So What? How to Analyze Survey Data with Generative AI

Hey, I don't hear you. I hear you. Katie's here. She was just here, too. I don't know what's up with that because we were just talking to her.

Oh, there you go. Oh my god, I'm here. I promise I'm here. I'm gonna I'm gonna do the whole show with like, I don't know, signals or something. Mine.

Let's start again. Happy Thursday, everyone. Welcome to So What The Marketing Analytics and Insights Live Show. I'm Katie, joined by Chris and John. Take two.

How's it going, fellas? Excellent. If you ever wondered if we're really human, the answer is yes. Well, and Chris, you're on the wrong mic, so we're off to an excellent start today. Oh my god.

There we go. Woo. This week, we are walking through how to analyze your survey data with generative AI. We did this a few years ago, but so much has changed. And one of the things that we're really striving for in 2026 is making sure we're showing you both lower tech, easy entry solutions, but then also the higher tech, more advanced solutions, because there's room for both of them.

And we know that our audience, and then also internal to our team, there's a need for both versions. So Chris was showing me this earlier, and we actually uh on the podcast this week. If you want to go to TrustInsights at AI slash GI podcast, Chris and I spent a decent amount of time talking about the low-tech solution that you can use now with your Google Sheets, which I think is very exciting. I think it's a really good game changer for people who are like, I have all this data, I don't know what to do. But then we can get into sort of the bigger, more advanced stuff.

So, Chris, where would you like to start this week? Um, we should probably talk this through. So, this is going to use Google's Gemini and Google Sheets. I uh we are not paying subscribers to Microsoft Co-Pilot. My understanding is that Copilot fundamentally does about the same thing in the current version of Excel, and it uses co-pilots' models on the back end.

And for what we're doing today, that's okay. I mean, I have a lot of uh ranty pants things to say about uh co-pilots models, but that's for another show. That's for another show. It totally is, or at the bar, you know, at the next event. Um so first things first, we are going to need uh to get our data, of course, which comes uh in our case.

This is coming straight out of uh we run our surveys on Gravity Forms in WordPress. So it's just literally a plugin for WordPress. Someone comes in, fills up the form. Uh, I've removed the personal identifying information so that we're not sharing people's private emails and stuff on on the air. And what we've got here is just a spreadsheet uh where every column in every cell in Rosie is an answer.

Some of them are multi-paragraphs, some of them are single paragraphs. Now. Can I back up even a step further from that to sort of share where this data came from in the first place? So we have tried, and you know, to be fair, I have been a little remiss in getting this done. But one of our goals is to run a one question survey every quarter.

And the goal of that survey is to help us understand what's going on with our audience, what's going on in the industry, what's going on with people who are taking the time to read our content. So it's the Trust Insights one question survey, still live. You can answer it at any time. And our question this quarter was if you could attend one deep dive webinar in 2026 to solve your biggest business or marketing challenge, what specific topic would you want to cover? And the reason we asked this question, not only for our own uh, you know, planning purposes, but the topic of AI is obviously very prevalent.

Everybody's talking about different versions of it, what do we do with it? But it's not enough for us to just say, okay, here's a webinar on how to do AI. Like that's too broad. So we wanted to know specifically if you could deep dive into anything. We didn't want to limit it to AI or just marketing, but anything.

And we got a wide variety of responses, very helpful responses, because there's still a lot of people wanting to know about just digital marketing that doesn't include AI or measurement or KPIs or organic search or you know, a lot of topics that aren't necessarily AI, which I found to be incredibly helpful. But unsurprisingly, majority of our data is around what do I do with a gentic? How do I do this? How do we do that? And so that gives us a really good understanding of how we as a company can better serve our audience, making sure we're giving them what they're asking for, not just what we think they're asking for.

So I just want to set that context of here's where this survey data came from. Exactly. And that's that's a great segment because you can see this is a paragraph, right? This is a big text box. So you can put whatever you want in there.

It's not multiple choice, it's not uh anything. And by the way, slight sidebar. This is the way we should be doing surveying now in 2026 because the days of a static drop-down that you have to you know check a box or whatever and things are gone. Generative AI, language models are very capable of parsing language, so there's no excuse anymore to say I, you know, for eight data analysis, we can only have a drop-down with these eight choices. No, let people say what they want.

You know, you can use language models to understand that language. Drives me up a wall when I say, How'd you hear about us? There's five choices and they're all from choices from the 1990s. Like, no, you know, I haven't gotten a postcard that I remember quite a bit. News slash print, email, referral, radio search, radio.

Oh, I miss radio ads. They were good. Anyway, let's go on. Um, so we take those responses, I mean put them into a Google sheet. Now, here's what's new in pretty much everybody's office software.

You can now invoke a language model like Google Gemini right inside of a sheet on a sell-by-cell basis. The formula for this in Google Sheets, unsurprisingly, is Gemini. And you have it, there's a prompt, and then you give it a range. Because we're processing individual survey results, we don't want to use a whole range of literally cell by sell. But before we do that, we have to know what we're analyzing.

We have to figure out like what are the things that we would want to know. So, Katie, what are some of the things that you might want to know out of these survey results? Um, I want to know the large buckets of categories of how people responded. So, yeah, topic category, but then within then I would want to know within each of those topics, sort of what the subtopics are. Um, because like, so let's say the topic in category is agentic AI.

Well, that's a pretty big topic. So I'd want to know within that, you know, people are wanting to know like how to's or you know, what is, you know, so that we can make sure we're appropriately answering those questions. Right. So let's call it this topic, let's call it a subtopic. What else would you want to know?

Um, I think you had put in, which I thought was interesting, um, and an earlier version, sentiment. And so sentiment is basically at a high level. It's is this a positive thing? Is this a negative thing? In this context, I think it's an interesting data point.

I don't know how necessarily useful it is. Um, but I think at the very least, like we can say, oh, this is what it is, but it then it's not necessarily useful, and we can always take it out. Yeah, I think the reason I put that in is because I might want to know, you know, do people feel differently about agentic AI than they do, say, regular AI? Do people feel differently about analytics? Like, oh, I hate Google Analytics for like, okay, clearly this is still a very sore topic with people.

John, is there anything you would want to know out of this data? You know, the thing that I just love about this is because I've been on the other side of it so many times where you do these surveys where you get 10, 20,000 results in and most of the time it just goes to a bunch of bar graphs. Like nobody pays any attention to the individual stuff. So yeah, sentiment is the number one thing for me because there's never been an easy way to you know more easier than it is today to just say hey, you know, categorize them as far as how many are the top tier champions that love us to death and you know show me the five most horrible ones and all those kinds of querying tools are more powerful than ever before. So I love all that stuff.

The other one is to find all of those things and issues that you just can't do if you manually go through the results. You know like uh for software stuff you can say show me the three features that people have the biggest problems with and show me the four you know features that need to be fixed or added you know to be to be able to mine it for that kind of stuff all that is just that's the most exciting thing going in AI for me right now. Yep. Here's a question as our biz dev guy would you be interested in a like maybe a zero to five scale of urgency where somebody's like ah just curious about the same whereas I want to buy something right now. Yeah that's a good point.

I mean, you know, and Katie has done a good job with the fact that you know that we don't just send the survey out and then I start pummeling people the next day cold coming out I know it's tempting, but please don't you know, there's a lot of, you know, well, oh, you want that. What does it take to get in this car today? But we don't have to, you know, we don't want to jump all over them. The big thing is to yeah, but yeah, there always will be too. That you know, I say that, but on the other side of my mouth, a lot of times when you get big enough, there will be four or five people per thousand that like it just happened to hit them on the day that yeah, we're looking to buy right now, we're in the middle of the purchase.

So, and then that you know, if somebody raises their hand and like, yeah, I'm actually looking to buy, those people do get a call back right away. So, um, but yeah, yeah. Knowing they're he, I guess you could qualify all of the you could do all of band too. You know, it'd be great to be like, hey, segment these results by budget, you know, let me see them by company size. Is there stuff that the bigger companies like or don't like that's totally different from the small ones?

Um, authority too, you know, did are we talking to a bunch of weaklings? Are these the decision makers? Um, and then you know, need how much do you do you like this stuff? And uh and timeline too, you know, how many people are talking about doing something in zero to three months versus over the next four years? So all that stuff is worthwhile.

Okay. So I think those clicky. I was gonna say, I think for today's purposes, I think urgency is good, but I think even just picking up on um almost like a yesno column of, you know, did there did they say in there, can you help me? Or can you do this? Or, you know, so sort of looking for those cues because you know, we have a couple of hundred here that we can sort of go through manually, but it would be helpful to know uh, sort of even just sort of like to sort down to a smaller list, you know.

You know, did they say immediately, like, oh, and do you do this, or can you help me? Or, you know, is this something I can buy from you? Like that's and when you're getting a lot of data, being able to pull that out and immediately give it to someone like John who's like, give me more people to talk to, like, happy to do that. Because if they raise their hand and say that, that's great. And we've done that on previous surveys where we've added an additional question of, and do you want us to reach out to you, you know, for this service?

A hundred percent of people said no, which is why I stopped asking that question. I was like, okay, that's pretty clear. They just want to give us their feedback, they're not looking to get pitched. Exactly. Okay, so we've got two categorical columns, topic and subtopic.

Um, we have a binary column, intent, yes or no, and we have two numeric columns, sentiment urgency. Now we have to actually build this thing. So here's how we're going to do this. First things first, we need to figure out what those topics and subtopics are. So for this, we're going to use regular Google Gemini.

And the reason we're going to use regular Google Gemini is because regular Google Gemini, A, you can specify a the the most powerful model, which is the Pro model, and B, you can give it the entire data set all at once to think through broadly what are those things. So what we would say here is, well, first we need to give it the file itself, which would be uh a useful thing to have. So let's go ahead and drop in the file. Even though this is a CSV, it's only got one column, and that's the the text of the responses. So that's perfectly safe to use here.

And we're gonna give it a prompt. We're gonna say, read through the corpus of survey results attached. What we want to understand are the five to six major topics in the results, plus two to three subtopics per major topic. Present your results as a markdown list with the major topic on the first level and the subtopic on the second level. And so uh we want Gemini to to try to think this through.

We do not have to get exact counts here. And in fact, that would be a terrible idea because a straight language model like this can't count. They're incapable of doing math. Well, that's why we're using Google Sheets. But we it absolutely can explore the text space around all these freeform responses.

This, by the way, is what people should be using language models for is to take that unstructured, you know, uh freeform text and distill it down so that you can then apply categories to it. So let's see how Gemini is doing here. Oh, actually, it looks like it was writing some of its own code along the line. Uh you can see that because it's popped a little code bubble there. But uh, based on the survey responses, here are the major topics and subtopics identified in the results.

AI strategy and organization adoption, change management, AI governance, strategic planning, AI agents on automation, building and deploying, workflow automation, technical execution, marketing strategy and analytics, attribution reporting, content strategy and SEO, role-specific assistance, practical AI implementation, aka vibe coding, coding and app development, tool integration, practical use cases, major tool ecosystems, co-pilot workspace, notebook LM, and business operations on scaling, small business, knowledge management, and operational efficiency. So, Katie, does that seem to fit with what you've read in the results? It does. Um, and I, you know, even at a high level, this to me would be incredibly helpful. I'd be like, I could take this, just this, and build out a content calendar.

But I'm excited to see, you know, where we go next with this. But yeah, so far, because unsurprisingly, I've you know been paying attention and read every single result that has come in and have been mentally banking them. So it's really nice to see that that's what Gemini is also coming up with is in terms of categories. Right. So now what we want to do is we want to start assigning categories to do this.

If you want to save your sanity, uh please start a new sheet. We're gonna call the sheet prompts. And in the first column, we'll have uh the prompt name, and the second column will have the actual prompt. And we're gonna have a category for a topic, subtopic, we're gonna have one for sentiment, we're gonna have one for urgency, and we're gonna have one for intent. We should rename that rearrange that so it spells the word suits.

Um it's no acronyms today. We're not we're not developing anything new. Surprise. Oh my goodness. I can't.

Now, here's the catch with um Google Sheets. If you're used to writing very long prompts, they have to fit in a cell. And cells can occupy a lot of text, but they can't have things like line breaks in them, which is highly problematic. So, what we're gonna want to do here is have Gemini help us write the prompt that will fit inside one basically big honking paragraph, and that is what it can then reference in the sheet. So let's give let give you a concrete example of this.

I'm gonna start off by saying I'll start a new document here. Let's classify the text in the cell based on the major topics provided. The major topics are the topic headings. Generate and return just the major topic. And so that's our first part of the prompt, and our second part of the prompt is going to be what those major topics are.

And we're going to copy and paste that. Now here's what we're going to do to make this work in Google Sheets. I went ahead and I created a gem called the Google Sheets Gem. And all it was going to do is here is the prompt to optimize. It's basically going to read through this and figure out how do I put this all in one paragraph so that Google Sheets will not blow up when it when we get it back.

And it will have to think this through, obviously, to make sure that it is intelligent. But once it's done, we can then put that in the sheet and then reference it. So it's thinking through the taxonomy, it's thinking through all the pieces. And it should, if I did it right in the gem, come up with two different options. And then we can decide which of the two options makes the most sense.

As a pro tip, don't ask gem, don't ask any AI to give you the thing, because it will only give you the most high probability. Ask it to give you options so that it forces the model to think a little more broadly. One of the pro tips that we had talked about last year, which um, you know, perhaps we've sort of debunked it, is um, you know, what questions do you have, or ask me one question at a time until you have you know a satisfactory amount of information. Is that still in this a tactic that you would employ, or are you really just looking for options? Uh for this specific use case, we're asking it just for the options because the gem is like four pages long of.

So I use those tips for like ask me one more question to build the gem itself. Got it. Okay. And that's a helpful distinction. Because I think that when a lot of us here, like, oh, and here's a pro tip, and here's the thing, we're trying to employ all of them at once, and it gets very overwhelming, and then the prompts themselves become very cumbersome.

Yep. Okay, so now we've we've got our topic prompt, and we now need to generate the subtopic prompt, uh, which is going to look very similar. Um, we're gonna say, let's classify the text in the cell based on the major topic in the first column and choose which of the subtopics is the best fit for the survey response text. Um so we're gonna take the exact same thing, start a new chat with our gem, and put that in and let it do its thing. While it's gonna while it's doing that, I'm gonna wire up this first one.

So we've got our prompt, it's in cell B2. So let's go here to our topic. We're gonna go Gemini prompt is in prompts $, sign two, and then the range is just E2. And yes, I know Jim and I can make mistakes. And so it has now executed that prompt.

And this is why we want the prompts in a separate sheet, because if I were just drag and drop that prompt uh all the way down this entire sheet. Uh this would have that prompt repeated over and over and over again, and nobody wants that. It would take forever and a day for it to do that. So instead, we reference the prompt on the other sheet, and it it does this whole thing for us very nicely. So here in our first pass through, we have the topics themselves.

Now let's go back to Gemini. And it has come up with a a second set of instructions. So we're gonna put this massive gigantic huge blob into the subtopic. So now we have the subtopic prompt. Let's get the third prompt, our our um sentiment prompt baking while it uh while we get this second one in place.

So go into our optimizer, and this time we're just gonna give it uh something relatively straightforward. We're gonna say uh do some sentiment analysis on this block of text. How would you score it? Minus five to positive five, and let it think its things through. While it's doing that, I'm now going to use Gemini prompts.

This is B3, right? And then the range for this is A2 F2. I believe that's how you do range. And closing parentheses. No, it's not how you do that.

Oh dear. Is it colon instead of come? Is it colon? It will oh that it's it's it what it's running into the issue. It's it's self-referencing itself, it's creating a circular reference.

All right, let's ask Gemini because one of the nice things is Gemini is built into itself uh here. How do I fix this formula so that it references cell A2 and F2, and then I'll just give it this. Oops, stop, and we'll see what it does. This, by the way, again, this is available in Copilot as well in Excel. Uh oh look, that was easy.

It says uh you just need to put it in commas. Shows you how good I am at this warning that it says, okay. Well, that's great. Great, but this but this is yelling at me. Okay, so I did that.

Well, do it again. Clearly. AI function cannot be used with other functions. Oh, okay. So what does that mean?

Yeah, basically, it it it can't have uh a subfunction in there. So here's what you do, but that's secretly. We're just secretly gonna put together a column that says concatenate. This is a like spreadsheet judo today uh interesting things. But you know what's interesting about it, John, is that this is the low-tech version.

Because we're not writing code, we're literally just trying to summarize and categorize the qualitative data, which is where a lot of people get stumped of like there's so much information. How do I know? And this is the low-tech version of that. We could bring it into Colab or Claude Code or something and have you know Python code written. But what I wanted to make sure we did was show if you're savvy with Excel, great, you're already ahead of the game.

Now you can still do that, and but get the benefits of generative AI. So it's that nice blend of you know, you're high tech in Excel, but you want the low-tech AI solution, but you still want to benefit from it. So let's see if our little magical hack there. So by just doing a concatenate and a column along the side, we now have the subcategories for each of these things. Nice.

Now let's check in on our sentiment prompt here. Uh C. Let's do the it says do the record the version A is robust. We'll use the version A. Paste that in.

And while I'm gonna have it do the urgency one next, start a new chat. And for the purposes of the live stream, obviously, we're on sort of a bit of a limited time. So it's that like, you know, and magically it's already cooked in the oven kind of thing. We still highly recommend, you know, human intervention, making sure that the data is coming out correct, making sure that the prompts that are being written make sense to you, um, making sure that sentiment analysis is reading as the way you want it to, the positive, negative, that it all makes sense. So, you know, for the purposes of the live stream, Chris is going through this a little bit more quickly, but in real life, outside of you know, a time constraint, you would want to make sure that you're really double and triple checking everything that AI is giving you back.

Exactly. So here we have our prompt for urgency, and we're going to do Gemini prompts B2 Urgency level low urgency. All right. No clue whatsoever. But I don't have to.

I can literally say to Gemini, I need help tabulating. I need help tabulating the answers in column A of sheet one by count. Topic by count. Can you do that for me in sheet two? And let's see what it comes up with.

This to me is the most useful thing about having you know your AI here in the system is to say, like tell me help me how to do this. This sheet appears to be anti. Of course it's empty. Oh, in column A of uh sheet one let's call this answers counts. Uh and I'm gonna say in column A of the answers sheet.

Can you do that for me in the counts sheet? Let's see if it recognizes the names that we've we've given it now. No. Let's I'm going to cheat here and just go over to the uh the formulas we used in the previous version of the sheet, uh, which is we're gonna do a unique, so this is topic, and this is count. We're gonna do a unique sheet one.

Oh no, it's answers. And this is A2 through A999. I think I'm missing a closing quote. There you go. What parser error answers?

Why does this not work? Does not use backslash explanation point. Backslash is called the formula fail corrector. You just do this. There we go.

And now in the next column, we have to do the count if. And the formula for this. Answers. A2 A999 A3. A two.

All right. So now we have our basic chart. We have our basic results. We have eight uncategorized. We have marketing strategy analytics.

Let's do call it. Let's put this into a nice bar chart. I said bar chart. And there you go, Katie. There's the answers from our survey.

I like it. Um to satisfy my uh OCD. Can you go to the and I need you to do a couple things for me? Spreadsheets are very can you please go to the answer sheet? And can you please freeze the top row?

People who don't do that are maniacs, and I can't believe people like can work that way. Okay. I think that did. No, okay. Cancel that.

Okay. Go to view. Go to freeze. Freeze row one. One row.

Yep. There you go. Now, can you do me a solid and sort the topic? Sort the topic. So that we have sort of the uh yeah.

Because everything was sort of like mixed together. Mm-hmm. So that's more helpful to me. It's like, oh, here's everything under AI agents and automation, and then the subtopics next to it. Got it.

Okay. So basically I'm trying to look at the clusters together. So then if you go to the counts. Or you can uh sort and filter the count column, which is column B, sort of biggest to smallest. Right.

And we could actually do this with a subtopic too to drill down into what the subtopics are. But um to your point, yes, this tells me what I'm looking for. The first thing I would go looking for as me is what's unassigned. What's in those that it can't be assigned? Right, which we go down the bottom, and then how do you pick winning lottery numbers?

Tell me everything Katie knows so I can be just like her. So it's the junk stuff. Yes, exactly. Got it. Which is also helpful to know.

Yep. Um, here's here's the next thing is what do we do with all this information? So one of the things that we want to know that you could do in here, but you can also it uh do elsewhere. And uh this is where I would personally um spend the time is take this into a system like Colab to have it start doing that statistical analysis. Like what is the relationship between sentiment and intent or sentiment and urgency and things so that we could ask those questions because maybe there's some topics that yes, it's a frequent topic, but there's no buying intent.

Nobody wants to, no one's gonna buy anything from us on it. They just want to broadly know something about it, or it's at things, and that's what you can do that inside here, uh, but it's probably faster to do it inside Google Colab. And I think that's where we're really talking about the lower tech versus higher tech solutions. This, in and of itself, without that additional analysis, can take you pretty far. This gives you a content calendar.

This gives you, you know, a direction to revise your services, or you know, it gives you a starting point for whatever your speaking topic is this year, or it tells you everything that you're doing is the exact opposite of what your customers want. Now, fortunately for us, we looked at this and none of it was really a surprise. But it for me, it's helpful to know that it's not all agentic AI. There is other things that people want. So if you go to the counts, for example, Chris, you see obviously you have a gentic AI, but the biggest topic, marketing strategy and analytics.

That is our bread and butter. That is super helpful to know. So we're not just gonna dump everything we've been doing for the past eight or nine years. We're gonna figure out where agentic AI fits into it. We're gonna figure out where business operations and scaling fits into strategy and analytics and so on and so forth, so that we're addressing all of those things, both separately and uh as combined topics.

Exactly. Now, when you dig into the prompt for that, one of the things that makes that category so prominent is AI ROI is in the analytics category. And then so that's again where those subtopics really uh start to help out is for us to be able to look at and go, oh, well, people want to know how do I measure the ROI of my AI efforts. Which completely makes sense in that category because AI agents and automation, that's just how to build the thing theoretically. That's what it's in the category.

That's not saying, and I want to measure my progress. Right. So ROI, regardless of the of the technology, really does belong under strategy and analytics. Exactly. So I'm gonna fill in this last column here because I do want to get the the Gemini prompt in place for intent.

Go away. Um, and as I do that, Gemini prompts B six, F2. This will give me my intent scores. Then we can take this over to Colab and have it do the data analysis on those columns because we'll have the the quantitative variable. Although it looks like it's spinning out a bit more stuff than I wanted there.

That's right. We can have it just detect the integers. Interestingly, just eyeballing this, um, there's there's a fairly diverse range here. There's there's nothing here that's like a zero. So it's it's you know, one, two, and three is sort of the the background of of where this intent is.

And that's actually pretty good to know for John's purpose. Like, yeah, there's there is some buying intent in here, it's not just looky loose. You can tell by John's face, he's excited. Let's make that money. Yep.

And there's obviously and as expected. Look, if you look in the zero, like this is irrelevant chatter, no buying intent. Like that, okay. That's that too is useful to know as as a diagnostic if you put in some intentional junk questions uh or junk answers. Uh, it's a way of fact-checking these things.

All right. So our next step then would be to say we want to do a more thorough statistical analysis of this of this data file. Uh, we we have survey data. So let's drop that in and let's give it some instructions. We'll say this is survey data that also contains sentiment, urgency, buying intent, topic, and subtopic.

We want to analyze this for a few different dimensions. First, we obviously want to know the counts by topic and by subtopic. We also want to know if there's any correlation between sentiment topic and urgency and intent. So creating urgency and intent, they should the urgency and intent should be about similar uh in terms of their uh their values. But we want to know how do things like topic and subtopic and sentiment influence urgency intent, if at all.

And they may not. So that's the prompt I've given it. Uh, and we're gonna have it go ahead and and for those who've may not may not have watched previous sessions where we've done work in Google Colab, Colab is going to use the Gemini model to write Python code to do the data analysis and ultimately come up with uh first it's gonna come up with its plan. Uh it's like here's my plan. I'm gonna read it, I'm gonna analyze and explore relationships, and then I'm gonna produce a final output.

Are you sure you want to run this? Um by the way, be careful, read it before you hit yes auto run. Sometimes things just go bad. All right, so we can see it's already starting to build our Jupyter Notebook, which is a Python notebook, and is looking at the condition of the data. Uh, it shows formula like strings, shows some weird stuff in there, so it's gonna attempt to do its own cleaning on this to try and even understand what it's looking at.

Um, and it's already flagged by the way that there's some might be some data quality issues in the what Google Sheets spit out for itself, which I think is is interesting and helpful. What I like about Google Colab, and to your point, Chris, we've talked about it on previous episodes, is that it is still technically a low tech solution because you, the user, you're not writing code, you're just saying, hey, I need you to help me do this thing. Can you write the code for me? And that is specifically what this tool is meant to do. You can write code in Gemini in Chat GPT, but that is not speci like that is not the single function of those pieces of software.

This is the single function of this piece of software. Therefore, from my from where I sit, that makes it the best at it in terms of you know what people have access to and what it is you need to do. This is the one that you want to use. If you're like, okay, I need to do something more advanced, but I can't write code. So let me find the tool that's going to do it for me.

This is a great low barrier to entry way to do those things. Yeah. Now, this based on the work plan and the fact that it's running into as many data quality issues. My guess is it's going to take probably 15 to 20 minutes to finish go through going through the baking. So the process is think through carefully what you want that prompt to be, and maybe even use your prompt optimizer to come up with it and then let it go and churn through.

So like it's starting to come up with just topic distribution. Um it's looking at the subtopics and things. Interesting in the in the subtopics, practical AI implementation coding app development uh showed up there specifically in subtopics, uh, which is is a nice little interesting thing. Again, that doesn't surprise me. We know just from our because like you did, Katie, I've read every single response.

And I want to make AI agents has was definitely a very, very uh hot topic. Right. All right. Distribution of sentiment. There's a bit of a mess in there.

That's gonna need some cleaning up. And let's see. Uh yeah, now it's going to go back and do that cleaning. So when this is done, ideally we'll have an answer for John. Um, and also for ourselves to say when people are talking about this particular topic or subtopic where they have this sentiment or about it, they are going to buy something from Trust Insights when they have this or this or this, they're not really buying.

And it's good to create that content maybe for awareness, but we probably don't want to build a service around it because there's not enough buying intent about it. Well, and I think that it's, you know, again, sort of if you just stopped at the counts in the spreadsheet, it does give you a content calendar. But to your point, it doesn't give you that additional focus because to be fair, Chris, we're only two people. We can only create so much content. We can create a lot of content if we want it to be really kind of crappy, but if we want to create a lot of high quality content, I'm being very specific, uh, that's actually helpful to people.

My first question is where do I start? How do I prioritize what people want the most? Just having the number of mentions isn't enough for me to say, okay, this is what's going to then translate into sales. Because if we're being completely honest, at the end of the day, we are looking to sell stuff. We're not looking to pitch these people back and say, okay, you said you have this problem, but we're looking to create things that then help them trust us to solve that problem for them or to serve or to serve themselves.

Either way, we're building that relationship through this one question survey. But then we can pass along uh those contacts to John and be like, this is what we did. Do you want to be the one to follow up? Like, hey, we answered your question. Here's the thing.

Let me know if you want to start that dialogue. Exactly. And what's so cool about using language models like this is we didn't have to ask, you know, what is your level of buying intent, right? Uh, because clearly, uh, anyone who's ever filled up on is like, none, I don't ever want you calling me. So you know not to fill that in.

But if we rely on language models to parse out and think that through, we don't have to ask that question. And so part of the assignment for everybody who's watching this is to say, how much how much mileage can you get out of a free form response? How much can you infer from it? And we didn't even do feature engineering. Um, one of the there's a couple things you could do with feature engineering, one of which would be how long was the response?

The longer the response is, the more care someone put into writing it. Um, and that itself can be an intent signal, right? If you got a one-quest survey, like agents said that you went out to they say, okay, clearly you didn't really have a lot of intent. You gave us a page worth of this is the eight things I care about right now. I'm I'm so involved in this and this.

You have a either you have a lot of free time, you have a lot of time. Or you have a lot on your mind, and someone took the time to ask you the question, and you're like, Finally, I have the space to say everything that I've been thinking about. I mean, Chris, I did that to you earlier, and you were like, Oh, you have a lot on your mind, you know. So I very much I see it more as that versus the free time. It's like, no, I haven't nobody's asked me what I'm thinking, and now they are.

So let me go ahead and tell you. But so what I see with this, I feel like we have a really clear next step. Is once we get the analysis sorted out through Colab, we can then go back and say, what do we currently have that answers these questions already? Now, John, you can go through, it's not a pitch, but it's a touch where you can go through and say, Hey, thank you for responding to our survey. Did you know we have this content already that answers your question?

If it's if you still have questions, feel free to reach out to me. You know, I'm happy to help. So it gives you that opportunity to build the relationship in a meaningful way, not just, hey, you filled out our survey. Here's the five things you can buy today, which everyone's gonna just be like unsub, block, get out, no more John Wall. We don't want that.

We want to do it in a more thoughtful way. Katie, do you want to talk about what else we used the survey for this week? So the other thing we use the survey for, which I know we're starting to run up on time, was to help us figure out internally where the heck we're going with the company. Now, the when I say that, it's not that we don't know. Like I have a very clear vision, very clear plans on what we're doing next.

But what this survey did was two things. One, it validated the plans that are in place for services and growth, but two, it gave us a more concrete plan of specifics, which we were kind of missing before. And one of the things that Chris, you've been working on that I thought was actually really interesting, that I'm still wrapping my brain around, is you brought it into Claude Code and deployed agents against it. Now, these agents are basically stand-in team members in this case. So you have a co-CEO, which is me.

So I feel very confident because I helped build it that it's going to represent me. And results we got back, it very much was, which was fantastic. We have a CFO, we have a chief revenue officer, and then we have the voice of the customer, which is always important because we want to make sure we're not doing these things in a vacuum, that we're actually doing it on behalf of the person who's going to buy it. So rather than Chris and John and I sit and go, what do we do with this? How do we move it forward?

Chris took everything he knew about us as trust insights and us as leaders, and then took the survey feedback and said, What do we do? How do we move forward? And it gave us the start of a strategic plan and said, these are the options of things that you could do. These are the likely probability things that are going to make money. And so again, this isn't a, okay, take those things and then you know sell them back to the people who filled out the survey.

It's for us internally to say, okay, how do we take what we have and refine it so that it becomes a service that people who are having these similar issues, we can solve them for them. So now Chris and I are working through the results of that, you know, Claude Code strategy session is that, hey, we have some really good viable options that aren't going to take us a lot of extra bandwidth to put together, and then we can get out there, and then John can sell the heck out of them. Exactly. And this is all based on things like the Trust Insights ICP service, uh, the Trust Insights Casino Prompt Framework for deep research. Uh, one of the things we did was we said we had Claude Code go through our sales playbook and say, fact check this.

What do you, you know, what things in here is there not supporting information for? And it came back and it wrote us 40 deep research prompts that I now have to go and get commissioned and set up and running. But it did a fantastic fantastic job of researching even our own assertions about who we are and what we do, combine that with the survey data that we have, all the real words of the voice of the customer to come up with, hey, here's what trusted says should do in 2026. And unlike if we had done this the old-fashioned way, you know, it ran through this entire process in about 70 minutes with everything. Um, whereas, you know, if I was doing it as a human, it'd be like a five-week project and I'd be, you know, uh crying into my coffee at the end of every day.

Now, to be fair, it's the kind of work that I thoroughly enjoy doing, but to your point, it's really about bandwidth. And so I can take all of the uh I can take all of them. I can take the survey feedback, I can take our sales playbook, I can take what I know about our audience, voice customer, and I can manually put all of that together. I enjoy doing that work, but what I like about this is it gives me something to react to, and I still have to do the critical thinking. I still have to put those plans together.

I still have to be the one to set the vision for the company. I'm not gonna let AI do that for me because it doesn't know what I know thoroughly. So it just it it brought me closer to getting to the point of an execution plan versus to your point, like a couple of weeks of getting through everything to get to the same result. Right. It's like a basically a team of analysts that brought you the analysis, so now you can make decisions on it much, much faster.

Right. All right, so that was survey analysis, the low tech, medium, low tech. The high tech way, obviously, is to spend some time with a tool like Claude Code. Say, here's my survey data. Give it even this example of what we just did.

Say, write me some Python code that will autonomously execute this against the LLM of your choice. That's what you would want to do if you have survey results. Say, you know, maybe your McDonald's and you have hundreds of thousands of survey results every day. You're not going to do that in a spreadsheet. Yeah.

Once you get beyond, you know, 10,000 results, you're really going to want to automate it fully. Um, but that's for another show. Uh any final words, uh, John and Katie. Just thrilled to dig into all the numbers. You know, like I said, this is the kind of stuff that people would be churning for eight months.

Like by the time you get the results, the people are dead. So it's great to have a nice uh, you know, have it actionable. Uh well, and I will say, quick plug. I mean, this is the kind of work that we do for ourselves, but also for our clients. So if you are buried, you know, we call it dark data.

That's always sort of been our, you know, mission since day one is to help shine a light on the dark data. This is still that. Give us a shout. Go to trustinsights.ai slash contact. Uh, you get to talk with John.

John will figure out how we can help. Yeah. Light it up. All right. Thanks for tuning in, everyone, and we'll talk to you all on the next one.

Thanks for watching today. Be sure to subscribe to our show wherever you're watching it. For more resources and to learn more, check out the Trust Insights Podcast at TrustInsights.ai slash TI podcast at our weekly email newsletter at TrustInsights.ai slash newsletter. Got questions about what you saw in today's episode? Join our free Analytics for Marketers Slack group at TrustInsights.ai slash analytics for marketers.

See you next time.