★ Top-rated on Maven🔥 Featured in Lenny’s List
AI Analysis for Customer Insights
Your AI told you something you already knew. You wanted it to dig deeper, zoom out to connect the dots, or pin down the exact moments that make or break the experience. It didn’t. This course shows you how to get it there.
- Self-pacedStart any day
- 37 lessons23 videos
- 4 build guidesI walk you through each
- Any AI chat toolChatGPT, Claude, Gemini
4.8/5 from 100 reviews · Taught to 700+ students on Maven and with client teams
“This course is an amazing investment to immediately improve the quality and speed of AI assisted research analysis across LLMs. Reliable, rigorous principles and tactics that I’ll return to daily.”
Coming soon
$499
Early bird price. Ends Dec 1, then it’s $649.
Compare with the live course: $1,150+
- The full live course materials, updated October 2026
- Lifetime access
- 3 months of updates
Get the discount. Sign up and you’ll hear the day it opens.
Students from teams at
Is this you?
You have interview transcripts, survey results and customer messages stacked up from four different sources. Collecting it all wore you out, and the analysis hasn’t started. Stakeholders keep emailing: where’s the analysis, when can we share it, we need to decide about this now.
So you tried AI to move faster, and what came back didn’t hold up. Vague themes. Quotes pulled out of context. Findings no customer ever said. You can’t shake the worry that it missed the one thing you would have caught, so you redo the analysis by hand anyway.
I’ve been there too. With the right instructions and a system around them, AI stops being a gamble. Your analysis gets faster and more complete, and you can show your work.
Where you start
You use AI on customer data, but you don’t fully trust what comes back.
Where you finish
You know how to get insights you can trust from any LLM, in any interface, with workflows you can put together your own way.
Better analysis this week. A playbook for the year.
The course is built to show you improvements in the same week you start, and to support you over many months.
This week
Build and fix, today
Start with your own analysis process and the AI you already use. The first build guide gives you a context brief you can use on your next analysis, so the difference shows up in the first week.
All year
A playbook you come back to
Surveys, interviews, user tests, stakeholder write-ups: when a new kind of analysis lands on your desk, the workflow for it is already here. It supports your insights work for the rest of this year and into next year.
Self-paced, but you won’t be guessing
This isn’t a live cohort. There are no scheduled sessions and no 1:1 support. Start any time and go at your own speed. Two things make it feel like I’m there anyway:
Build guides
Each week has a build guide where I walk you through making one thing step by step (recorded). You see what good looks like before you try it on your own data.
Real student questions, answered
Over two years of cohorts, students asked me the same questions at the same points. Those questions and my answers are part of the course, so when you get stuck, the answer is usually already written.
Is this course for you?
Tick what’s true for you.
You’ll need: any AI chat tool (Claude, ChatGPT or Gemini) and your own interview or survey data from work.
Your answer shows up here.
This course is built for you.Get the early bird price ↑
Claude Code for Customer Insights fits you better.See that course →
What you’ll be able to do
Cut your analysis time in half and find stronger insights, without giving up quality or trust in your work.
AI analysis systems that run 10x faster
- Repeatable AI workflows that hold up on relevant, reliable insights
- Frameworks for finding where your analysis slows down, and where AI pays off most
- Agents that connect your tools into one multi-step insights workflow
Speed without giving up quality
- AI instruction tactics that pull out the details in an analysis that matter to you
- Instructions that tell AI what you need, so days of tedious work turn into minutes of focused work
- Quality checks placed where they catch the most and cost the least
A partner for reducing risk
- AI that challenges your own assumptions and sharpens your judgment
- Hidden bias caught on both sides, yours and the model’s, before it reaches your findings
- Bias checking as a systematic step that takes minutes, instead of something you do when time allows
Hidden opportunities and a point of view
- Output that goes past thematic synthesis to a stance you can defend
- The opportunities sitting in your data that one person reading alone would miss
Fast tool experiments
- My framework for running consistent experiments with any AI tool
- Benchmark tests you rerun inside free trials, so your read on each tool stays current and stays free
Confidence in your AI decisions
- Hundreds of my experiments on when and how to apply AI to research, so you skip repeating them
- A clear line on which tasks you can drop entirely
- The standing to be the person your team or client asks about AI for customer research
What’s included
- Recorded tutorials
- 4 build guides, each ending in something you keep using: a context brief, a run sheet, a codebook and your own Analysis Copilot
- Answers to the questions cohort students actually asked
- An AI instruction library: tested workflows and prompts to copy or compare against
- Lifetime access to the course content
- 3 months of content updates
- Certificate of completion
What’s inside
Week 1 · Communicating with AI
- Where AI works best in your analysis
- The context your AI needs, and where to put it
- Context windows and context rot
- Prompting skills that still matter
- Teaching your AI how experts think
- Guided exercise: pull your analysis process out of your head
- Build guide 1: your context brief
Week 2 · Running interview analysis
- FINDSIGHT: nine principles for analysis you can trust, whichever AI you use
- Analysis step by step, with multiple lenses
- Hunting for opportunities in your data
- Tested workflows: when more than one prompt pays off
- The rules for building an analysis workflow you can trust
- AI as your devil’s advocate, plus a bias audit
- How to know your AI analysis is actually right
- Build guide 2: your run sheet
Week 3 · Surveys and insights
- Preparing survey and short-feedback data
- Teaching AI to code responses, with counting done in code instead of guessed
- The full survey workflow: coding, counting and verifying
- Large data sets and troubleshooting
- Build guide 3: your codebook
- Synthesis and prioritizing insights
- Personalizing insights for each stakeholder
Week 4 · Your analysis copilot
- Agents vs. automation for analysis
- Designing agents that work
- Build guide 4: your Analysis Copilot
- Analyzing user tests and video
Reference, yours to keep
- The prompt library (October 2026)
- The Common Questions Archive
- FINDSIGHT principle cards
What students say
I gained confidence on applying AI, as well as ideas for new areas where AI could be applied. I’ve struggled with some sort of mental bias or imposter syndrome when it comes to heavy AI use, and this course alleviated that.
She provides excellent guidance…suggests approaches to take to have the most impact in the shortest time. Many times, I’ve been very thankful we’ve been able to lean on Caitlin’s experience & support rather than trying to re-invent the wheel and wasting time.
This course is phenomenal. Caitlin packs so much content into the written and video lessons. I knew within the first few days that it was 100% worth the money. This course has completely changed the way I approach using AI throughout my analysis process. This is an 11/10 course.
Every single time, I’m able to say to myself, “Ok. Great. I know exactly what to do now. Let’s go.”
Great course with lots of content! I’m eager to apply what I’ve learned in my everyday work and can highly recommend the course to anyone looking to upskill themselves in using AI for research.
This course has changed the way I think about using AI in the analysis and synthesis process. I have a lot of experimentation ahead of me to bring my new vision to life, but I’m way ahead of where I was 4 weeks ago.
This course offers a systematic process to structuring qualitative analysis using AI tools. There is a ton of content that covers all manner of things from analysing written interview transcripts and survey responses, to analysing usability test videos.
I loved this course and learnt so much. I found Caitlin’s teaching style to be exactly right for me and enjoyed the level of detail she provided for all the different modules. The course was well structured and well paced.
Get it the day it opens
AI Analysis for Customer Insights is coming soon at $499 (early bird price, ends Dec 1). Sign up and you’ll hear first.
Questions people ask before joining
How do I know the AI’s findings are right, and that it didn’t miss or make anything up?
This course builds the checks into the workflow. Every quote gets traced back to the source, you confirm the AI read everything instead of skimming, and you compare the findings against your own notes before anyone builds on them. The AI’s own check counts as a first pass only, because models tend to confirm their own output, even quotes they’ve slightly changed.
Will this course work with the AI tool my company approved?
Yes. What you learn works in any LLM, including ChatGPT, Claude, Gemini and internal company versions like Microsoft Copilot. Check with your IT or legal team that you’re allowed to put research data into the approved tool.
How much time does it take?
Plan for about 3–5 hours a week, plus a little time to try it on your own data at work. It’s self-paced, so you start any time and go at your own speed.
I’m a PM or designer, not a researcher. Is this for me?
Yes. This course is built for anyone who works with customer data, whatever the job title. Past students have been PMs, product designers and UX designers, as well as insights-focused pros.
Which course should I take?
If you want results you can trust from the AI tools you already use, without Claude Code, start with AI Analysis. If you want to turn that into repeatable workflows and agents in Claude Code, take Claude Code for Customer Insights. Still unsure? Email me at caitlin@aicustomerresearch.com and I’ll tell you which one fits.
What’s your refund policy?
14-day refund. If the course isn’t right for you, email within 14 days of purchase for a full refund. No questions asked. Refunds end course access.
Can I buy seats for my team?
Yes. Course seats are 20% off for 2+ people and 25% off for 10+, and larger teams can book a private cohort. Email me at caitlin@aicustomerresearch.com with your team size and what you want to get out of it.
Something else? Email me at caitlin@aicustomerresearch.com