38 min read
AI Content Best Practices for Inbound Marketers with George + Liz
Liz Moorehead
Jul 17, 2024 10:39:46 AM
This episode of HubHeroes between George and I has been more than a year in the making. After months and months and months of bashing our heads together over the marriage of AI and content — with a dash of George dragging me kicking and screaming toward a mindset that embraces AI at all — we're here talk about what we've discovered.
Our conversation comes on the heels of the unveiling of my Human-Powered AI Content Framework for Inbound Marketers guide, which is not something I ever thought I would write. It also is leading up to George's upcoming talk on creating human-powered content with the assistance of AI at #INBOUND24.
💥 Related HubHeroes Podcast Episodes:
- AI Mindsets for Inbound + How Most Organizations Fail with AI Content
- What the HubSpot Content Hub Is (+ Isn't) in This New AI World
- The Ethical Implications of AI, ChatGPT, and Inbound Marketing
In this discussion, we discuss their experiences with AI in content creation and share tips and tricks for using AI effectively. We emphasize the importance of viewing AI as an assistant rather than a replacement for human creativity.
I also pull back the curtain on my bumpy journey of initially resisting AI and then embracing it as a valuable tool. Through our collective experiences, we talk about the the need for clear inputs to get meaningful outputs from AI and the importance of understanding your audience and their needs. We also highlight the value of creating a strategic process and using AI as a creative brief to enhance content creation. Our conversation also touches on the challenges and funny moments we've encountered while using AI.
Keywords
AI, content creation, assistant, replacement, inputs, outputs, audience, creative brief, AI, content creation, human-powered, AI-assisted, research, optimization, getting started, leadership, challenges, funny moments
Key Takeaways
- View AI as an assistant rather than a replacement for human creativity.
- Clear inputs are essential for meaningful outputs from AI.
- Understand your audience and their needs to create effective content with AI.
- Use AI as a creative brief to enhance content creation. Embrace AI as a tool to enhance and optimize content creation
- Take a human-powered, AI-assisted approach
- Leaders should lead by example and bring their teams along on the AI journey
- AI can be used for research and optimization
- Start with small, incremental asks to see outsized ROI
- Consider the human element and the people who need to come on the AI journey
- Be open to learning and iterating as you go
- Don't ignore AI, but rather embrace it and put the right mindset to it
And so much more ...
Additional Resources
- Why Go HubSpot? Your Free Ultimate Business Growth Guide
- Why Go HubSpot Service Hub?
- What Is Customer Delight + Who Is Responsible?
- How Have Inbound + HubSpot Changed?
Episode Transcript
Liz Moorehead: We don't need to disclose
George B. Thomas: the, yeah, by the way, Chad, I love, uh, in the chat pane, it's funnier if you watch Linus Tech Tips, which I have watched Linus Tech Tips. So I, I will say that yes, Lack is whack.
Liz Moorehead: was here with Nick from Fargo. Yeah. Nick from Fargo. Lack is whack. I'm, I'm here for that. I love that.
George B. Thomas: Yeah, Lord, Lord Lack, you whack, homie.
Liz Moorehead: I am going to admit that while I'm excited for the conversation you and I are about to have today, George, I'm a little nervous. So typically for those long time listeners, first time callers, I typically am in the role of the question asking. I get to be a fussy little bee interrogating everybody about their thoughts and feelings.
Periodically, George will catch on to my scam and be like, no, no, Liz, you tell us your feelings too. And then I have to say something smart and I hate it. But today, I am so thrilled to be having this conversation with you, George, because let me, let me pull the curtain back for our listeners right now.
Today, this is another conversation about AI, because George and I in parallel on our own, and then also collaboratively have spent what, the past nine, 12 months deeply experimenting with AI in terms of how we integrate it into our systems, our processes, content. You have a talk coming up at Inbound. You want to plug that real quick.
What day is it? What's that topic?
George B. Thomas: it's September 19th at 9 45 a. m. to 11 15 a. m. It's one of them 90 minute, uh, their talks at inbound and it's future of content, AI harmony and human touch. Uh, and so hopefully you can get, uh, scheduled for that. Save the seat, be there because it's going to be, it's going to be real interesting down and dirty.
A lot of, uh, Information that you'll expect and maybe a good amount that you won't expect to be in that room, to be honest with you.
Liz Moorehead: Oh my gosh. I will be front row. So anybody who's in the second row, I apologize. I'm a six foot tall monstrosity and I will be there with my little party hat and my little confetti poppers. It's going to be a good time. And then on my side of the fence, and this is where I get a little bit anxious. This isn't just me grilling George about AI.
No, I get to go in the hot seat this week because I recently published an article at sidekickstrategies. com which talks through my human powered AI content framework, which are the processes and tactics and tips that I use to actually use AI. And George, other, let's start here. Other than you sending me a quick Slack going, so you made a reference to Victorian children having consumption and I don't get it, but this article is fire.
I, I'm so curious what your thoughts were, because you give me a lot of latitude and leeway to just write what is ever in my dumb, my dumb brain. And given that AI is something you are so passionate about, I, I am curious.
George B. Thomas: Well, so let's back up a little bit because you started to throw out like a timeframe there. And I was like, well, I've been messing with AI for 12 plus, you know, maybe almost the last two years, like since it came out a little bit. You may be nine months. I feel like, uh, dragging you kicking and screaming to be honest with you into this AI conversation.
I I'm just being
Liz Moorehead: but true.
George B. Thomas: being honest. And so what's funny is when this article, first of all, when you came up with the idea of the article, I was like, Oh, I, I like this idea of this article. I think there's going to be some key things. Cause I was super curious of how does a human who comes kicking and screaming from a content strategist standpoint, how does their brain evolve?
What changes, what do they start to look at? What did they. Think about, because if you take historical SEO and content creation, and you think about this, just like plethora, deep layer of AI powered, blah, blah, blah conversation that was happening on the internet. Like, how do you disseminate to like, and here's some valuable pieces on this.
That is the other side of what everybody's talking about is AI generation from the beginning, you can create. To like AI generation, here's how you can finesse. And so I kind of giggled when the article came out and when I was reading it, but, but the first question that I want to ask you is, and again, this is a safe space.
It's all about being honest when you were in the mode of kicking and screaming. And I sent over the first, what I'll call two to three massive pillar posts that were like, Kind of AI generated, uh, human powered, my brain research, but AI generated. Like, what was your initial thought of like, Oh God, this is where we're headed.
Like unpack that for me a little bit.
Liz Moorehead: It was a two things are true kind of situation, George. On the one hand, I knew this was something that we needed to do. And this was a process we need to go through together. You know, I, I knew even though my initial reaction was, am I being punished? You weren't actually trying to punish me. But I think what happened is that I faced a situation and you and I have spoken candidly about this.
So most of this won't be a surprise. We had to just start from ground zero and throw AI spaghetti at the wall. Because the one thing I will say about it is that I think I was in a different position than maybe some other content strategists who are working with big visionary leaders on their content were in.
In a lot of cases, I could imagine scenarios where a boss is showing up saying, this is what I want you to work on and publish. I would just, I think I would have had a Like a meltdown because they wouldn't have seen quality deficits and they weren't coming to the table in the way that you were, which is okay, this is something we need to do.
This is something we need to work on. And by brute force, I am, I am, I am dragging us both over the cliff and we are going to work on this together. Now, beyond that though, when I first looked at these pieces of content, I think I looked at it in a way that a lot of people who are a little bit purist about content creation are.
Which is, there is a big difference and a much different locus of control when you are a content strategist who, I have the vision, I am setting the strategy, not just at the high level of what we're talking about over the next 90 days, but at the topic level, and then who am I talking to, to get that information, and what are the precise questions I'm asking, so I know Plucking the exact expertise out of their brains that I need in addition to their voice and tone.
So we go from that method, right? We go from that to suddenly being on a, I have been given the biggest lump of clay from which I am unclear on which the strategy came from. Not what the topics were. I understood what we were talking about. I understood why we were writing about it. But when I was initially looking at things, it was just this like.
I have to do reverse engineering to understand what this is. I have to, this is less of a strategic job and more of a cleanup job, and so there was this weird kind of topsy-turvy thing where it was like, it felt really painful because even though it might on the surface look like, well, I'm giving Liz a pizza of content, fix it.
It is an entirely different. set of skills that are kind of painful because you always feel like this is not the way I would have done it. This is not the way I would have done it. And I'm not saying that from the perspective of I would have written it from scratch with quill pen in the dark. On a Tuesday night in the rain, like, no, I put, no, I put this in the article.
I am an inherently lazy person. I just, my, my assumption for a while, and I don't think I'm wrong. I think in some cases, this is true. AI can be used to solve the right kinds of lazy in the content process, but it goes back to a conversation that you and I had. And this is, and I put this in the article. I said, what actually flipped the switch in my brain.
Was when you said, Liz, I'm, well, one, I'm not trying to hurt you. I, I want your brain to help us solve this problem because you and I both saw the same problem. People were picking up AI and churning out garbage. And they still are. Like, you and I hear this probably multiple times a week. We hear our peers going, well, that's clearly generated by AI.
Or we think it to ourselves. How many more ever evolving landscapes can we deal with in copy? Really, like, stop it. So I felt good in that you were forcing me to solve a very big problem, right, which is similar to what happened to me six or seven years ago. How did I come up, how did I become a pillar page and topic cluster expert?
Because Kathleen Booth made me, and I didn't want to. How did I become a prolific newsletter writer? Uh, at Impact, I grew our newsletter from like 1, 200 to like 40, 000 people and I'm doing it now with Beyond Your Default because I was forced to by Bob Ruffalo and Kathleen Booth and I did not want to. So we are back again doing the same thing.
But the thing that switched, the switch that flipped in my brain was when you said to me, George, do not look at AI as a replacement, look at it as an assistant. And we were off to the races.
George B. Thomas: So that's where I want to go next is because again, I think this episode is going to be for those folks, whether it's small business owners. Whether it's content creators, uh, whether it's content strategists inside of organizations, I think there's going to have to be this unlock from what is maybe the old purist form, uh, into this new AI model and understanding, or there's going to be the people who just jumped in, started creating crap, and now they're trying to backtrack into something that will actually work.
In a year of 2024 and beyond where SEO and content is becoming even more extremely difficult with changes that Google's making and the way that users are actually finding information in these large language models versus Google being a little less of the all source knowledge light and truth. So I want to go back to that.
Like when I said, Um, think of it as an assistant instead of a replacement. Talk to me about the dominoes that had to drop in your brain and kind of the inner journey, if you will, of like, Oh, that mean, cause there has to be dominoes that dropped in a level of curiosity that went in multiple directions.
Unpack that a little bit for
Liz Moorehead: Well, the first part is going to be the, probably a brutally honest answer. That doesn't sound as sexy or strategic as a one might think. But George, I think this is what you will deeply relate to as well as any solopreneur or entrepreneur or business owner who spends most of their time working in the business and on the business.
I desperately need an assistant. I cannot hire one right now. I can't. And so the moment you said it to me, I'm like, Wait, I can have one? I can have one. And so here's where the dominoes started falling, right? One of the things that I work with, you know, so we work together, we do content, uh, for our clients.
Uh, we do strategies, things like that. But one of the things that we also offer, uh, is I do content manager training. So I work with new content managers, whether you're a content strategist, content marketing manager, whomever it is that owns the content strategy. And is responsible for the production, right?
And one of the things that I always teach new content managers is this idea that I will teach you frameworks, I will teach you best practices, I will teach you rhythms, right? Be building that strategy every 90 days, every two months you should be doing this, every week you should be doing this, right?
Teaching you the foundational elements. But one of the most important things you will need to master in your first 90 days as a content manager is what your unique processes are. And so this is where I think it gets a little bit tricky, but this is where it got really exciting for me. Because I'm sitting here going, no wonder I don't relate to most of these articles trying to tell me how to use AI.
These are not process, these are not my processes. This is not how I operate. This is not where I work. So one of the first things I did is I started sitting down and I said to myself, okay. Okay. When I look at the content work that I need to do, I created, I've created two buckets, painful, heavy lifting that I have to do category one.
In order to get to category 2, the high impact value stuff where my brain actually needs to be fully present. And so that's where I started filtering in little things, right? Like, George, you know how many hours I spend, like, I think people would be baffled to learn how much I actually spend talking for my job.
I spend, Probably 50 percent of my time when we're like at full capacity, interviewing hours and pillar pages, messaging strategies, website pages, brand voice and tone workshops. I will have transcripts that are literally a hundred pages long. And what I, but that was my only way I couldn't take notes during the workshops because I had to be.
Okay. Fully present as an investigator tapping into my superpowers of asking great questions and building rapport and building trust with the people who were working with me, right? I can't be there in, but then I also afterward,
George B. Thomas: could be.
Liz Moorehead: here's the thing. So then that transcript pops out, right? That's a hundred pages worth of stuff I need to go through.
So that's where I started. I'm like, well, what if just, I tried for funsies one day. Take this 50 page transcript. Can you just tell me what the list is of things I'm supposed to be running through? And it spit out a list. And what was great is that it didn't do the work for me, but it saved me probably about two to three hours of highlighting, combing through, making sure I didn't miss something.
Like it gave me immediate direction and immediate actionable stuff. So that's where it's like little stuff like that, where it's like, Oh, Well, what are the parts of the process that hurt the most and how can I really reduce that pain?
George B. Thomas: Yeah. It's, it's interesting because I think you stumbled into a thing I want to pull out here is that, um, a lot of the time for the past two years that I've been doing AI or working with AI, it, it, it hasn't been for worksies. It's been for funsies, meaning I've been curious and just like, will it do this?
Will it do this? Can it do that? And like trying to push it and myself to like new levels of understanding what it could help as far as pain relief. With the busy day that we have. The other thing that I love that you said in that is it allows you to get past phase one to get to phase two. And when I heard you talking about phase two, it's actually where you can start to insert the
Liz Moorehead: Yeah, baby,
George B. Thomas: right.
That, that Liz adds into this. And so what's fun is, if I think about you kicking and screaming me creating the original kind of docs that came over is, there was a human level at the beginning, meaning I was doing research, I was telling AI what I wanted to pay attention to, there was an output of some information, I was putting my little thoughts and brain children around it, Then it was coming to you and you were able to get through some pain points and start to humidize it on this other level.
So that's what I want to kind of get to, too, because you started to write in this article, like, things that you were asking Chatsheepie to do and this, that, and the other thing, but for sure, I want you to start to dig into these, and I loved when we got to this section, the My AI Content Tips, Tricks, and Hacks.
And so when you look at a piece like I, uh, sent over, this list that you have, like, who's the audience? Be specific. How do you, how do they feel about this topic? Like, Unpack some of this list and the fact that it's drawing from your historical content strategist, brainset, and knowing how to interact with your AI assistant.
And what you've seen as far as like good, bad, and ugly that kind of got you to this list.
Liz Moorehead: One of the first things I will tell you right off the bat, it's something I wish I had written in there, but we're going to say it for our live audience here in the studio and also maybe on LinkedIn. I have no idea. But. You have to ask yourself to start. Are you an artist or are you a masochist?
Because you have to make that decision because if you're sitting there clinging to parts of your process because of the purity and the whatever, you will die. I will tell you one thing that has made me feel really good about what I'm about to dig into with some of these tips and tricks is the fact that.
I know because of my level of expertise, I know what to ask. I have never once since the AI machine has come along ever felt like my job was in jeopardy, even when I was kicking and screaming. Because you're right. I was like, I was a whole ass year behind you, George. I'm stubborn. My first word was no. So like, we know what we're dealing with here.
Right. But you have to ask yourself that question because you have to ask, am I clinging to the parts of my craft that matter? The artist. Or am I clinging to the parts of the process out of fear that's just itself inflicted wounds? Masochist. So when I started digging into this, I started thinking about, well, what are the ways in which, cause it started small, right?
Can you just read this and tell me the thing I forgot? And I'm also like you, I use chat GBT. I'm like, so these are the four weird things I have in my pantry and my freezer. What am I going to make for dinner? And then it'll spit something out and I'll still eat a burrito instead. It's just, you know, it happens.
But when
George B. Thomas: I like burritos.
Liz Moorehead: exactly. So when I think about my content tips, tricks, and hacks, I started to think, okay, so if I excel in a particular area, I also have to do the same to my assistant. Right? So I treated my assistant like a human. It is a robot. I named him Greg. That is my chat GBT. His name is Greg.
George B. Thomas: Nice.
Liz Moorehead: I like him.
He's fun. Um, I started to think, well, if I know what my strengths and weaknesses are, strength, interviewing, storytelling, personality, jazz hands. When I think about Greg, strengths, he can move through things quickly. He can organize things for me quickly. He can find things for me quickly. He can do lots of things really efficiently.
Weaknesses. Doesn't know how to tell a human story will only ever be as smart as my inputs and it is a Greg is very literal. Greg is Fabio Greg is a beautiful golden retriever who is just there to make me happy and give me the shit I need bleep quickly. Sorry. Hi, you know what I mean? So that is the first thing I would say you have to recognize that your Outputs will only be as clear as your inputs, because let's go back to one of the first things you and I ever did together in all fairness to you, you and I, you did the same thing I've done, but I just didn't like, it didn't get past whatever.
Cause I didn't have anybody to share it with, which is like, we're just going to see what happens. I'm a little unclear on the direction of a particular piece, so I'm going to feed it some inputs and just see what comes out. basic one on one experimentation. And then I was like, George, do you have, do you have 15 minutes?
Do you have 15 minutes? Can we talk? And we hopped out and we started reading through it together and we were both like, Oh, Oh no. And that's when we learned a really valuable lesson that if you are confused, your outputs will be confused. And when I say confused, I don't mean like you're having some sort of deke existential crisis about a topic.
I mean, If you're looking to AI to tell you what your story should be, you are already lost. You need to be showing up with those things. Now, the other piece of that is like, so don't expect clarity. If you are unclear on the flip side of that, this is where I think people struggle with how you and I use AI.
And I'm very excited to start tapping into what your processes are. Cause this hot seat is getting very steamy.
George B. Thomas: Yeah. Yeah. I like this way though. This is good
Liz Moorehead: know I'll get you back on Monday when we record beyond
George B. Thomas: This is the
Liz Moorehead: you worry. Um, When I think about how people get antsy with AI is they want it to reduce workload, right? So I think people get a little squeamish when it's like, well, I'm now asking you to do different, more front loaded work, right?
So let's go back to a different time. Once we had that weird experience where we're like, well, this is a big pile of word salad and we don't know what this means.
George B. Thomas: Yep.
Liz Moorehead: again. You and I sat down and said, okay, so let's mash our heads together and understand. What are the things that we need to tell it to do? And so we came up with a big strategy process. We basically decided what are all the questions we need to answer and feed to our Greggs and our chat GBTs and it's who is your audience? Be very specific. Don't just give me a persona. I want to know how they feel about this topic. Excited, proactive, reactive, stressed.
How do they talk about their problems? What do they want in their words? Like we really start digging into a lot of the stuff that I teach with the content compass, which is who, what, why, how. Those are your four pillars of any great piece of content. Who are you talking to? What do they want? Why are you the one to answer it?
And how are you going to help them? That is the blueprint for any great piece of content that you do. So that is the part I think that kind of turns people off a little bit. You mean it can't just do it for me with a one word prompt? No, genius! I'll tell you what I told a friend of mine over the weekend.
I can't, I can't meet your needs if I don't know what they are
George B. Thomas: Yeah.
Liz Moorehead: robots feel the same way.
George B. Thomas: Yeah. It's so cool because there's a couple things that I want to unpack here that, by the way, some of these are just like life sayings that you may have heard or philosophies that you may believe in. But. One of the things that when I was going into playing with AI, that I was keeping in mind is something that my grandpa always used to say, and that's garbage in garbage out.
And so I immediately was like, well, I don't trust the garbage pile. That is just the overarching large language model. So let me make sure I'm putting air quotes here, good garbage or good data or good information in to the conversation that I'm trying to have. So you got to pay attention to like what you're putting in first.
To then tell it what you want to get it out. And here's, what's funny too, is it's like some people when they start using AI, I feel like they just kind of like lose their brain or lose their way. Because if you asked any creative. What's one of the most important things that you have in your process?
They would use these words. Oh, we have to have a creative brief.
Liz Moorehead: Oh God.
George B. Thomas: we were creating Liz was basically like an AI version of a creative brief of what we wanted that assistant to do and the things that they needed to know about the project that we were building. And so did an, did a brief take time in the past?
Yes. Is an AI version of the brief going to take time? Yes. But is the output in both situations going to be much better than if we tried to run through the scenario without it? The answer is yes. And so here's the other piece that I'm going to throw in here is like, we got to remember that there's fundamentals in life for a reason, like somebody did just wake up one day and say.
You know what? For business, we should have personas because it'll be something cool to talk about. No, it's because you literally know the humans that you're actually trying to serve. And that's what I want to go into for a second here too, is like, I was not coming at AI to how can I get the work done quicker?
I was coming at AI from a standpoint of how can I create something much richer? So talk to me about this mindset of yes, it will speed up your process in certain areas. But at the end of the day, you're going to be able to build something more rich and robust that actually serves the
Liz Moorehead: Oh
George B. Thomas: better than what you may have done before, like unpack your brain around
Liz Moorehead: Absolutely. I mean, one of the things that AI, Greg, my, my pal Greg, or any AI tools, we're talking that HubSpot's AI content assistant that pops up in all of your rich text fields on website pages and, and blog articles. It It challenges me to be more and, and what I mean by that is, is that it, I can sometimes get very tunnel visioned in my ideas and I'll get really stuck on a particular topic and I won't see the forest for the trees and I'll forget things, right?
Whether we're talking about, Hey, can you just look at this transcript? One of the things I love to do when I am challenging myself, not even in writing. I will feed it a bunch of research. So we have a podcast that we do called Beyond Your Default, and it requires a lot of research that I need to conduct very efficiently.
Once I pull all that research together, I will sometimes throw it all in there and I will give it a huge brief. I will say, this is my vision for this episode. This is some, this is an example of stuff that George has talked about with this topic. These are the basic bullet points of where I think I want to go.
Um, And here are a bunch of raw materials, like, for example, just pulling back the curtain, like, super radically, we're doing a, an, uh, we're doing an episode next week about spirituality and feeding your soul, but I'm also bringing in secular perspectives, and Carl Sagan, who is a famous astronomer, And writer and philosopher and famous atheist has some incredible work that talks about soul and the human experience in a, in a large, vast, infinite universe.
I've read Cosmos a thousand times, but I have not read everything that Carl Sagan has written. So I had him pull stuff out. But then I also said, now I have the, all these things. Can you just spit out, I'd like, I usually say like 30 questions. What would you ask in this conversation? I usually only pick maybe two or three of the 30, but all of the stuff that's there tells me, Oh God, I would have never thought of that.
I'm still the originator of the original questions that end up going into our conversation, whether that's cherry picking the ones I want or being inspired by something else. Oh, what's up, Nick from Fargo. That's right. Beyond your default. com forward slash newsletter twice a week. I will rain hellfire in your inbox and tell you to do things you should be doing.
Um, but that is where I start thinking about the other way I like to use it too, is that sometimes there's just meat and potatoes. Yes. Okay. There's meat and potato stuff in your content. Like, here's another thing, know your AI gaps. There is no aspirational rhetoric that AI robot will not try to pull into your piece of content.
I don't care. It doesn't matter how many times I'm like, can you please just describe the problem? Can you please start by describing the, it will start with the solution every time. So I always know I have to be much heavier there, but it allows me to expand things out. Like, can you just give me 400 words on something about like, I'm just making something like Salesforce versus HubSpot or something like that.
I know I'm not going to
George B. Thomas: Sounds like a great video, by the way. That sounds like a great video in the future.
Liz Moorehead: like, I just pick it up and it gives me some messy stuff to play with. It inspires me. I usually will still write from scratch. Maybe I'll pull a couple of things in, but it'll still do two things. Shorten my time to phase two, where I'm actually creating.
My brain is actually moving. I'm not just staring at a blank screen. But it also will usually say, well, shit, I would have never, I would have never thought to bring in that parallel. I would have never thought to compare those factors because we always go for things like pricing costs, like it just, it makes me clear as an investigator that I'm, I'm tackling all of my arguments.
George B. Thomas: Yeah,
Liz Moorehead: So, George, are you ready? Because I have a few questions for you.
George B. Thomas: okay, yeah, yeah, I mean, I might not be done with you yet, but go ahead and let's, let's go after
Liz Moorehead: We'll get one final question with each other. How about that?
George B. Thomas: Okay,
Liz Moorehead: my, here's what I want to know from you because you are in a very interesting position right now.
George B. Thomas: yeah,
Liz Moorehead: One of the reasons why I'm going to be in the front row of your talk is not only because I'm your number one fan and you're my friend and we work together, but also because I'm very interested to learn from you on this topic.
But this is a bit of building the boat while you're sailing it. Like you are actively learning. Every single day and iterating the processes that you are going to be teaching live on stage at inbound. And so I'm curious for you, what are some of the most recent light bulb moments you've had? Cause I've looked at this deck over the past few months and boy, is it changing?
George B. Thomas: yeah,
Liz Moorehead: the most recent light bulb moments you've had where you're like, Oh crap, now I got to throw that out.
George B. Thomas: yeah, that's, it's, it's rough because by the way, my rough draft is due and I'm like, guys, you don't understand the world that I'm living in right now. Like, can I just show up and give the talk? Because this could change like a thousand times, which I know I'll have the ability to change it. Cause I'll be doing it for my computer and different stuff like that, but you still have to follow the rules.
But, uh, Liz, here's the thing is like. So many things have changed. Like, first of all, let me tell you a little story. One of the things that I think makes me interesting or maybe unique is I have this mindset or this ability or power to say, how can I optimize the crap out of something? So, so let's, for example, let's remove AI for a second and let's talk about video production, right?
When I worked back at the sales line, um, Marcus and I were talking all about like, it's the year of video, by the way, this is 2014 people finally caught up, but we're like, okay, so how can we optimize this? We used to live in a day where you'd set up the gear, you'd film the stuff, you'd have to pull the cards out.
And early on, I was working on systems where the camera was plugged right to the computer and the file could just go into the editor. And then it was like, well, this switcher and this board and this mic, and if you set it up this way, then you could just show up. And I got to the point where, you know, it's, it's nothing for me to probably create three, if not five videos a day, a day and publish them onto the internet.
But that's because we removed any insufficient, like, pieces of the process where it was just like, step up, hit record, uh, 30 minutes later, 45 minutes later, there's an edited version, and it's published on YouTube. So, I have this mindset with AI, and so I've been working on like, sure, you can generate content with it, but what's the most optimized way to do research?
Sure, uh, when I type, I peck, like, you know, one finger at a time. Well, how can I use AI to be the most optimized typist? Not a writer! There's a difference. A writer has to be creative. A writer pays attention to a creative brief. A writer does all of these, but a typist types. You give them ish, and they type
Liz Moorehead: They produce.
George B. Thomas: do I turn AI? Yeah, they produce, right? So how do I turn the AI into the typist, not necessarily the writer? Or the human or creative side of what they're actually outputting or producing. So the reason I'm bringing all this up is because everything is about how is it human powered and AI assisted to remove any efficiencies.
And I'll give you a great example of one thing that I'm going to add to a PowerPoint presentation that I kind of giggled, but I was like, it sounds simple. But I know that nobody's doing it and the fact that nobody's probably doing this and the amount of time it will save people. I have to talk about this as dumb as it sounds.
So here's probably how most people use. A. I. They input some information. Because they know they got to do some research before they actually create the thing, but every time they put in, let's say, a URL, it will tell you about the URL that you put in and you literally have to sit there and wait for it to give you the information on the research that you gave it in realizing that you don't need to know about the information that you gave it.
It's literally giving you a summary of it.
Liz Moorehead: I told Greg to stop doing that. I told Greg, read this and just tell me you read it. Stop summarizing. And that's what we do now.
George B. Thomas: literally the prompt that I put in there is I, when I go to start research, I give this prompt of do not give me any output on what you learn. Just add it to this conversation. Summaries go away at least 5, 10, maybe 20 minutes, depending on how much research I'm doing is saved. And now I can just get all the research in and turn to the like, okay, now let's get the creative brief portion of this going so that we can get real creative and start the typing portion of this.
But again, that's. Using your brain, the human power to get the AI to assist. So that's just a small example of something that if you asked me 30 days ago, are you going to put that in the presentation deck? I would have been like, no, a week ago, four days ago, I was like, this has to be in the presentation deck.
Liz Moorehead: Okay. So the next question I'm going to ask you, I don't know if it's something you've already tackled in your presentation and if not, I hope you do because this is coming from a content nerd who is working with. And you know, we're both business owners, but when we work together, like I'm working for you, you know, like we partner on things, but when we're doing this, I'm working for you and I'm working for sidekick.
And one of the things I will tell you is that the only reason I'm probably ever got over the kicking and screaming hump is because how you approached it with me. And because you could George, you know, but you could drag me anywhere and I will still say no. I would just, I would just say, no, I'm a toddler in a crate and barrel with a mom who does not want to be there looking at spoons.
Like I want to be watching Teletubbies or whatever. So like,
George B. Thomas: That's a flashback.
Liz Moorehead: know. So my question to you is this, what advice would you have for business leaders? Marketing leaders, people who are probably going through the same process where they have put upon people or maybe people who are excited but resistant or confused and they can't wrap your brain.
What are the tips that you have for them in terms of how they approach this relationship with their wary would be content nerds?
George B. Thomas: Yeah, well, so there's a couple of things that jump into my brain there. One, um, you can't just sit on your mountaintop and say, Hey, I want you all to start doing this. Um, you'll find that most leaders go through the trenches with their, uh, soldiers and, you know, and so if you're going to ask people in your organization to do this, you should probably also be leading the way and figuring things out in yourself.
So like, for instance, with, with us, Liz, you know, That one of the things I immediately did was create a slack channel that is the AI conversations and any prompts that I would create that had positive effect, I would share in the slack channel. So the rest of the team would get these at that point and knowing pieces of information of prompts that they go.
Until they realized it was a library of prompts that they could go to and actually copy and paste and tweak to do the work that they were trying to do. So I was leading the way. So if you're listening to this and you're a leader and you want to create these efficiencies in the places where it matters with AI, then you need to kind of lead the way, be testing, be, be going with that.
The other thing that I will say. Is you have to also give them time. Liz, I talked about and showed you AI stuff long before you were like, okay, let me go ahead and dig this a lot. And so, but it was that space to like, let them brew through it in their own brain. Um, let them have, as we've talked about on this podcast episode, their own aha moment, because that's the thing.
What, what I feel like us as leaders were meant to do is we're to help people, um, understand, gain insight. And because of that insight, have an aha moment. And then therefore take action because the actions that you've taken after that insightful aha moment. are dramatically more impressive, um, and expansive than if I would have been trying to be like, uh, I'm going to beat you with this ruler till you use AI.
Like that would have never, never worked. It would have never
Liz Moorehead: completely agree. And one of the things I will also throw out there too, just on the other side of it that I thought you did really well was that I, uh, I never felt at any time. Like I was at, like, I was joking about it, but I wasn't dragged along. You, you looked at me and said, I want you here. Working with this on me because I value your content brain and we are going to solve this together.
We are building this together. And I think that's another point that I really want to just double click on for any listeners out there who are marketing leaders or anybody who is in some sort of top down or management situation with someone where you're leading the charge and somebody may be a little bit dicey about it.
You have to genuinely believe this, right? You can't just like ask him a couple of questions and say, I checked the feedback box. But it was literally like, we're going to mash our brains together. We're going to work on this together. And the reason why you are the person working on this with me is because you are the content expert.
And that just like was radically different for me. Um, One question I do have for you though is like, what are some of your most, I don't know what you call it, funny, hilarious, but moments where you're like, well, I'm not going to do that again with AI, that was a fun life experiment. And we're just, we're, cause you talk a lot in this deck and this presentation about like the things we should be doing, but what are those behind the scenes?
Like, no, no, no, no, no. So let's not, let's take a step back.
George B. Thomas: Yeah. Um, it's funny because one thing I've learned is that the more complex You try to make it the more I will start to hallucinate and you'll just get some real radical information. And so 1 of the things again, I'm always trying to kind of simplify the complex and keep the conversations and the prompts very specific.
very actionable, because as soon as you start to, and we're good at this as humans, as soon as you start to come into the conversation a little bit sideways, um, or you start to use a lot of jargon or things like that, where it'll just, it'll start to just deteriorate and fall apart. And it's hard to see that it's deteriorating and falling apart until you try to get that actual output.
And then you're like, oh crap, here's a real small thing for you, by the way. And I was like, this is interesting. Um, I'm going to change this moving forward as a creative Liz, when you talk about what's on the page, have you ever used the word copy? I'm going to go ahead and create the copy for the page.
Like that's what we always called it at different agencies that I would work at is yeah, create that copy for AI it's text. AI really doesn't know what copy. Or to copy something is to duplicate something. And so as soon as I switched from saying, create copy for this landing page, and I switched it to create the text that I need for this landing page, the output was dramatically different.
So again, not a like. Massively brain exploding thing, but the difference in a word to the output of what we got was substantial. Um, the other thing is that I, I wish that I would have learned something and implemented something. Earlier, and it took us a while to like research it and get it in place the certain prompt, but the amount of articles or content that I created that I handed over to you for us to go through this process that said delve,
Liz Moorehead: Oh, delve.
George B. Thomas: or, or other AI like loved words.
Um, God, I wish I wouldn't have done that, or I wish more people would learn to do, we literally have a prompt, which now, by the way, it's not even a prompt, because I have it in my chat GPT memory, so it knows with every conversation I have to ignore these words, but it's literally, uh, a prompt that's, oh my God, it's, it's It's a prompt.
If you're listening to this, you can't see my hand gesture, but it's like the size of my head long of words not to use when creating content. Um, and so just like that element. In the difference in what is output. Um, it's, it's funny. I've had, I did a session for one of our clients, uh, for one of their communities on AI and how to use it.
And, and it was kind of like the precursor to what's going to happen at inbound. It's definitely not the same thing at all. But what's funny is this, this human, this human said that, you know, George's chat GPT is the most tuned in GPT that I think I've ever seen anybody use. And I want everybody to understand that because here's the fatal flaw.
For probably the first six to nine months, Liz, I used it without thinking about what car I wanted to turn it into, or how I wanted to tune the engine. And the magical piece for me, when we're thinking about what don't ever try that again, is I would never use a vanilla type style chat system again. Like the fact that HubSpot now has where you can put voice and brand into the blog tool.
Beautiful. I'll use it. I'll have fun with it. I'll attempt it. I'll test it. If there is a tool where I can't give it instructions, or I can't program a memory, or I can't give it the, you know, what my friend Remington would say is guardrails and goalposts of what we're actually trying to do on a day by day.
But I would never do that again. Because again, when you take time to tune it into what you want it to be, Okay. To eradicate the efficiencies that you know you have in your process and to build something that's deeper and richer that you want to provide for your community, I don't run into a lot of those, like, Oh, what's this pile of heaping crap?
I used to have those. I don't necessarily have those that much anymore.
Liz Moorehead: Amazing. All right, George, you get one more bite at the Liz Apple. Make it good.
George B. Thomas: So I'm super curious. What is the thing that That you've asked AI to do that you never thought you would ask AI to do in your life But now you have done it.
Liz Moorehead: I would say poetry, but if anybody's been listening to this podcast for any length of time, we know that's not true.
George B. Thomas: we've done that.
Liz Moorehead: What is the one thing I ask it to do that I never thought I would?
This isn't going to sound very sexy. This isn't something mind blowing or groundbreaking, and it just comes back to something very simple. And this is where I think people will, you know, I think we're looking for big, audacious, crazy ideas of how we should be using these tools. When in actuality, the most impactful things it's that 80, 20 rule, right?
Like 80 percent of your efforts come from, or any 80 percent of your results come from 20 percent of what you do, right? It's the same thing. Like a small incremental ask will suddenly have an outsized ROI. A couple of things, one on the keyword research side. So this is something hyper technical for my big content strategy nerds out there.
So we've all been there. We've been in SEMrush. We get one of my favorite things about SEMrush is that was where I get to actually be an investigator because I have to use my little, my little human hat. I have to say, well, what are people actually searching for? What are people actually looking for? It's very interesting.
I love data because data, data is the most human. Part of content strategy development for me because data is just the quantification of human behavior. It tells me where the people are. It tells me what the people want, right? So I love those little ones and zeros and I'm like this in my personal life, too.
I have one time somebody told me it's like you're like arguing with a calculator. I'm like, thank you. But you know, I I love that, but here's the thing that's tricky is that when you finally find that right broad match keyword that you really want to dig into, you will get hundreds, if not thousands of results.
So one day I just got curious and said, I'm going to give you a big prompt. I'm actually going to give you a transcript of a conversation. I'm going to give you that. And the conversation had to do with, uh, a pivot and content strategy. And it was between us and a client. Then I fed it a bunch of different prompts and said, this is my overall vision of what I'm looking for as this company's content director.
These are the things that I'm moving away from. These are the things that I look at when I'm evaluating SEMrush results. I'm looking for a particular competitive density. I don't want commercial or transactional intense. I want informational. I don't want anything that's geo specific. I don't want this. I do want that.
Like it's a big laundry list. And again, that's something that took me five to 10 minutes to write up. But again, this is usually a two to four hour activity. And I fed it all that stuff. And I said, could you give me, I think it was 50 or 60 opportunities. List out the volume, sort it by keyword density, which is usually how I like to sort it, and maybe give me a few topic ideas of where you would take some of these keywords. Did I still have to do some refinement on the backend? Absolutely. Again, I'm a strategist. This is just about, this is too much data for me to go through. And then also I still manually went through some of it, but I was able to skim much faster. I was able to move much more quickly. But what was really cool is that there were a couple of opportunities where I'm I would have never.
George B. Thomas: Yep.
Liz Moorehead: never thought of that. The other piece of it is on the backend. Hey, I just wrote this totally dope, totally human piece of content. Can you just give it a quick check over to see how you might have optimized it differently based on this keyword, based on this attempt, intent. And by the way, here are the top link to our 10 URLs to the top 10 websites that are currently ranking for this term.
And I want to beat them. I still do the human work on the backend of it. I make the changes. I decide what goes in, what goes out. Uh, It's still all my work. It's still human powered. So that, that would, I would say like, again, these are not sexy things, but if we think about the things I never thought I would ask, you usually think about AI for inputs and outputs in terms of create this piece of copy text, whatever for me, but it is an incredible optimization tool because it is, it is a machine.
And it knows what to look for it. Now, again, you have to have the expertise, right? Like if you have a more green content manager who hasn't been doing strategies like I have for 10 years, they're probably not going to be as detailed in the prompts that they're doing again, but I've never worried about being replaced.
So George, here's my final question to you.
You've been listening to a lot of this conversation. Obviously you're here. You've been asking me questions. I've been asking you questions. What is one lightbulb moment you are walking away with from today that's either going to influence how you use AI going forward or the talk you're about to give in a couple of months?
George B. Thomas: light bulb moment. I think for me it's it's interesting because Sometimes when I have this conversation, I'm so bought in and I'm, I'm so like down with it and I, I have to realize that there's probably going to be a large part of the audience that is there because they're down with it. but there's also going to probably be a large portion of the audience that is there because they're just curious of what's possible.
And so one of the things that you're, you're asking me questions and, um, I thought you were going to ask me a question that was like, how would you, um, tell leaders, um, or what steps would you tell people Um, to get started, right? You didn't ask me that question, but my brain went there and I'm like, I don't have anything in my presentation deck right now that covers the like getting started portion of this.
Like I really do jump in to the deep end of the pool of like, check out all these cool prompts. And by the way, you should talk to it. Like it's a human. And did you know that it can actually be your assistant for the things that are in a Like, I mean, I just dive in like immediately, like, here's what you're trying to do.
You're trying to hit the easy button and creating a piece of crap, but you don't want to create crap anymore. Let's create good things. And so I'm, I'm sitting here understanding that my draft due date is passed and I probably need to be like, All right, here's the five things you need to do and think about to even get started.
And I want to ask the question now, when I first start, like by a show of hands, how many of you are right now using AI to generate any type of content in your organization? Because it could be like years before, where all of a sudden, I think it's going to be a room full of people that raised their hand.
But only 10 percent of the people raise their hand and I've got to be ready to pivot and have that piece of conversation. Now, what's nice about that is if like 100 percent of the audience raised their hand, I could skip through those like 5 or 6 slides real quick and be like, okay, we don't need that and actually dive into the deep end.
But I don't I don't have that safety net right now. And so, for me, that was an aha moment of like, oh, God, she's about to ask me a question that I don't necessarily have an answer for.
Liz Moorehead: Well, I,
George B. Thomas: I was immediately like, I should probably have an answer for
Liz Moorehead: I would make the argument though that the conversation we did end up having though about how you bring your people along with it is a, is a critical part of that getting started process. Because I'm coming at it from the selfish perspective, right? Let's talk about me and how you managed me, George, because I'm a pain in the ass and that's what we had to do.
Right. But when we think about it, you know, the thing I want to point out here is that this is the part where it's like, it's so easy for you that you don't realize it's magic to others. Like you natively default to thinking about processes because the human part becomes so easily to you. And so I think one of the things that people need to remember, and hopefully this makes it into your talk, is that part of the getting started process is taking a look at the humans who need to come on this journey with you of AI and catch the vision.
And I think that's really critically important. And then one last question for you, George. Because you always, you always put a bow on these conversations so well. What do you want our listeners to take away from this conversation today? What, what are the big things that you want to stick in their brain?
George B. Thomas: I mean, the, the big things for me is, you can't ignore it. You really can't run from it. And so the quicker you embrace it, and the quicker that you put the right mindset to it. Like, for you, Liz, the unlock was, think of it as an assistant, not a replacement. Yeah. There's going to be different unlocks for different humans.
And that's going to be an interesting part for me over time. The more people I talk to about AI and what we're doing and how we're using it is like, okay, what do I need to search out their unlock? But here's the thing that I want to leave you with is that the rest of the world is probably preaching and teaching this to you in a way that you don't even understand, but it's fundamentally wrong. Take a look at the internet. And what you're going to see is that AI powered this, AI powered that, AI, AI, AI. Ladies and gentlemen, this all changes when you flip it to the way that it should be. When you're using AI, your mindset has to be, should be, it's human powered, AI assisted. That's the big takeaway that I want you to have.
Human powered. AI assisted, in each step that you use it, where's the human power? Okay, now let's assist that. Here's the human power. Now
Liz Moorehead: Uh, Nick, Nick from Fargo has a, has a question for you and that's, it's what powered, George?
George B. Thomas: Oh, it's G
Liz Moorehead: That's right. And by the way, if you think we're ending a whole podcast episode of, about AI without an AI poem, sir,
George B. Thomas: here we
Liz Moorehead: ready for me to take us out with a little haiku,
George B. Thomas: Yeah, let's, let's hear a haiku.
Liz Moorehead: AI types away. I pretend to work hard, but I'm napping all day.
And with that, ladies and gentlemen, we'll talk to you next week.
George B. Thomas: We're outta
Liz Moorehead: Bye.