Caitlin Sullivan

Cohort 2

Cohort 3

Cohort 4

Cohort 7

Cohort 5

Cohort 1

Cohort 6

Cut through the AI tool flood. This free database maps 140+ AI tools just for customer research in one place— <ul><li>Tagged by task</li><li>with quick differentiators </li><li>and data-storage notes </li></ul> So you can shortlist fast without overwhelm. Grab it if you’re tired of the "did I miss a tool for {task}?" FOMO and trial-and-error. After download, you'll also get my AI Customer Research newsletter to learn to 10x how you use AI for discovery immediately (for free).

140+ AI Tools for Customer Research: Find the Right One Fast

You pasted customer feedback into ChatGPT. It gave you a beautiful summary—five themes, neatly prioritized. And now you're thinking: <ul><li>Is this *everything* - or did it miss something?</li><li>Can I actually present this - or did it make things up?</li></ul> You can't trace the priorities back with certainty.You don't know what the AI ignored. You can't answer "how confident are we?" This cheatsheet gives you 4 approaches and copy-past prompts that fix that—analysis with real counts, cited evidence, and traceable reasoning. More trustworthy insights in 24 hours 👇

The PM's AI Analysis Cheatsheet: Copy-Paste Prompts for Insights You Can Actually Trust

Cut analysis time in half and uncover more powerful insights—without sacrificing quality or trust in your work.

Ex-Head of UXR, Spotify Business/SYB

Ex-Head of User Research, Spotify Business. 2000+ Hours Testing AI for Research.

AI Customer Research Analysis for Designers, Researchers and PMs

Principal Service Designer

Senior UX Researcher

Director of Research

UX Researcher

Head of User Research

Independent researcher

Lead User Researcher

Founder

Strategic design consultant 

Design/Product

‼️ Revamped 2026 Curriculum + format. Price will go up after March 1st 2026. Includes content updates from 2 future cohorts! Is this you? You’re sitting in front of a mountain of raw feedback—interview transcripts, survey results, and customer messages from multiple sources. Your brain is fried from collecting it all, but the real challenge hasn’t even started yet: turning that pile into clear, actionable insights. Stakeholders are emailing:👋 “Where’s the analysis?”👋 “When can we share this with the team?”👋 "We need to decide about [product change] NOW." -You've been trying to use AI to speed things up, but it gave you results that didn’t feel reliable.- Vague themes, out-of-context quotes, or flat-out fake "user insights."- You have a constant fear that it missed something critical that *you* would have found.- And after all the time spent, you decide to re-do the analysis manually anyway.-I’ve been there, too. 🙋‍♀️If you’ve felt stuck in endless ChatGPT loops, redoing AI’s work, or wondering if there’s a better way—this is for you.With the right knowledge and systems, AI stops being such a gamble and becomes an ally you can trust for faster - and more complete analysis.

2026: AI isn't optional. It's a core tool - and most people haven't mastered it.

Learn AI Analysis Systems That Deliver 2-10x Faster

Gain Speed Without Sacrificing Quality

Develop a Partner for Reducing Risk

Uncover Hidden Opportunities + POVs

Run Efficient Tool Experiments

Get Confident About AI Decisions 

You’re expected to deliver actionable insights for product decisions fast, even when research isn't even your "main" job.

You're a Designer, Researcher or Product person spending hours digging through customer input, searching for insights that matter.

You’ve tried AI tools, but the results feel “meh.” Instead of saving time, they create more work -- and you're back to doing it yourself.

Get from scattered prompts to semi-automated systems that run with minimal time from you—and run evals to find what works best.

Trains teams at YouTube, Canva, Maze +

YouTube

Canva

Figma

Ramp

Ex-Head of UXR in Product Team - Spotify Business, Advisor, Top 100

In 2 weeks, you'll go from Claude running superficial scans of your discovery data to rock-solid decisions from continuous workflows. No missed hidden gems. No hallucinations you didn't catch. Just insights you can actually bet your career on. From workflows that let you gather customer signal on repeat. Free Claude Code tutorials teach you how to use the tool to go faster. But faster can also = superficial. This course teaches you to build insight systems that are fast, validated, and run with minimal effort. -- You'll build systems that: - Have trustworthy verification built-in - Apply your own style and process that you know works - Minimize unnecessary context window/token use - Deliver customer signals in half the time

10x faster insights—that are actually right.

Build systems for insights that are FAST *and* RIGHT

Connect + update everything, export nothing

Know what works -- and prove it

Ship with confidence

You have a computer you're allowed to install Claude Code on and give file access to (Mac, Windows, or Linux).


You're allowed to use Claude Code on your computer

Claude Pro or Max subscription (from $20/month), or Anthropic developer console account with credits (pay as you go).

You have Claude Pro or Max

You'll get the most out of the course if you can practice with some of your own data - and even connect to sources like Google Drive, Notion

Recommended: Your own data

You're a PM who's watched free Claude Code tutorials--but know you can't trust their superficial synthesis. You need systems for big bets.

You're a Researcher or Designer who want to 10x your analysis speed without sacrificing reliability.

You're making product decisions from qualitative or quantitative customer data<ul><li></li></ul>

Claude Code for PMs: Data to Decisions Workflows

Ex-Head of UXR at Spotify Business/Soundtrack ᐧ Advisor ᐧ AI Customer Research

Stop slow, basic synthesis with big holes. See how a system can map your evidence before every product decision - fast.

Audit all your customer data with minimal commands

The riskiest assumptions are the ones your team never wrote down. See how this workflow finds them automatically.

Find the hidden assumptions that could kill your launch

Don't guess what to validate. My workflow ranks risks and gaps, then builds an expert test plan automatically.

Get a prioritized test plan generated from your data

Synthetic Users for Product Discovery: What People Miss

Ex-Head of UXR, Spotify Business/Soundtrack ᐧ Advisor ᐧ AI Customer Research

You have behavior data, customer calls, surveys... But not all data is right for synthetic users. Know what to use.

Not all data works for synthetic users

Not every research question fits. I'll show you which use cases can deliver real value — and which are a gamble.

Where synthetic users actually work (and where they don't)

Academic studies did specific things most Product teams skip. Learn what, and how to apply it to your own data and setup

What big studies did that you're not replicating

AI Customer Insights: Prompt Fixes + Top LLMs for Analysis

Ex-Head of UXR at Spotify Business/Soundtrack. 550 Hours Testing AI for Insights

I’ll show you how different models perform across parts of the analysis process, so you know exactly which work best.

Demo: What Different Models Do by Default for Analysis

See real examples of "okay" vs expert prompts, and learn how tiny changes lead to clearer, more accurate insights.

Prompt Tweaks That Sharpen Your AI Analysis

Discover simple ways to work with your current LLM’s defaults and quirks for consistently better output.

Get More From the Model You Already Use

Claude Code for User Insights: What Agents Can Do

Ex-Head of Research at Spotify Business/SYB ᐧ AI Advisor for Customer Insights

I'll run live analysis: Premium, free + churned users — three agents analyze in parallel. Final step tests reliability.

DEMO: A multi-agent workflow run end-to-end

Breaking analysis into parts changes  output quality. I'll show you what each agent handles, and why this beats one chat

What each agent does (and why the chain matters)

Agents aren't for every task. I'll tell you where this workflow shines — and when a simpler approach is enough.

Where this approach is worth it (and where it's overkill)

3 Tactics for PMs: Immediately Better AI Analysis

Ex-Head of User Research at Spotify Business/Soundtrack. AI Advisor.

I'll use the same prompt and transcripts across 6 models. See why we get wildly different results and counts - and a fix

Why AI analysis outputs feel so inconsistent — and a fix

Same feedback messages, same prompt, completely different themes. Small additions solve this if you know why.

Categorization that's finally consistent across runs

Most people upload and prompt. I'll show you what to do first — it takes 30 seconds and prevents entirely wrong analyses

Caitlin Sullivan

Caitlin Sullivan

Courses

Lightning Lessons

Resources

Reviews