You’ve got your notes — now what?
So you’ve talked to your end users and learned a lot. You’re excited to implement what you’ve learned! But first, you need input and buy-in from the team and leadership, which means you have to analyze, synthesize, and communicate your findings.
You take a deep breath and assess your assets. You have video recordings of the interviews, the transcripts that your conference tool of choice generously provided (although they’re not perfect, they hit the main points), and your own scribbled notes.
You may feel a bit alone and overwhelmed, especially if this is your first time in this situation. It’s natural to feel torn between wanting to do justice to what people told you and trying to make sense of everything within the scope of the project. How will you gather all the information, both subjective and objective, along with its associated context, and present it in a digestible form for your team while maintaining complete accuracy?
Enter: The AI intern
What happens next is not what you expect.
Outside your office sits a jaunty, well-groomed, eager young intern with a slightly spacey smile. He (or she, or they… pronouns seem indeterminate to this entity) looks at you earnestly and asks, “Can I help you with this?”
Of course, you’re taken aback. Your connection with the interview participants feels almost intimate and personal, and you feel the weight of telling their stories just right. So your initial response is a slightly stiff, “No thanks, I got it,” mainly because you don’t know quite what “it” is, and you can’t ask for help if you don’t know what kind of help you need. So I’m here to tell you: Invite this (thing, robot, whatever) in and accept their help – not to take over, but to nourish and strengthen you while you carry the flame. Your first response can actually be, “I don’t know; how can you help me?”
Working together: Your prompting loop
Putting personification aside (because AI is not – and will never be – a person), ChatGPT, Claude, Gemini, NotebookLM, TetraBot, or whichever tool you choose, can be a great research assistant for organizing your results.
Each tool works differently, but your basic steps will be the same:
Provide an initial prompt. Numerous online resources can help you with this if you’re new to it, but an easy way to think of it is Context + Details + Intent + Format. Don’t worry; if you forget one, the AI will ask you for it.
Provide the data. Depending on the tool and your subscription level, your data can be more or less raw, and more or less extensive (you get what you pay for). But even free tools can handle a few interviews easily.
Review the output. You already told it what you wanted in your prompt, but you may have made assumptions or left certain aspects open to interpretation. Usually, the first output isn’t exactly right; how could it be, when it can’t feel what you felt in the room?
Refine your prompt. This is where the co-creativity begins, and your curiosity, experience, and mental sparks come back into play. Even if the initial output gives you exactly what you asked for, the output may inspire you to add something additional, or focus on a new angle you hadn’t seen.
Review the output.
Refine your prompt.
Review the output.
Refine your prompt.
Review the output… until you’re happy with it. Some might call this a conversation, and it sure feels like teamwork when you get going! But it’s your reputation on the line, not the bot’s, so own the outcome, and make it something you can stand behind.
Export the output. The AI can help you with the format, but you’ll still need to present and share it somehow. So whether it’s a corporate template, an email, or an old-fashioned binder, here’s your chance to massage the output into the deliverable best suited to your team — something clear, grounded, and gracious toward the people who shared their stories with you.
Back to you: What only humans can do
And now, just when you think you have a new best friend who will help you with all your work, tear yourself from that super-powered data processing machine. Presenting your findings, listening critically and empathetically to your teammates, establishing credibility through broad expertise and contextual understanding, adjusting to the immediate human factors in the room, relating personal stories… all these activities fall squarely on your own darling, oh-so-human shoulders.
As expressed by UX experts at the Nielsen Norman Group (NN/g):
“Uniquely human abilities are the future of our field. Critical thinking, creativity, and taste — the ability to discern and curate a series of outputs and decisions — will become the differentiators.”
Thanks to the nerdy intern sitting at your desk, staring blankly into your absence, you reached this point sooner, with mental energy left to take the next steps. All that and there’s even cold brew left in the fridge.
Posted on July 31, 2025 by Lamar Goodenough in Commentary
You’ve got your notes — now what?
So you’ve talked to your end users and learned a lot. You’re excited to implement what you’ve learned! But first, you need input and buy-in from the team and leadership, which means you have to analyze, synthesize, and communicate your findings.
You take a deep breath and assess your assets. You have video recordings of the interviews, the transcripts that your conference tool of choice generously provided (although they’re not perfect, they hit the main points), and your own scribbled notes.
You may feel a bit alone and overwhelmed, especially if this is your first time in this situation. It’s natural to feel torn between wanting to do justice to what people told you and trying to make sense of everything within the scope of the project. How will you gather all the information, both subjective and objective, along with its associated context, and present it in a digestible form for your team while maintaining complete accuracy?
Enter: The AI intern
What happens next is not what you expect.
Outside your office sits a jaunty, well-groomed, eager young intern with a slightly spacey smile. He (or she, or they… pronouns seem indeterminate to this entity) looks at you earnestly and asks, “Can I help you with this?”
Of course, you’re taken aback. Your connection with the interview participants feels almost intimate and personal, and you feel the weight of telling their stories just right. So your initial response is a slightly stiff, “No thanks, I got it,” mainly because you don’t know quite what “it” is, and you can’t ask for help if you don’t know what kind of help you need. So I’m here to tell you: Invite this (thing, robot, whatever) in and accept their help – not to take over, but to nourish and strengthen you while you carry the flame. Your first response can actually be, “I don’t know; how can you help me?”
Working together: Your prompting loop
Putting personification aside (because AI is not – and will never be – a person), ChatGPT, Claude, Gemini, NotebookLM, TetraBot, or whichever tool you choose, can be a great research assistant for organizing your results.
Each tool works differently, but your basic steps will be the same:
Provide an initial prompt. Numerous online resources can help you with this if you’re new to it, but an easy way to think of it is Context + Details + Intent + Format. Don’t worry; if you forget one, the AI will ask you for it.
Provide the data. Depending on the tool and your subscription level, your data can be more or less raw, and more or less extensive (you get what you pay for). But even free tools can handle a few interviews easily.
Review the output. You already told it what you wanted in your prompt, but you may have made assumptions or left certain aspects open to interpretation. Usually, the first output isn’t exactly right; how could it be, when it can’t feel what you felt in the room?
Refine your prompt. This is where the co-creativity begins, and your curiosity, experience, and mental sparks come back into play. Even if the initial output gives you exactly what you asked for, the output may inspire you to add something additional, or focus on a new angle you hadn’t seen.
Review the output.
Refine your prompt.
Review the output.
Refine your prompt.
Review the output… until you’re happy with it. Some might call this a conversation, and it sure feels like teamwork when you get going! But it’s your reputation on the line, not the bot’s, so own the outcome, and make it something you can stand behind.
Export the output. The AI can help you with the format, but you’ll still need to present and share it somehow. So whether it’s a corporate template, an email, or an old-fashioned binder, here’s your chance to massage the output into the deliverable best suited to your team — something clear, grounded, and gracious toward the people who shared their stories with you.
Back to you: What only humans can do
And now, just when you think you have a new best friend who will help you with all your work, tear yourself from that super-powered data processing machine. Presenting your findings, listening critically and empathetically to your teammates, establishing credibility through broad expertise and contextual understanding, adjusting to the immediate human factors in the room, relating personal stories… all these activities fall squarely on your own darling, oh-so-human shoulders.
As expressed by UX experts at the Nielsen Norman Group (NN/g):
Thanks to the nerdy intern sitting at your desk, staring blankly into your absence, you reached this point sooner, with mental energy left to take the next steps. All that and there’s even cold brew left in the fridge.
Tagged With These Topics
About Lamar Goodenough
Lamar plays a leading role in multiple web projects for healthcare organizations, especially medical certification boards. Her focus is the user’s experience and how it can be improved while meeting client objectives. She specializes in end user research and converting findings into actionable online improvements.
Related Posts
UX Research Starter Kit
Post by Lamar Goodenough
5 questions that will shape your UX research efforts
Should I Edit This Down or Do I Just Need a Bigger Hat?
Post by Tony Rose
How long does a sentence need to be? It depends how long you need it to be.