 2-paragraph summary.
Summarize the competitive landscape for [product/service category] in [geography/market]. Cover: top 5-6 players, how each positions itself, pricing models, and any notable moves in the past year. Format as a comparison table, then add a 2-paragraph strategic overview.
View full prompt →Tip: Always verify the AI's specific claims (pricing, product launches) before putting them in client materials — use it to build the structure and framework, then fact-check the details. For niche markets, add "focus on [specific segment]" to narrow the output.
A narrative synthesis identifying 3–4 overarching themes across your coded verbatim data — with supporting quotes and an explanation of how each theme connects to your research question.
Here are coded open-ended responses from a survey about [topic]. Codes used: [list codes]. Identify the 3-4 overarching themes, write a 2-paragraph narrative for each, and pull 2-3 supporting quotes. Data: [paste coded responses]
View full prompt →Tip: Share the research question ("we're trying to understand why customers switch providers") so the themes are framed in terms of what matters to the client, not just what's common in the data.
Each NPS comment tagged with a sentiment label (Positive / Negative / Mixed) and up to 3 topic tags from your predefined list — in a table ready to paste into your analysis file.
Tag each NPS comment with: Sentiment (Positive/Negative/Mixed) and up to 3 topic tags from this list: [your topic tags]. Return a table: Score | Comment | Sentiment | Topics. Comments: [paste NPS data]
View full prompt →Tip: Paste in batches of 50 with score and comment together. After the first batch, review the tagging and say "adjust: tag anything about wait times as 'speed' not 'service'" to refine for later batches.
Your statistical findings rewritten in clear, jargon-free language a senior executive can act on — without having to decode significance levels or crosstab notation.
Rewrite these research findings for a [CMO / VP of Marketing / CEO — pick one] who doesn't read statistics. Remove all jargon. Make each finding feel actionable. Findings: [paste stats/crosstab data]
View full prompt →Tip: Specify the decision the executive needs to make — "they're deciding whether to relaunch the product" — and the rewrite will naturally frame findings in terms of that decision.
Your statistical output rewritten as clear business insights — no p-values, no mean scores, just actionable "so what" statements your stakeholders will actually understand.
Rewrite these research findings as business insights for [audience, e.g., "a VP of Marketing who doesn't think in statistics"]: [paste findings with means, percentages, or significance flags]. Lead with the implication, not the statistic. 3-4 sentences per finding.
View full prompt →Tip: Specify the audience in the prompt — a CMO brief sounds different from a product team update. After the AI rewrites the findings, ask "Which of these would be most surprising to a typical [audience] and why?" to help you decide what to lead with in your deck.
A vivid, client-ready consumer persona with a name, day-in-the-life narrative, key motivations, frustrations, and brand relationship — based on your segment profile data.
Write a 1-page consumer persona from this segment profile. Include: persona name, age/life stage, day-in-the-life paragraph, top 3 motivations, top 3 frustrations, and how they relate to [brand/category]. Profile data: [paste segment stats]
View full prompt →Tip: Add "write it like a story, not a bullet list" for a more engaging persona that clients remember. For multiple segments, run this prompt once per segment and compare the resulting characters side by side.
A professional 3-paragraph executive summary with a headline finding, supporting data points, and a strategic implication — ready to place at the front of your research deck.
Write a 3-paragraph executive summary from these research findings: [paste bullet points]. Paragraph 1: headline finding. Paragraph 2: 2-3 supporting data points. Paragraph 3: one strategic implication for [audience, e.g., "the marketing team"]. Tone: clear, professional, no jargon.
View full prompt →Tip: Paste your actual data bullets — numbers, percentages, significance flags — not interpretations. The AI will do the translation into plain language. If the output is too generic, add a sentence describing the client's industry or business goal.
A complete 90-minute moderator guide with timed sections, probing questions for each topic, and at least one projective technique — ready for a moderator to run or edit.
Write a 90-minute focus group guide for [research objective]. Target audience: [description]. Topics to cover: [list]. Include warm-up, category exploration, [concept/stimulus] reaction, and wrap-up. Add probing prompts for each section.
View full prompt →Tip: Include any stimuli (ad concepts, product prototypes, packaging) in your description so the guide includes proper reveal and reaction sequences. Ask for a "moderator notes" version with guidance on handling quiet or dominant participants.
Three alternative versions of a "Key Insight" callout box for each finding — short, punchy, and written to stop a skimming executive in their tracks — so you can pick the best one for each slide.
Write 3 versions of a "Key Insight" callout box (2 sentences each) for this research finding: [paste finding]. Make each version punchy and memorable. Avoid passive voice and hedging. Write for a skimming executive, not a researcher.
View full prompt →Tip: Run this for your 5–8 most important findings. Paste the three options side-by-side on the slide while deciding — the contrast helps you see which framing lands hardest. Shorter is almost always better for callout boxes.
A flowing 400–600 word narrative that connects your key findings into a coherent story — complete with transitions, implications, and a clear "so what" — that you can use as the narration track for...
I have research findings from a study about [topic] for a [client type] client. Write a 500-word narrative connecting these findings into a clear story: [paste findings as bullet points]. The narrative should flow from context → key insight → implications. Use business language, not research jargon.
View full prompt →Tip: Give the findings in logical order — the AI will follow your sequence. If you want a more assertive "point of view" tone rather than a neutral reporting style, add "Take a clear point of view — don't hedge."
For each chart or data point, a punchy insight headline (the "so what"), two supporting bullets, and a callout stat — formatted so you can drop it straight into your deck.
For each of these research findings, write: 1 insight headline (the "so what" in one sentence), 2 supporting bullets, and a callout stat to highlight. Audience: [e.g., retail marketing team]. Findings: [paste chart data or findings]
View full prompt →Tip: Lead with the implication, not the observation. "Awareness is highest among 35-54s" is a finding; "Your core buyers already know you — the gap is conversion, not awareness" is an insight. Ask the AI to rewrite any headlines that start with a number.
A slide-by-slide outline for your entire presentation — with a headline, 2 supporting bullets, and suggested chart type for each slide — organized into a narrative arc that builds to a clear strate...
I have a [number]-slide research presentation for a [client type] about [research topic]. The key finding is: [one-sentence key finding]. Suggest a [number]-slide story structure. For each slide: slide number, headline, 2 supporting bullet points, and suggested chart type. Build toward strategic implications.
View full prompt →Tip: The story structure is a planning tool — build it before you open PowerPoint. If the recommended structure doesn't match your data, tell the AI "slides 4 and 5 need to swap because [reason]" and ask it to revise.
Use AI in your tools
AI features built into tools you already have
No new subscriptions, just features you may not have noticed
Set up an AI assistant
Step-by-step guides for dedicated AI tools
10 to 30 minute setup, then ongoing time savings
Go further
Advanced workflows, automation, and custom AI setups
For when you’re ready to connect tools and automate
Recommended Tools
5Ranked by relevance for research analyst
- 1
ChatGPT
Open-Ended Response Thematic Coding, Survey Questionnaire Generation + 2 more
Beginner - 2
Claude
Research Report Narrative Drafting, Executive Summary Drafting + 3 more
Beginner - 3
Perplexity
Competitive Intelligence Synthesis, Industry Background Literature Review Synthesis
Beginner - 4
Otter.ai
Qualitative Research Transcription & Summary
Beginner - 5
Beautiful.ai
PowerPoint Slide Generation from Research Findings
Beginner
Common questions
- What is the best AI tool for a research analyst?
- 1. ChatGPT: Open-Ended Response Thematic Coding, Survey Questionnaire Generation + 2 more. 2. Claude: Research Report Narrative Drafting, Executive Summary Drafting + 3 more. 3. Perplexity: Competitive Intelligence Synthesis, Industry Background Literature Review Synthesis.
- How can a research analyst use ChatGPT or another AI chatbot?
- Start with copy-paste prompts that work in any free chatbot. For example: A review of your draft survey questions flagging leading language, double-barreled constructions, acquiescence bias, and ambiguous wording — with suggested rewrites for each problem. A vivid, narrative persona profile for a market segment — complete with a name, motivations, frustrations, decision triggers, and a day-in-the-life vignette. A table coding each verbatim response to 1–2 themes, plus a frequency count of how often each theme appeared — ready to paste into your report.
- Do I need technical skills to start?
- No. Level 1 prompts work in any free AI chatbot with no signup beyond the chatbot itself: copy the prompt, fill in the bracketed details, and paste it in. Later levels add AI features in tools you already use, then dedicated AI tools and automation.
New to AI?
The Big Four AI Assistants
ChatGPT, Claude, Gemini, and Grok do roughly the same thing. Pick one and start.
Four Levels of AI Skill
From your first prompt to building automated workflows. Where are you now?
How to Keep Up with AI
The landscape changes fast. A low-effort system to stay informed without drowning.
We update this guide when the tools change. See what's changed →