23 | Turning AI into a Key Player in Your Marketing Team with Liza Adams

In this episode of The AI Marketers Playbook, Audrey Chia sits down with Liza Adams, AI executive advisor and marketing leader, to explore how businesses can integrate AI into their teams. Liza shares her unique perspective on AI as a thought partner, not just an automation tool, and explains how AI can enhance strategy, not just execution. Learn how to create a human-AI org chart, leverage AI for go-to-market strategies, and ensure your brand shows up in AI-driven search.

Join my weekly Newsletter: https://lp.closewithcopy.co/welcome

Transcript

Audrey Chia: 0:00

Hello and welcome back to the AI Marketers Playbook where we cover actionable frameworks to help you leverage AI and marketing strategies in your business. I'm Audrey Chia, your host, and today I have with me Liza Adams, someone I have been looking forward to speaking with. So Liza is an AI executive advisor Who has been transforming the way B2B businesses approach AI integration. With more than 20 years of senior marketing experience at Tech Giants, Liza brings a really unique perspective on AI implementation that really goes beyond the basics. Today, she's here to share her insights on building effective human AI marketing teams and creating amazing content today. Welcome to the show, Liza.

Liza Adams: 0:40

Hey, Audrey, it's so exciting to be here. I've been so looking forward to being on your show. Thank you for having me.

Audrey Chia: 0:47

Awesome. And for listeners who don't know you, you know, tell us more about your background, how you got started and how did you land in the world of AI?

Liza Adams: 0:55

Yeah, so, a little bit of background about me. I was born and raised in Manila, and I came here as a foreign exchange student when I was here in the United States as a foreign exchange student when I was 14. Ended up in engineering, so my background is engineering, but I realized that I can tell stories and I can make complex things more approachable and decided that I wanted to be in marketing because of that storytelling aspect of it. And. You know, I have not looked back, right? Although I still have that very engineering mindset, very structured in my thinking. and I think that's helped me in the world of AI, especially in, in giving it instructions and in guiding it. And in thinking about strategy and making big things. more consumable by chunking it down at the bite sized pieces. But I, like I said, I've never looked back. I really just enjoy capturing hearts and minds. And when you capture hearts and minds, the wallets will follow. I love

Audrey Chia: 1:57

it. Let's, let's get it on a

Liza Adams: 2:00

t shirt. So, you know, I've, I've seen. Spent a big chunk of my career in Silicon Valley working for some very large companies. but then I, I found a, a company where I headed up marketing. Here in Colorado and Audrey, honestly, I thought that it was going to be my last one. cause I, I, you know, I'm in the back nine of my career. If you play golf, I can see the clubhouse. I don't need a golf cart to get there. And, very purpose driven company and and had exceptional product market fit had a culture that felt like home and I thought, you know, maybe 3, you know, 3 to 5 years and I will be done and do my own thing, but it wasn't meant to be about a year and a half into my tenure there. We were acquired. I was a little bit sad about it, right? And, and I was actually pretty distraught. And I started thinking about serving on boards. And I, realized that there were only 41 marketers on Fortune 1000 boards and less than 3 percent of all, board members, regardless of company size, had marketing experience. So. I was further distraught by that because the chances of landing on a board and survey would be a lot lower. So I dug in deeply to figure out why it is that marketers are not on boards. And 1 of the biggest there are many reasons, but 1 of the biggest reasons is that we are. Perceive this tacticians rather than strategists. So people see the websites or social posts or an event, or just a logo. And believe that that's all that marketing does. And the, the things that we do, where we deeply understand markets and customer journeys and great messaging do positioning do targeting great categories and even lead categories. Those kind of fall by the wayside. So, I, And having come from product marketing and go to market strategy, I was deeply passionate about the strategic work of marketing while my exit from that company just happened to coincide with the launch of chat GPT in November, 2022. And, I, when I looked at chat GPT, I automatically saw what most people didn't see initially. Most people, even today, think about it as a question and answer engine. Or they use it for, basic tactical things, like create me a blog, or, summarize this email, or, you know, create an image for me. But when I looked at it, I saw it more as a thought partner. Where I can actually collaborate with that, share with it with, a, I, my ideas and it could give me different perspectives. I also see it for automation. so not just content creation, I also saw it as a way for us to personalize experiences and content. And then, lastly, you know, for research and analytics, be able to. process a lot of data and, and get a lot of insights from that data to inform decision making. So, given that I felt that this is what we needed as marketers to flip the script to allow us to really lean in on the strategy, because we can now do those things a lot more quickly. Unlike in the past, where people didn't have a lot of patience to do market research, a lot of patients, not a lot of patients for strategy work. But in today's environment, now that we can do that fairly quickly, almost in real time, we can now really lean into the strategy piece of this. And when we get strategy right, the execution gets A lot easier and it also becomes more impactful. So that is essentially my journey into AI. And then now I am advising companies, CMOs, go to market leaders and inspiring them with what's possible with AI. From the really tactical things, they're very strategic things and helping them create this human AI org chart where we have AI tools or AI custom GPTs that are now teammates that are in support of the humans, in our organizations.

Audrey Chia: 6:02

What a journey and there is so much to unpack in just, you know, that short introduction, right? So the first thing I wanted to highlight was actually what you said about AI being your thought partner I think this is very interesting because a lot of people still see AI perhaps as their intern to get work done But at the same time AI can also serve as that strategist that works with you, you know At that same level of productivity or that same level of, you know, efficiency, if you train it well, and if you know how to prompt it. And also you mentioned something interesting about building like human AI teams as an organization. How do you see that happening? Because most people are looking at AI as a tool, but not looking at it as, you know, a part of the organization.

Liza Adams: 6:42

Yeah, so this is a huge unlock, Audrey, when You know, I wrote a post about this. I think it was the title was like human AI organizations. And all I did was actually create an org chart and it's not my original idea. I saw it somewhere else. Right? And there was actually a CMO that she had a very small team and she. Needed a lot of help. She didn't have the budget nor the resources to add more people. So what she did was she essentially said, alright, let's create some custom GPTs to help us with the mundane and more tactical work. So as an example, they had a custom GPT for brainstorming content topics. Based on a specific persona and based on where that persona is on the buyer's journey. So they called it a content topic brainstormer. And then they also had a custom GPT for creating LinkedIn posts. And they had a custom GPT for creating competitive battle cards in product marketing. So the human beings, in essence, train and build these custom GPTs. And the big unlock when I saw it was she had it on an org chart. So she had product marketers and that underneath the product marketers was a box. That said, competitive battle card builder. That was the GPT, right? And then there was another box that was a market analyzer. So at that point, I was like, my goodness, people are thinking about this as a question answer engine or thinking about it simply as a tool, like Excel or PowerPoint. But when we actually give it a specific role and a specific task, and we put it in a, into an org chart, then the thinking shifts. It's now more of a teammate. It's now more of an entity. It's not human that you can essentially give work to and assign work to. And they do that work. And they're super good about just doing that simple task. So I think it is such a big unlock when we think about AI. As a teammate that, that we build, that we guide, manage, and maintain in support of our work. Now, all these AIs can now do the repetitive tasks, which frees us up to do some of the more interesting and more impactful work. And that depends on the role, right? It could be more strategy, more insights, better aligning teams. Thinking about people, how do you progress your careers? And oftentimes we don't have a lot of time to just really think right. But when AI does all those repetitive things, we now, it frees us up to do more work and it frees us up to be more curious about AI. Right? what also has changed is our roles. Audrey, because our roles are now, you know, the addition to our role is the fact that we now guide AI. We now build, maintain and manage AI. So that is then a shift in the role. But when we guide, we're training these things to think like us, right? We're training them to be creative. Yeah. Training them to be mindful of, you know, fairness and transparency and ethics. And I think that is also an unlock for people. You know, AI, when you give it very generic prompts, you will get a response that is, You know, based on the 85 percent of the Internet that it's trained on,

Audrey Chia: 10:14

but, but when we, it's not great, it's not great,

Liza Adams: 10:17

but when we teach it and we guide it based on what's highly relevant to us. Then it can give us answers that are more impactful and absolutely more relevant. So that, that unlock with an org chart, like I simply put an org chart up there, and then people are like, ah, that's how I need to think about AI. And right now it's a lot of like mundane tasks. But over time, right? Use it as a thought partner, and maybe there's a thought partner AI, you know, teammate in there. and then as we go into agents, they'll be able to do more rather than us just, you know, you know, guiding it to do specific things. We can automate and it could do things on our behalf.

Audrey Chia: 10:59

Yes, and I really loved it. How you know it does. It's just basically a matter of framing even. And it's so interesting that a simple shift in perspective on how you leverage AI can also change the way workflows are created, how the organization sees it. And I think it's very interesting when it comes from a maybe a founder or management led approach, right? Because now it's almost like at the top, the team leaders are saying, Hey, look, let's adopt it, embrace it, integrate it, and I think it's a lot easier to. Build out a productive team rather than having it from ground up where they are leveraging it, but just as tools, and there are no proper systems, no proper processes in place. I know you also have done a lot of go to market work for founders and of course, companies of different sizes. So can you give us perhaps like three really actionable use cases for a go to market strategist or a founder who is trying to use AI, but have no idea where to start?

Liza Adams: 11:53

Yeah, so, this is one that I talk about quite a bit. because in today's environment, the markets are shifting. So customer behaviors are changing. The way we search has changed, right? Like, many are using AI perplexity, chat GPT, cloud, to do searches. and then, you know, we, in addition to the way we search, the way we buy also has changed. Our preferences have changed. And in fact, you know, there's the saying, in marketing perception is reality. But in the AI era, I would contend that we're heading towards the, this whole notion of the AI response becomes reality, right? So you could actually ask perplexity or chat GPT, a question about a company. And it gives you an opinion. About that company, so it forms an opinion about a business or a person or a leader prior to us making and having an opinion about that person or leader, right? Because it takes signals from everything, whether it's the company's messaging or employee feedback or customer reviews, it takes all those signals, good or bad, and it amplifies it and whatever it says. Becomes the truth or becomes the reality now people always say, oh, you should actually, you know, check AI's work and we should. Right. But we are human beings. If we like shortcuts, like, Audrey, when was the last time you actually scroll down to the 2nd page of a Google search rarely, right? Same thing is going to happen with AI, where perplexity responds, who will actually check individual citations to make sure that it's relevant and accurate, right? So there's this whole thing around changes in buying behavior, buying patterns, preferences, and journeys. So, given that, I contend that in most companies, that product market fit is not there, because the buyers have changed, but our products have not changed. So, if the buyers have changed and our products haven't, then we have a misalignment in meeting that market's need. Or, if we changed our product, but we haven't looked into how our buyers have changed, then Then we could potentially have a misalignment there. Right? And I don't believe that any company is immune from this. So even when we look at Google and Google search, there being, there's a lot of pressure on them, right? That product market fit in search, it has a lot of pressure from the AI companies as a result of a zero click search, like what I've described. So, and then when people say, hey, we want to go up market into the enterprise, well, going up market to the enterprise is not an easy thing. You might have to change your product to make it scalable, to make it, more secure and all sorts of things. So in using AI, I always say. Before you just start infusing a I in the tactical things like in the campaigns, ensure that you have good product market fit, because if you don't ensure that you have good product market fit, what happens is it is literally like doing an all out campaign for snowblowers in Singapore.

Audrey Chia: 15:07

Not a great idea.

Liza Adams: 15:10

Because it's not a problem, right? It doesn't apply. So AI will amplify that misalignment and product market fit. So what I guide people towards is okay. Yeah, use AI and some of the campaigns, but ensure that you also use AI. With some of these more strategic decisions to make sure that we have good product market fit, right? So to get really, practical about this is identify the top market segments that you can address based on the products that you have. And that could be, based on a set of criteria, the set of criteria could be, let's say, market size, market growth, when loss ratio, strength of partnerships. And I've actually used AI using all this data to parse through all that data and begin forced ranking by segment of the market, which one has the highest market size, which one has the highest win rate. And then when you look at all of these segments and look at all of the criteria, AI can actually do that analysis and then down towards the bottom, you could have a total that says, the best fit for your business based on your criteria. That is X, Y, and Z, because AI can go through all of that analysis, parse all of that data due to calculations to help you with some baseline decision making based on data. Now, whatever AI says doesn't mean that that's what we're going to go do, right? But we now have something, from which we can align the executive team and have a conversation because it is based on data that we can then validate with the market. So that's just our one example. Well, the other example that I love to use, Audrey is using customer reviews. To develop content are, you know, in, in marketing, we, we use, we fall into the trap of using it ease enables. Like we say, scalability, flexibility, you know, and no one talks like that, right? AI driven, flexible, future proof, all sorts of things. What I love to do is, looking at real customer language. And without having to spend a lot of money doing customer surveys or customer interviews, I look at customer reviews online. I look at transcripts of sales and customer calls or customer service and customer calls based on all of that. Obviously, we redacted to ensure that we don't have any confidential or sensitive information in that we use that to then think about. What strategies might it inform us? you know, based on what they say, like, based on what they say, what are the key problems based on those key problems? Is it something that we can address? And then we can also use it to help us with messaging. It could help us with positioning. And we can use the exact same language that they use rather than coming up with, you know, marketing lingo that none of our customers truly use. So, that's a very tangible use case. Like, if, if, if marketers are not using reviews and, and transcripts. And they're not using the language that a customer uses. I think we're in a dangerous situation where we could be highly irrelevant because if they see language that we don't understand, then customers tend to think that it's not for me.

Audrey Chia: 18:30

And I think what you said about that is like also drawing from real customer insights Like a lot of messaging that converts eventually is based on actual customer insights There are pain points which a lot of times marketers might be like I know my customers really well But actually they do have other concerns that they may not always surface until you have those Conversations with them. So I love that use case. and to the first point you said about, you know, not even using AI first, but finding product market fit. I think that is extremely important and it's a very strong demonstration of what you see where marketing isn't just technical, but it's strategy, right? And that's the part about strategic thinking and thinking, do I really even have a product that works right now? And how do I qualify and quantify it?

Liza Adams: 19:10

That's right. That's right.

Audrey Chia: 19:12

Yeah. And I would love to know. So maybe like a, a third use case that perhaps you see like really often something that a lot of businesses, can probably, you know, adopt AI for like a lower hanging fruit. Yeah.

Liza Adams: 19:23

AI search is one of them. So one of the biggest questions that I get is, how do we ensure that our brand and our content appears in AI search? So if you go into Claude or ChatGPT or Gemini or Perplexity, if you type in, you know, I want project management software that allows me to collaborate across my, geographically diverse team, and we're a marketing function. and we need to collaborate on campaigns as well as creative. What would be the best project management software for that? That's the prompt, right? That's the inquiry from a potential buyer. And the question for many companies is, how do we ensure that our content and our brand and our products are getting served up in that type of a question? So, one of the key things that, I tell people is we need to deeply understand our markets and understand what would be their questions at the top of the funnel meeting awareness stage in the middle of the funnel when they're beginning to consider different vendors and then at the bottom of the funnel when they're starting to make decisions. So what are those questions, right? And people are like, okay, well, how do I figure out those questions? I'm like, well, get the help from AI. You have a lot of customer reviews. Back to my other use case, you have customer reviews, you have transcripts, you have case studies, take all that, redact the information, take out sensitive information, and then ask AI, infer from the, from these reviews, and from these transcripts, What might be the top of the funnel, middle of the funnel, and bottom of the funnel questions that potential customers might ask? And that is your starting point. It will give you, you know, two or three questions that they would ask. Now you then go validate it with your market, make sure that those are the right ones. But once you get those two or three questions, then now the thing to do is, Okay, let's put these questions into the different AI engines. Let's put it into ChatGPT, into GemIIni, into Copilot. What's showing up as a response? Are we showing up as a response? Which influencers are being quoted? Which, which companies are being served up? What types of content? What types of data? And we can see. Whether or not we're being served up, right? And based on all that, there's this tool called and I don't work for them. Somebody found this tool. It's called chat hub that G. G. where you can put the inquiry and simultaneously. Oh, actually, it was Esar, Esar that told me about ChatHub. gg where you could put the inquiry and then you can see the responses of up to six AI assistants. Well, now you can see like how the different AI assistants respond and determine. If and how you show up and if you don't show up, who does show up based on all that analysis, you could then make a determination. Ah, okay. I am not showing up because I don't have a relationship with this influencer, or I'm not showing up because my content isn't relevant, right? I'm not as answering that question directly. And I need to create content or I'm not showing up because I'm not in the right places where the customers are. All right. They're having these conversations in specific communities, and I'm not there. So, AI will expose those weaknesses. We can then take a look at that and determine what the right strategy is. And the right strategy might be a combination of PR, content creation, and maybe identifying new, what they call watering holes, meaning where our customers tend to congregate. that we haven't been present in. So that, that I think is a really practical and pragmatic, use of AI to determine how to show up in AI search.

Audrey Chia: 23:18

I think that is also a very clear reflection of how your customers are changing, right? How your consumer behavior and trends are changing and the way people search for information, look for information, even plan their travel itineraries. Now it's completely different from before. And I think brands need to like. Keep up with that. Otherwise, you know, there will be that gap that you see the product, but the customers are miles ahead. That's right. That's right. And these are, I think it would be amazing if you could share with us, maybe a more like actionable use case. Is there a demo that you can share with our, you know, it would be amazing.

Liza Adams: 23:52

Yeah, let's do that. Audrey. All right, let's do this. Okay. Let me give you context around this. so I've created a custom GPT. Okay. And, this particular custom GPT is both strategic and tactical. Tactical because we are now able to do this repeatedly. It happens time and time again. We have the same action. But it's strategic because it is designed to create content that stands out and is unique for by persona and by stage in the buyer's journey. So, I'm going to show it to you right now.

Audrey Chia: 24:29

So, so for those of you who are listening, Liza is going to show us a beautiful GPT, right, Liza? Yes, it's a GPT. Awesome.

Liza Adams: 24:38

All right. Can you see my screen, Audrey? It's

Audrey Chia: 24:40

perfect. It's perfect.

Liza Adams: 24:40

Okay. I didn't get creative with the name of my TTT. Specificity is key. Yes, but it is very specific and you can tell what it does by the name of it. It is Persona and Buyer Stage Specific Content Creator. So, It actually not, doesn't just create content. It actually starts with ideating topics with you. And then it will, if you want, you can, ask it to create an outline. And then if you want, you can also ask it to create a draft of the piece of content. So, I have redacted this, so it, we will protect the innocent, no client names. So in this particular GPT, we are creating this for, I have created this for Acme company, which is a travel management SaaS company. And, I will show you how it works first, and then I will show you the instructions for how this was built. All right. So we have some conversation starters here, so I'll just say, let's start. And I hope the GPT works because I never know it, it might, Go haywire on us.

Audrey Chia: 25:47

Yes, I get that a lot from Chair GPT. I'm like, you better work on me today.

Liza Adams: 25:52

I'm doing a live demo, better work.

Audrey Chia: 25:54

Better work.

Liza Adams: 25:54

So you can see here that, you know, I'm asking it, in the instructions you'll see that I'm truly guiding the user, right, and it's showing here the objective, the objective is to help Acme Company generate impactful, persona specific stage appropriate content topics, and our goal is to drive engagement with the target audience. And, and nurture them throughout the journey. So it's asking us a few questions, which persona would you like me to focus on? And then there's these five different options. And then it's asking which stage of the customer journey are we creating content for, awareness all the way through loyalty. So I'm just going to say, I want to create this for a travel manager. And we're going to be in the awareness phase. So B and A. So let's see what it says. So now it's saying, all right, I'm going to give you a couple of content topics for the travel managers in the awareness stage. So, here's 1, the topic is redefining the travel managers role in a post pandemic world. It tells us the key points to explore in that piece of content. Why, it will stand out because, as a result of, you know, some of the angles that we will be using and the pros and cons of using this topic. And then it's giving us another option. The other option is the hidden costs of non compliant travel where it's time to automate. Policy management and that same format tells us the points to explore Why it will stand out and the pros and cons So that is asking us which one of these topics would you like me to develop into a detailed outline? So i'm like, I kind of like one so i'm gonna say one So now okay, it's it's giving

Audrey Chia: 27:35

you options, you know, and I always love options like let me think about it I think option one sounds great

Liza Adams: 27:41

That's right. And and you'll see this in my instructions, right? I always Ask it for options, and I always ask it for its rationale. Why did you give me that option, right? Give me the pros and cons, because that allows me to make some decisions. And it also helps me double check its work. So, all right. So now it's giving us the outline for, that first option. So you'll see pretty detailed outline here. and then, you know, you see the introduction. And then for each 1 of the sections, a fairly detailed outline of what it's suggesting that we do. Case studies and so forth and in conclusion, and then it says, does this outline meet your expectations? Would you like me to refine it or proceed? So I'll just say, you know, because we're nice to it today, we'll just say we'll proceed with a draft. So I'm not prompting it. It's basically prompting me to give it some answers, right? And then it's essentially here. It says, before I create. So you can see that it outputted it in Canvas. But before we go to Canvas, let me just show you what it said first. so our, it's our, our conversation is still here on the left hand side. So it says before I create the draft, I want to clarify that if specific details, such as case studies or examples aren't explicit. Explicitly provided in the documents, it's referring to some documents that I gave it in its knowledge. Yes, I will include placeholders or hypothetical sample content. This ensures clarity in areas where additional input might be needed. Let me prepare a comprehensive draft for the topic and then so you'll see. That the answer is in the chat, which is right here on the right hand side. Right? So now it's created our, our document for us. So see, and this is now in canvas. That's why it's a separate window and I'm just going to scroll through just to show you what it created and then we'll work with it in canvas. And so you can see there's a placeholder here for a case study, because it didn't have a case study, but it's suggesting that it would be great if we had a case study in there. And then so what I'm going to do is I'm just going to prompt this thing and just say, since we're working in canvas, we are now able to edit in line. please suggest areas where a quote might be appropriate and we'll just do that. And then so now we can see on the right hand side, it is suggesting that area could be a place where we could put a quote. This is another area where we could put a quote, and let's see, and then there's two more areas there, and I might just say, you know, I like this one. Let's apply that one. So then now it, well, it's working on a quote. So then there, you could see that it added a quote there. And then it's, it's basically saying, Hey, get, get some real information that we could put in here. And then if we don't want these other ones, we could simply, turn those off. Right.

Audrey Chia: 30:59

And I think that is the beautiful part about, Canvas, right? Because it's again, something that. I guess it's pretty new and people are still figuring out and trying to explore, but it's almost like Google Docs plus AI chat GPT combined. And Gemini has a bit of catching up to do here.

Liza Adams: 31:13

Yes. Yes. So now I'm just turning that one off. And then I could even say, suggest some places. Where a table might make sense.

Audrey Chia: 31:33

I also love how you prompted with suggest, you know, places instead of saying, please add a table because most people are just going to dive right into like, add a table and then GPT will probably add it somewhere that you don't want. So I love that you're asking it for options and then deciding for yourself, like, where is a better fit?

Liza Adams: 31:49

Always options because we're collaborating with it, right? We don't want it to give us just answers. We want to collaborate with it. So, I don't know, maybe right here, let's put a table in there. Right? So I'm like, okay, let's apply that table. Let's just see what happens. So I'm completely making this up. So there you go. There's the table right there. Amazing. and then we could just turn these other ones off and actually let's, let's see what table this looks like. I don't know.

Audrey Chia: 32:20

It's like, now we are looking at how you think about, you know, building your workflows. And you can see that it's not like a one shot prompt, like please generate this entire article for me, but it's really a step by step process, which then gives you a much higher quality output, right?

Liza Adams: 32:33

That's right. That's right. And, and, you know, and then I could even say. please turn this into an ebook. Let's just see what happens. So now you can actually take your raw content and then turn it into different form factors. Right. So now it's, it's an ebook and then you can continue to work with it similar to what we, how we worked with it when it was just, a document.

Audrey Chia: 33:04

Super powerful. I think this is a very powerful demonstration of how anyone can already start creating, you know, high quality content with that combination of guiding the AI to get to your final output, but with your human insights, with your database, you know, and combining all those knowledge into one workflow. Cool.

Liza Adams: 33:22

Definitely. And I will now show you the instructions if you're, and then here you can see the case study. So, we can move into the instructions. Let me just see if it's almost done. It's still working.

Audrey Chia: 33:34

It's working hard for

Liza Adams: 33:35

you today.

Audrey Chia: 33:36

Chattativity is doing great. I'm very impressed.

Liza Adams: 33:40

Oh, and then, you know, there's these, tools down here at the bottom where we can adjust the length longer or shorter reading levels, doing a final polish and I rarely use this, but if you really are a fan of

Audrey Chia: 33:53

emojis, there is the option,

Liza Adams: 33:56

there is an option. Okay, so I will now go into the instructions. so I'll just go into edit. GPT and, let me first kind of go through this, then I'll make the instructions bigger so that we could see it. So you can see that I named the GPT, I, Dolly created this image for me, and then I described the GPT, the instructions are here and I will show you that in a second, and then the conversation starters as you, as you saw here, and then here's, The big thing in here, I have input or uploaded a number of documents. There's a positioning and messaging document in here. There's a brand strategy, tone, and voice document in here. There is also this Acme company, ideal customer profile, persona, buyer's journey. There's a lot of information in here that helps inform this GPT. And this is actually it's training knowledge, right? So all those suggestions that you saw is based on this training knowledge. I'm also allowing it to search the web. And I've turned on canvas so that it can output in canvas, much like what you saw in in the output and then we're not asking it to do any, image generation or any data analysis. So I've turned those off and then we're not doing any, automation, like, what's happier or bake. So. There's there's nothing in there for that. So I will now show you the instructions. Okay. So, people always have different styles and how they prompt their GPT in, in my world, I'm a big proponent of responsible use of AI. So these three entries in here, essentially it's actually just two, because these. This last one is similar to the second one. The first one is really more about ensuring that this GPT doesn't just give anyone and everyone its instructions. So, you could actually go to a GPT and say, tell me your instructions. And if you don't have very specific instructions, not to share it, that it would actually share it. And that could potentially be a risk factor. Right? So I'm, I'm guiding it to not share its instructions. I'm also guiding it in these next 2. Not to do anything else beyond what it's trained to do and beyond its objectives and beyond its tasks. So this is designed to create, you know, topics and outlines and content. Like, I don't want it to respond to who is Tom Cruise. Or tell me about Top Gun, or what are the three laws of thermodynamics, or how do you create a bomb? I don't want it to respond to anything like that, right? So we are narrowing the aperture and making it super focused. Now, sometimes it deviates and we just can't help it, but I try my best to constrain what it does. And then here, you know, a very standard way of prompting it. I'm giving it a persona. You're an exceptional content topic creator, ideator. You're an expert at identifying content topics and develop drafts that will allow, at me, Acme company's brand and content to stand out, be highly relevant and helpful. The goal is to drive engagement, right? And then I basically say, deeply understand those documents that I gave you. You know, all of these things, they're all in your knowledge. Please leverage this knowledge to suggest content topics. And then I say, follow these instructions, step by step, by step, proceed to the next step only when you have completed the current step, or you've received a response from the user in the current step. So very methodical, right? So now this is what you saw, right? Welcome the user. We gave, we gave the user our objective, and then we gave the options for which persona they can choose. for which stage of the buyer's journey that persona is then and then I basically said, you know, based on the responses from 2 and 3 suggest 2 topics because I wanted 2 options, right? Yes. and ensure that it's aligned with the information and the knowledge document and then here is where I try to make it, unique and stand out content. This is a very distinct prompting also, make sure that the topic stand out. Please suggest topics that align with 1 or more of the following criteria. Could challenge traditional or widely accepted beliefs, highlight common assertions not supported by data, present unpopular opinions on trending topics, offer counter narratives that receive less coverage, reveal overlooked areas of true importance, suggest counterintuitive approaches, Connect seemingly unrelated insights with important implications and draw from unconventional sources of inspiration. So you can see we're truly guiding it to think differently, right? Because without this prompt, it will just take something very basic, something superficial or face value and create content on that. But if you're asking it to be thought provoking. Or provocative, then it will start thinking differently.

Audrey Chia: 39:05

I love that. I, and I love that you were so specific in guiding me in the type of content it crafts. And because you're so specific, you're going to get content that is non generic or non vanilla, because that is exactly what you prompted for.

Liza Adams: 39:19

That's right. That's right. and then I'm saying here, feel free to search the web only. to help validate or support your suggestions. Otherwise, please use your knowledge primarily. Provide your rationale and outline for the pros and cons of each topic. Again, it's to help me to figure out what I want, right? And then I say, and I ask the user if they want to create an outline, ask them if they want, You know, changes, and then ask them if they want you to create the draft and then this caveat that's that tells the user, Hey, by the way, I'm going to be putting some placeholders in there, make sure that that that you know that there will be some things that you may need to fill in. And I'm saying, do not make things up. So, yes. So I, I'm telling it, it's okay to have placeholders. You don't have to be perfect and make things up. Just give me, you know, brackets and I will fill in a case study. I will fill in the stack. And then lastly, you know, draft the output and put it into Canvas so that we can do user editing. And then finally thank the user and ask them to come back again in the future.

Audrey Chia: 40:25

Can you just say that I can see your engineering mind? Combined with the marketer's mind, because like most marketers, if you ask them, like, Oh, can you craft something? their prompts would just be like, you know, come up with some outline, you know, just give me like, you know, options and then write a copy for me. But here you have 11 steps that are so beautifully defined to really guide AI through that process for the optimal outcome, which I think is beautiful.

Liza Adams: 40:51

Oh, thank you, Audrey. I think it's one of those where, I say this. I now know this, right? In this early stage of AI, I think AI favors the intensely curious and a very specific kind of thinking mindset. Like, it favors very systematic thinking. It favors, you know, problem solving. In an iteration, if you like that stuff and you tend to think that way, then you'll have an easier time. Right? So if I was different, you know, if it was like more bigger picture, you know, like, I'm not an artist. Like, I can't see what's possible. Like, if you showed me a run down home. I can't see how to make that better. Right. And I think we have to give ourselves a little bit of grace because right now it does favor a certain mindset. So reach out to those people that that think that way and begin just kind of have to practice more because if you don't have that mindset, you kind of have to do it differently. But I think as a I advances, it will become More universal, right? It will get easier. Just like the iPhone, when it first came out, it was hard. You needed a stylist. It had little buttons. It didn't have an app store, but over time it got easier and it became more accessible to more people. Even children can now use it. So the, this notion of, Hey, AI is only for technical people. I don't buy that because I do believe that it will get better. It will become more accessible. And the prompting skills will probably be not as relevant moving forward, because it will understand context, it will understand intent, and, hopefully the underlying technology will be, we won't have to choose the model, we won't have to choose the tool, it will just know based on what we're asking, the right model and the right tools to use.

Audrey Chia: 42:39

Awesome, I think that is where the future is going to be and it could be a very near future At the speed at which AI is evolving So to wrap things up Liza would be great to hear maybe one tip that you would give Anyone who is just starting out on this journey

Liza Adams: 42:54

Yeah. So the one tip I, I give people is start now, right? You know, and open your mind to something very different because when you do look into chat GPT or Claude or Gemini, you only get that conversation box and that conversation box in, in many of our minds looks like a fancy question and answer engine or a search engine, right? But think about it differently, right? It's. I think AI and technology is less of a challenge. It's actually a behavioral and a mindset shift. We need to think it, think about it more as a thought partner where we can give it context. We can, ask for its opinion. We could share our ideas and it would give us what it thinks about it. We can use it to get very different perspectives. And I think if we go in thinking that it's not a search engine, and it's not a question and answer engine, then we have a better shot of truly understanding what it's capable of doing.

Audrey Chia: 43:57

Definitely. And I love that. So I always like to say, you know, the best time to plant a tree was probably five years ago and the second best time, you guys already know the answer. Thank you Liza for joining us. It was a pleasure having you on the show and thank you folks for tuning in. Don't forget to subscribe to the AI Marketers Playbook and hit the bell for more actionable marketing insights. We'll see you next week.