Latest / Content, Briefly / How Content Marketers are Actually Using AI

Transcript

How Content Marketers are Actually Using AI · Content, Briefly

• 0:00
  [ on-hold music] This episode is brought to you by our friends at Brevo. Looking to level up your marketing?
• 0:12
  Meet Brevo, the all-in-one platform built to help you connect with your customers, drive engagement, and grow your business.
• 0:18
  With Brevo, you can keep all your customer data in one place, track prospects seamlessly through the CRM, reach them via email and SMS, and boost retention with powerful automations.
• 0:28
  Easily create personalized multi-channel campaigns to ensure every customer gets the right message at the right time.
• 0:34
  Plus, Brevo offers seamless integrations, a messaging API, advanced analytics, customizable forms, and more, all in one intuitive platform that gives you a full three hundred and sixty degree view of your customers.
• 0:46
  You can get started for free, or use code SUPERPATH to save fifty percent on starter and business plans for the first three months of an annual subscription. Try Brevo today and take your marketing to the next level.
• 0:57
  Hey, everybody. Jimmy from Superpath here today with Eric. Got a pretty cool topic today. Unwriting AI use cases for content marketers, and we have lots of them. So talk about that. But first, Eric, how's it going?
• 1:08
  You're in your new house, new background for you. Uh, hopefully you're getting settled over there. Yeah. It's probably echoey because I still haven't moved everything in.
• 1:15
  Got a nice echoey office right now, which I'm excited about, but it's been great. Yeah. We've like painted the whole exterior. We've replaced the fixtures. It's been a week of not...
• 1:24
  I think I've put in twenty thousand steps a day in my own house according to my phone. Yeah. [chuckles] Which is insane. That is insane. That is insane. All right. Well, you can stay seated for this one. Okay. Thank you.
• 1:34
  I appreciate that. So we have... Yeah. Well, I guess there's a little bit of backstory here.
• 1:38
  A couple weeks ago, I posted on LinkedIn about this idea of, like, the new T-shaped marketer, and I'll admit that it was kind of a half-baked idea.
• 1:45
  But the point I was getting at was that the, the horizontal axis of the T is getting wider. You know, we have new skills available to us thanks to AI.
• 1:53
  So a content marketer could, in addition to, you know, having strong editorial skills and understanding how to distribute content and things like that, they could also, uh, make a simple web app.
• 2:04
  They could also write some code. They could do image generation. There's a lot of other skills that you can do that you may have used to rely on other people for. Okay. And there was -- Go read the comments.
• 2:16
  People were like kind of upset about this idea, and it was a lot of like, "No, it's just making everybody do mediocre marketing and stretching us all too thin."
• 2:23
  And, like, I totally get where people are coming from, and I think part of it was that my idea was too half-baked. But, hey, it's Link- Isn't that what we're supposed to do with LinkedIn posts?
• 2:31
  Anyways, in giving this idea some more thought, a few things come to mind. One is, I think writing is actually the least interesting use case for generative AI.
• 2:40
  And whenever I talk to people who are, like, really AI, you know, like, early adopter type, that's not what they're doing with it.
• 2:48
  So, like, the impression people have of like, "Oh, we'll just all create mediocre content thanks to AI, and our bosses will wanna squeeze more productivity out of us," like, that's not...
• 2:57
  At least what I see is not really that out in the real world. So we asked people on LinkedIn, "What are you doing with AI that isn't writing?" And we got tons of cool responses.
• 3:06
  And so we'll run through some of that today.
• 3:08
  But maybe first, you know, Eric, as a content lead at a SaaS company, are there some use cases you wanna share, some non-writing use cases, um, at Dock that you've employed that have made your life easier, better, more productive, uh, helped Dock get more signups, more revenue, et cetera?
• 3:24
  Yeah, for sure. And I think we'll cover a lot of the themes as we go here, but I think, like, step one for me has just been a lot of efficiency gains. Like, here's a CSV, I want you to turn it into an image.
• 3:35
  Or even recently, we've been rewriting our, our sales decks. I mean, I'm getting into writing right away, but just as an example, I'm just pasting an image of a deck and saying like, "Give me ideas here."
• 3:45
  And, like, those... There's just little efficiency gains in, like, going from format to format that I think is interesting that before would've taken a lot of manual copy-pasting or- Yeah...
• 3:55
  here's -- Or for the Superpath survey I did the other day, I put it in ChatGPT, and I said, "Just run some analysis on the answers so far."
• 4:03
  Of course, it said there were thirty-six responses when there were actually twenty, and I said, "Are you sure about that?" [chuckles] And then I asked it to redo its math, and then all the math checked out.
• 4:11
  But yeah, there's some, like, instant efficiency gains from in terms of things I was going to do before but I've been able to do quicker. And then there's been some things I've been able to build.
• 4:22
  Like, I build myself a new dashboard of where every project is at with a freelancer, what the budget, you know, what my remaining budget is, things like that, that, you know, I could have built in Airtable, but I wanted to build it my own way.
• 4:34
  And so I have more flexibility to build these, like, micro SaaS tools and apps for myself. Yeah. That's really cool.
• 4:42
  There is, you know, at least at the mo- I find that, like, folks who are tinkering a lot are probably spending more time getting the same amount of work done- Mm-hmm...
• 4:50
  because they have to explore the tools and, you know, kind of go through the process of trying to create something new, and then it may or may not work, and you may still end up doing it the old way.
• 4:59
  So I feel like we're all kind of in this,
• 5:01
  you know, kind of very, like, exploratory phase of trying to understand, like, okay, obviously it's an amazing technology, but, like, what are, like, the really, really practical ways we could use it?
• 5:09
  And so recently I've been doing a content consulting project for a startup, and I'm helping them create content. And th- this'll be like AI native company. It's small. The founders, like, totally bought in.
• 5:20
  So I think they're kind of like the leading edge of companies who are aggressively adopting AI, and the way that we've gone about incorporating into the content creation process has been really interesting.
• 5:31
  Uh, like, I hadn't really used Granola very-- I've used Granola. I have not, but I've heard of it. It's wild. It's such a cool tool.
• 5:38
  So, like, we'll have calls that are transcribed through using a tool, using Granola, and the notes that you get from it are insane, and you can interact with them, and then you can also interact with them across multiple conversations.
• 5:49
  And then we can feed that into a cloud project, you know, alongside existing brand guideline, a style guide we made for the founder, you know, blog posts that were successful and that we like.
• 5:59
  You, you feed all this in, and while we're not using it directly for content generation, it helps so much when you're like, "Can you get me an example of a company who's done this so I can cite it in the article?"
• 6:10
  Like, that type of stuff has been- Yeah... just a massiveTime saver. And then when I'm not there, we can hand it off to the next person. It will already be built.
• 6:19
  We're using a similar tool at Doc called Fathom, which is like a classic sort of call recorder. And the other-- We're working on repositioning Doc.
• 6:27
  I won't get into the details, but I needed to do some research on a competitor, and I was able to search the competitor name and all the call transcripts, and just it was able to pull out the snippets of all those calls, and I could just watch people say a bunch of negative things about our competitors.
• 6:41
  And it immediately, within fifteen minutes, I would have had to have, like, watched... You know, they always say, "Watch sales calls," and that's very good advice.
• 6:49
  But here I was able to watch a hundred sales calls for a specific moment- Yeah. Amazing... and learn a lot more than I was able to before.
• 6:55
  And I had, in ten minutes, I had the exact ten-point document I needed to do my positioning work. So th-that was incredible efficiency gain. And again, something that would've just taken me twenty hours to sit through.
• 7:08
  Yeah. That's so cool. It's interesting because I think there's sort of like two large buckets of work that content marketers are doing with AI. One is efficiency.
• 7:14
  It's like a thing you used to do that took a long time, or maybe you didn't even do it because it would take so much time.
• 7:20
  And then there's the other category of stuff which is actually wasn't possible before, and now it is. And then within that, there's other buckets.
• 7:27
  So like, you know, we've made this document real quick before this podcast to, you know, kind of pull out some of our favorite use cases that people shared with us.
• 7:34
  And there's, you know, there's kind of a few buckets, like, within that. And we'll give some examples of each of these. One is, you know, research and analysis.
• 7:41
  There's just, like, all kinds of cool ways to do really deep research and to analyze data with AI. One is building stuff, so that could be prototyping, that could be creating internal tools.
• 7:51
  Uh, it could be creating external stuff like quizzes and graders and things like that. And then there's this other one, which I guess, I don't know, what do we call it?
• 7:59
  Memory or knowledge, where you're sort of, like, accumulating a bunch of stuff in one place so that it can be leveraged over and over again.
• 8:07
  And then, like, the efficiency gains happen over, like, a longer time horizon than some of the quicker productivity gains.
• 8:14
  And then I guess personal stuff too, like we'll probably share a few personal ones just because, like, this is a group of people who are definitely early on the adoption curve and are using it for all kinds of cool things for other parts of their life, not just the professional part.
• 8:26
  Yeah. I'll also add on top of that, there's just sort of workflow productivity things.
• 8:31
  So like, uh, SEO tasks that were monotonous, but not content creation ones, but doing, like, gap analysis or things like that or- Yeah. Yeah. So there's this sort of miscellaneous bucket of efficiency workflow thing.
• 8:45
  Yeah. That's so cool. I wish I had this. I wish I had all of this stuff like ten years ago. There's just like the amount of, like, things that I haven't done in my career because it was, like, not possible.
• 8:54
  Like, at one point in my career, I had this idea of like, "Wait, what if I made a WordPress plugin and it would integrate with Google Analytics, and it would help us track posts that were decaying, like getting less organic traffic, and then automatically add them back to our editorial calendar?"
• 9:08
  And I was like, "That's a cool idea." [chuckles] The end. The end. Yeah. Nothing else to do with that because I'm not gonna pay a developer to do it. Like, this is just like a little thing that...
• 9:17
  You know, anyway, but now, like, you could do that. You could probably do that in an hour if you really wanted to. I will add one more bucket before we get into examples. Yeah.
• 9:24
  A, a bunch of people commented on your post with what I'll call, like, personal growth or career coaching- Mm-hmm... use cases, which, like, I'm skeptical on how much the feedback is valid.
• 9:34
  I was talking to someone earlier today, and they said the problem with ChatGPT is every time you ask a question, you get an answer. And so when you say, "Could this be better?" It says, "Yes, it can be better.
• 9:43
  Here's how." So if you're doing personal coaching through this and you say, you know, "Criticize me," it's gonna criticize you.
• 9:49
  I'll put an asterisk next to this one on if I think this is the right way, but I have seen that's the number one use case for the average person in ChatGPT in the US, I saw some research that I cannot cite now, was personal feedback and personal growth, and, like, coaching.
• 10:04
  People are asking advice of ChatGPT for-- about themselves. That's so interesting. Uh, and so... Yeah. So I mean, I'll give some examples here, but why don't we just jump in, is- Yeah...
• 10:14
  uh, Pallavi, Pallavi Dhody, almost same last name as me, s-said they've been using ChatGPT as a writing improvement tracker, so th-they sort of continuously feed in their work, and they continuously ask for feedback and ask their GPT to, like, give progressive feedback on, like, "Have I achieved the advice you gave me?"
• 10:32
  So that's a writing use case kind of, but it's that sort of personal feedback, and there's a few folks who said things like that for personal growth and coaching. Yeah. That's very interesting.
• 10:43
  I very much appreciate that AI could help uncover blind spots, you know, and recognize patterns if you're, like, trying to avoid making a mistake or change something in the way that you write, if that's like a habit you fall back on and it calls it out every single time.
• 10:58
  Like, b- definitely see how that could be very useful. Totally. Do you wanna jump into some of the other categories? Yeah, let's do it. Can we do-- I think my favorite one is the build category.
• 11:08
  Can we start with the build one? One of my favorite that we saw was from Anki Parekh. I hope I'm pronouncing your name right. It says, "Vibe coded a makeshift employee advocacy social platform for internal use.
• 11:20
  It does have some kinks, but the proof of concept works." I think that's so interesting.
• 11:24
  Like, this is a problem that I've seen at every content organization I've ever come across where they're like, "How could our internal team help us get this stuff out the door, help us share it?"
• 11:34
  And, like, nobody ever does it. You know, like, there are apps, I'm sure. I don't... I guess I'm not sure. I would imagine there are tools built specifically for this.
• 11:42
  We have- we haven't at Doc, but yeah, it's nothing better than I could have vibe coded in ten minutes, and I think it was free up to a certain amount of users, and we just passed that number of users, and it turned off, and I'm not gonna pay for it.
• 11:53
  Yeah. So it's a perfect example of, like, a SaaS thing that, like, has some value, but not enough that I was gonna pay for it anyway. I could probably build it in ten minutes.
• 12:00
  All it needs to do is ping Slack and say, "Please share this," right? Like that kind of thing. Yeah.
• 12:06
  D- absolutely, and also, I do feel that part of the, the build, or maybe prototyping is the better word in this instance, is that
• 12:13
  wh- when you do it yourself, you have a much better understanding of what the problem is that you're trying to solve. Like, so rather than buying a solution that solvesYour problem and probably some adjacent problems.
• 12:25
  Y-you can build it specifically for your team, your use case, your tech stack, y-you know, the personality. You can build it in a way that is way, way, way more customized.
• 12:34
  And I do feel that by doing it yourself, you are able to, one, understand why your previous efforts haven't worked like you hoped they would, and then hopefully build some solution into your app.
• 12:46
  And also, I mean, to the degree that this matters in a professional environment, it's really fun. Like, it's a really fun type of work to do. Absolutely.
• 12:54
  And I think w-we could probably break the building and prototyping category into two subcategories.
• 13:00
  There's the prototyping tools for internal efficiency or, like, workflows building content, and then there's the building content side of things. Yeah.
• 13:11
  So, like Er-Erin Balsa said, for example, she built a-an assessment tool for customers where they can plug in some data, and it can spit out an answer, right?
• 13:22
  Or, and Jolly Joshi said that she built quizzes and landing pages via vibe coding.
• 13:28
  So that-that's like building content, like customer-facing content that they might not have otherwise created, versus like those previous cases we talked about were sort of internal tools. Yeah. Yeah. I like that a lot.
• 13:43
  Uh, kind of on that same note, Anjali Joshi left a comment about building quizzes in landing page with vibe coding. I think that's super interesting.
• 13:51
  I'll restrain myself from talking about the podcast bot, but there are other- You made a podcast bot? Did you? No. Like, I was like...
• 14:00
  [laughs] I think that there's a very interesting opportunity for software to be used in places where content was previously the default because the content was the tool you had. A-and now you have other ways.
• 14:17
  So like, if, say, you're trying to rank for a keyword, for example, but content is your... You know, what is it? Like, if all you have, every hammer, every nail... What's the thing? You know what I mean.
• 14:27
  If you have a hammer, everything's a nail. There we go. So- If you have a podcast bot, everything's a- Everything's [laughs]... answer question for a podcast bot.
• 14:35
  Just wait, just wait till people are using the podcast bot to get a summary of this conversation. Might break the dozens, I think less than that.
• 14:42
  [laughs] Anyways, yeah, I think there's this opportunity to like, you know, what-- Like, quizzes is just, like, such an easy example to throw out, like, 'cause there's cool ways to do it.
• 14:51
  Like, I saw one the other day that's actually really interesting. It was, like, a risk assessment calculator for product managers who are in the process of building AI into their own tools.
• 15:02
  And so it was like trying to understand, like, how, you know, of your user, where do they sit on the adoption curve? Like, what is the frequency of use of this product? Like, what's the price point?
• 15:12
  You know, whenever I ask you all these questions, it kind of gives you a score of, like, your risk in this category is whatever based on these. Th-that's really cool. It's actually quite useful too.
• 15:22
  And I'm sure we need to find some other examples of things people are building beyond quizzes and podcast chatbots, 'cause I'm sure there's many. But I feel like that's a world to me that's, like, almost limitless.
• 15:32
  Like, is whatever you can dream up, you could probably build. I absolutely agree. Let's see. What else do we have in here? Oh yeah. Uh, Diamaka, "ChatGPT helped me build a landing page in PHP.
• 15:43
  For context, I already know the fundamentals, but it helped me get eighty percent of the way there." That's very interesting.
• 15:49
  So that's-- I would imagine that's a circumstance where you're running a website probably on WordPress because that's built on PHP. You need a new landing page.
• 15:57
  WordPress is terrible for that, so you probably need a plugin.
• 16:02
  If you're in an environment where it's hard to get plugins for security reasons or whatever else, the ability to just write the code or basically vibe code your way there.
• 16:10
  So you make it look like you want, then generate the code, and then plug it in.
• 16:14
  That to me is really cool because it's an example of a scenario where you used to rely on other people to help you, and now you don't have to anymore. Yeah, exactly.
• 16:22
  I also have from Adam Fackler, he had said he wrote some custom code to get to build his own Calendly style pop-up. I can imagine what that looks like. From his own button instead of the one they provide.
• 16:35
  So yeah, cases where a piece of software says, "Here's the UI," and you can say, "No, I want it to look this way instead." Yeah. Yeah.
• 16:42
  I actually noticed I was on the V0 website, V0 is like a Bolt and Lovable competitor, earlier today, and actually one of the default prompts they have in the chatbot you can just test on their website is, "Give us a screenshot to replicate."
• 16:56
  So you can just give them a screenshot of the landing page or app or but-- whatever you want, and then it, it will just build, it will build that, which is wild to me that's possible. That's amazing.
• 17:06
  I think we've checked the building and prototyping category as far as I can tell from the ideas that people threw in. I'm sure people will have more ideas after this episode. Definitely.
• 17:17
  Maybe one of the things I'll just quickly throw out is there's a kind of a professional grade of this as well. I was actually talking to an agency owner just yesterday who said that they've always run...
• 17:27
  Uh, they, they built a dashboard for their clients using Bubble, which is like no code app builder, allows you to, like, deeply customize the experience.
• 17:35
  Uh, but Bubble has now shifted from no code to chatbot, essentially. So like for companies who have already been building internal tools, there's now a way to, I guess, kind of supercharge that.
• 17:45
  Like a tool like Bubble has like the complete infrastructure for an app, where like some of these vibe coding tools, they're not there yet.
• 17:52
  You know, like you can't build a Lovable app, or maybe you could, I guess, depending on how you integrate it with other products.
• 17:56
  It has like logins for your users, you know, whereas like a tool like Bubble has that all built in. So there is like a professional tier of this also. Okay, sorry. Yeah, let's go off building and into...
• 18:05
  What's the next one? You wanna do product knowledge? Yeah, sure. Let's do it. So Bridget from Creative Little Planet, they're an agency.
• 18:14
  They've started to create product expert buddies for their clients, so they can ask the GPT questions when they're coming up with-Content ideas. I think that's a great use case.
• 18:24
  I think at Doc I've done a similar thing where I built like a Doc GPT and then I've given it to the writers who can...
• 18:32
  So a-again, we're getting into content creation territory, but even they can just ask questions of it about like does Doc do this, does Doc do that? And it will get you eighty percent of the way there.
• 18:41
  There are some hallucinations, but it's better than nothing, which is what most, [chuckles] you know, most freelance writers, for example, just get handed the company website and a blog brief, and that's about it, right?
• 18:54
  So there's, this is much better than that world. Do you train writers on how to use that effectively? 'Cause I would imagine for most of them, they've never been handed something like that. I do not.
• 19:03
  I feel like I'm really only working with four or five writers right now, so it's not that bad. It's not like that huge of a lift. I-- and I think they get it.
• 19:12
  I think they're, most of them are in Superpath and are the type of- Nice... person who's already thinking about these things. [chuckles] Yeah. Yeah. I like that. I like that.
• 19:19
  Yeah, Ryan McCready dropped a s-somewhat similar suggestion in a comment, said he built custom GP to understand a company's brand, voice, values, messaging, et cetera.
• 19:29
  And I feel like it's like not a sexy use case, but it's such a useful one, you know? And I, I was doing something similar to just this morning for this consulting client I'm working with, but doing it in Lex.
• 19:41
  Lex is such a cool... Like, if you're gonna write and you want AI to make your life easier, use Lex. I don't think there's anything else really like it out there.
• 19:50
  Like, I don't know if you've messed around with like Notions AI. I think it like totally stinks compared to like a really purpose-built tool like Lex. And that's like a built-in feature where you upload style...
• 20:01
  It can even help you make style guides by just feeding it content. It will help you like parse out like what is it that's unique about this that you can then replicate in other stuff you create. That's awesome.
• 20:09
  I'll put one like sort of in this category. Tanya Svoboda said, "Running copy past a virtual customer persona and asking for critiques." I feel like that's similar to what we were talking about- I like it...
• 20:19
  before in the world of coaching and feedback. So I think it's a slant on that. But yeah, saying, "Okay, here's what my customer cares about. Do you think that checks the box?"
• 20:29
  And if nothing, it's just like a nice check of, you know, did I achieve the goal of... Yeah, I've done that a few times at Doc.
• 20:36
  Like, I've just run a draft through the thing and said, "Hey, what would a sale-- did this resonate? Will this resonate with sales enablement leaders?" Like, what have I missed from a seniority perspective?
• 20:46
  Like, are we hitting that? So doing that kind of feedback could be nice. Yeah. I like that a lot.
• 20:52
  You know, back in, uh, the agency days, like one thing we learned very quickly that you should never do is email a draft of an article to a client and say, "Hey, what do you think?"
• 21:01
  Because similar to asking ChatGPT, "Could this be better?" It can always be better.
• 21:07
  And so I really like the idea of like gut checking that against a trained AI that, you know, to the best of your ability, try to help the AI understand like the customer's perspective on this so that you can deliver something that's a few steps further along, so you don't, you know, kind of expose yourself to, you know, basically having to rewrite the entire article, which is what you usually have when you do that.
• 21:29
  Totally. Okay, and I think we more or less have the last category, which is research and analysis, which I'd say of the responses you got, I think there were maybe 70 comments on your post.
• 21:40
  I'd say at least half were some version of, "It helps me do research" or, "It helps me analyze a whole bunch of data." Have you messed around with Perplexity? I haven't yet. A few people did mention Perplexity.
• 21:51
  I'd be curious. It is fantastic.
• 21:53
  I feel like, I mean, I've been doing some of this kinda like quote unquote "deep research" where [chuckles] actually deep research, it's kinda funny because regular research is you ask a chatbot a question, and it gives you an answer in two seconds.
• 22:04
  And deep research is you have to wait ten minutes [chuckles] or 20 minutes or something. Awful. Like Claude calls it extended thinking. I think Perplexity calls it deep research.
• 22:13
  The same idea, like you just give it more time, and it will give you more information, but it's like crazy good. So I was using Perplexity for this client and just trying to like help...
• 22:23
  They were writing about sort of an analysis of this industry, so I was like asking, "Can you just like give me a rundown?"
• 22:28
  Like, here's the players I'm aware of, here's the problems I'm aware of, but like help me just understand the full picture of what's going on in the space. Hit deep research, wait a few minutes, and it comes back with...
• 22:38
  I mean, I don't know the space well enough to like go check it against, you know, someone who has like really, truly deep expertise in that world, but after I read it, I felt like, wow, okay, I have like a pretty good sense here.
• 22:50
  This feels like very comprehensive to me, and it was hugely useful. I mean, in this case, more like me getting myself up to speed, but then like that translates out into the writing and the distribution, et cetera. Yeah.
• 23:00
  I think it's a better version of look at the top 10 Google results and read all those blog posts, right? Yeah. Like, that's the old version that we would have done. We probably, as content marketers,
• 23:11
  jumping into a space for a few weeks, for example, you're not gonna read 10 books on it. It's just not gonna happen.
• 23:17
  You have to do something quick, and we used to read blog posts, and I think this is a better alternative.
• 23:22
  I had a similar use case where I used deep research in ChatGPT, where I was doing sort of personalization research for some outbound campaign I had to do, where I was like, okay, I wanna adapt Doc's messaging to people who work in finance.
• 23:36
  Like, if someone works in sales enablement at finance, what is their challenge compared to someone who works in sales enablement at healthcare? Oh, dang. So different customer segments.
• 23:45
  And so I said, you know, "Do some deep research on what a sales enablement person at these..."
• 23:50
  And so it did some meshing together of ideas for me in terms of, okay, they probably care about regulation, they probably care about whatever. Procurement is more difficult, blah, blah, blah.
• 24:01
  And so that was really useful. So I did exactly what you described. I asked it to give me like an analysis of the industry and the persona, and it gave me something pretty good.
• 24:08
  And it, that was enough for me to do my job at, which was, you know, write a four-sentence cold email that [chuckles] hopefully hit on one pain point they had, right? Yeah. Yeah. Well, that's really interesting.
• 24:18
  And one thing I find is cool about doing research with these tools is that the citations are really interesting. You know, like this is obviously like a big problem for content marketers.
• 24:27
  Like, it means people aren't really clicking through, but at the same time-Like rather than it finding blog posts that are like the top 10 problems that, you know, finance salespeople have in the finance world, it's not finding that.
• 24:38
  Like in my case, it was finding like studies that were in PDFs on university websites. Like that-- So like now I'm like instantly like that is niche.
• 24:48
  That is like really nuanced and deep, and it, I got there so quickly because I didn't have to wade through all of the stuff that was optimized to get to me first. Totally.
• 24:58
  And I'll bring up the counterargument to all of this in that if it's something that you need to internalize and you need to learn and that needs to be like a core part of
• 25:07
  like your knowledge as a content marketer in your job, there's so much value in doing the research yourself, right?
• 25:13
  I mean, there's those studies coming out that are slightly-- they're small sample sizes and et cetera, et cetera, but I think the general pattern is that the more you engage the material, the more you're gonna remember, the more you're gonna internalize.
• 25:25
  So there's risk of like overindexing on- Mm. Yeah...
• 25:28
  outsourcing all the research, because then you don't, like, it's not internalized when you actually ne- then every time you're going to ChatGPT for the knowledge as input into content.
• 25:36
  So I think like I'll put a giant asterisk there and say, "Don't overindex on this stuff."
• 25:40
  But it-- for the use case where I was like, I need to write one outbound email to someone, boom, boom, boom, did my research, done. I'll probably never have to think about it again. Great use case. Yeah. Totally.
• 25:51
  Totally. I feel like the same is probably true for data analysis as well. Like- Mm-hmm... you know, if for example you spend a couple hours in a spreadsheet, you will like truly understand that data set.
• 26:02
  I got an email the other day from a potential sponsor who asked, you know, "What are the job titles of people, uh, in the community? How much seniority do they have?
• 26:12
  And, you know, wh-where's like the geographic density?" I was like, "I have that data. It's gonna take me a whi-" And I was like, "Oh my God, what am I doing? I'm just gonna upload it to Claude."
• 26:22
  [laughs] You know, and like two seconds later, not only do I have the email written, but I have the charts and graphs to go with it. You know?
• 26:30
  It's like then I quickly package it up into a media kit, and it's, the email is sent off, you know. Yeah. I love doing stuff like that. Yeah. Exactly.
• 26:37
  And we did have a couple other people talk about data analysis kind of on a grander scale, you know, in the context of reporting, for example, because, you know, one, you can...
• 26:49
  It, it's easier to aggregate data and then analyze it. It's also easier to take the data and find ways to act on it because, because of the way that you can query it.
• 26:58
  And then one other thing is depending on what tools you're using, like let's say your data is in a place where you have to write SQL queries, but you don't know how to do that.
• 27:05
  Well, now you know how to do that because you just ask ChatGPT or Claude or whatever to do that, and now you can, like, you're instantly up to speed on like how to get what you want from the data. Yeah.
• 27:15
  It doesn't always have to be asking the tool to do the analysis. It could be, "Give me the code I need to write to do the analysis." Yeah. That's how all of my Airtable scripts work now.
• 27:27
  You know, I just say like, "Write a script to do this or that," and then that, rather than building like step-by-step automations, I just say, "There's a trigger and then a script, and then the script does everything I want," because I can't write that code, but Claude can.
• 27:40
  Yeah. So why don't I read off a couple examples- Yeah... of other research. So Tesh- Tashifa said, "I use it to do sentiment analysis on reviews and better understand users,
• 27:49
  and I do market research using GPT's deep search." There you go. That's exactly what we just said. Really, from Client Boost said they use a localized GPT to analyze social media data.
• 28:01
  So they put it in CSV uploads, ask for top performing themes, improvement suggestions, optimal posting times based on their posts. So yeah- That's interesting...
• 28:11
  basically being able to do some reporting and analysis on their own con- Uh, they said not so great at building graphs, but, uh, maybe they need to improve their prompt. I've found similar by using the wrong tools.
• 28:22
  That's super. I really like the idea of being able to analyze just your own data rather than like the kind of giant aggregated trends that you'll- Right...
• 28:32
  find if you go search for this information, which would tell you, like, "Oh, send your emails at ten AM on Tuesday," or something like that. Yeah.
• 28:39
  Everyone's Googled best time to post on LinkedIn, I'm sure, if they are listening to this podcast. And I'm sure, and every report says something different than I'm- Yeah... I'm assuming that your audience.
• 28:49
  Somebody at some point said the best time to send a newsletter is on Sunday, because suddenly I get every single newsletter on Sunday.
• 28:55
  I think it's either Sunday or Monday, and I now don't read any of those weekly ones that comes on Sunday. On Monday, I come in and I put tw- in the garbage.
• 29:03
  So that would be a great example of like where the best practice is overdone and chat AI of some kind could help you figure out your best performing times. Yeah. Yeah. I like that.
• 29:14
  Also, somebody used to get a boatload of traffic for what's the best time to send an email, and they're probably not getting any now. RIP. Yeah.
• 29:21
  [laughs] Um- It's probably like a garbage delivery company or- Yeah, yeah... garbage disposal. [laughs] Cool. I feel like we've covered this pretty well. There is this whole category of personal stuff. I don't know.
• 29:32
  Some of it is cool, but it's not like quite within the scope of this conversation.
• 29:36
  But we'll obviously link to the LinkedIn post, uh, so people can read through this, 'cause there's obviously quite a bit than, than we covered here.
• 29:43
  And like different stuff is gonna resonate with different people depending on what you're working on and how familiar are with these tools and all that kind of stuff. So I would encourage people to go check out the post.
• 29:51
  Yeah. And I'll read from Adam Fackler here, marketing director at Versa.
• 29:55
  He just summed it up really nicely 'cause he listed a whole bunch of things he's doing and said, "The takeaway for me is there's a bucket of things that I'm doing today with AI that, uh, would've been impossible for me to do without more assistance yesterday, like entire workflows, and B, even if they could be done, they'd be done far slower."
• 30:11
  So I think really that's sort of like the optimistic take on all this, right? That you were trying to get at with the T-shaped marketer, pulling that back up right to the top. Look at us. Yes. Look at that narrative.
• 30:21
  We've- Love it... looped right back around. That was perfect. I should've consulted Adam before I- [laughs]... post on that one. [laughs] Great. Cool. I imagine we'll probably do more of this stuff.
• 30:30
  Like, these use cases are changing, you know, I was gonna say monthly, but maybe daily, uh, just like depending on what kind of stuff you're working on. So we'll do more of this.
• 30:38
  I actually have reached out to a few people about doing some show and tells for Superpath Pro members.
• 30:44
  So these would be like probably, you know, quick, like thirty-ish minute sessions, not webinars, or just, you know, someone's gonna share their screen and walk us through something they've built.
• 30:52
  So we're gonna have a couple of those coming up, which is a good segue. Uh, come check out Superpath Pro, thirty-day free trial, courses, content, one-on-ones, monthly group calls, uh, co-working sessions, et cetera.
• 31:02
  We'd love to have you. And you get it thirty days free, so come kick the tire and see if it's a good fit for you. Love it. I'm convinced. I'm signing up right now. [laughs] Again, for the fifth time.
• 31:12
  Do you have a free membership? 'Cause I'm gonna change that, or we have to charge you. I think, yeah, I should probably be charged double. I use it a lot. Usage-based pricing. Yeah. Yeah. Cool. Awesome.
• 31:22
  Eric, thank you as always. Glad you're all settled in the new house. Keep an eye out, everybody, for more AI talk. Obviously, there's gonna be more of that. And otherwise, we'll see you in Slack. Take care.
• 31:31
  See you, folks. [outro music]