Introduction and Guest Welcome
00:00:00
Speaker
We're back! And I'm so glad you've joined me for another episode of More Than You Can Chew, and I hope that you've remained hungry for new insights. Enough of that cheese. Let's get to the meat of today's episode, which we couldn't cover without our incredible guest, Melissa Popp.
00:00:20
Speaker
Content Strategy Director at Rickety Roo. And what I like most
Melissa's Content Strategy Approach
00:00:24
Speaker
about Melissa's approach is that she loves not just storytelling, but storytelling that has strategic thinking behind it, coupling creativity with data.
00:00:33
Speaker
Now, that's the right way to craft a powerful narrative that really captures, engages, and converts consumers. If you've never heard or seen Melissa talk before, so many of the brands that she partners with are in competitive markets.
00:00:46
Speaker
Her sore focus is on the really important facets of content, authority, authenticity, and relevance. And she'll let you know, and rightly so, that this approach will help businesses gain visibility, build trust, and achieve results.
00:01:02
Speaker
That's because she doesn't look at content just as a thing to do, a thing to publish. For Melissa, it's about connecting with people, inspiring action, making an enduring impact. So Melissa, now the folks can highly finally hear your voice. Welcome
The Role of AI in Content Writing
00:01:14
Speaker
to More Than You Can Chew.
00:01:16
Speaker
Thanks for having me. I'm always excited to talk about content and how we can move beyond just the traditional marketing aspects to start telling more stories that resonate. I think that's more important than ever in this crazy AI now driven world we're in.
00:01:30
Speaker
Absolutely. Absolutely. So, you know, now your whole but background, your philosophy, your approach, you makes you the perfect guest to cover what I think is an important topic, and that is AI in content writing.
00:01:43
Speaker
And know there are agencies out there that either tout they're using AI to fully write content from ideation to publication to create content at scale. which I personally believe is a flawed perspective, or they're secretly doing so and not telling their clients. And what I think what worries a lot of marketers, writers, consultants, even agencies is really how to properly use AI, use these tools.
00:02:05
Speaker
So that's what I want the big bite that's more than we can chew to be. And there's sort of two parts that create one whole takeaway. So Melissa, could you, even if it's a 10,000 of you, get in the weeds as much as you want to, you know, kind of walk us through how your process of how you use these tools um help expedite that rote or grunt work. And then how do you know when it's time that you're at a point where you can now inject your own creativity and thoughts into the process so the content has some sort of shape, is something refined, is something that has some soul that truly connects that content with consumers?
00:02:42
Speaker
think the most important
Melissa's AI Journey at Rickety Roo
00:02:43
Speaker
thing is, is that we do have to recognize as an industry and as businesses in marketing, that AI is going to replace some of the work we do.
00:02:52
Speaker
And I think it's kind of like you have to go through like the five stages of grief here, where you know you're looking at acceptance of, okay, This isn't going anywhere. Platforms like ChatGPT, Claude, Perplexity, and others, you know, are continuously evolving, are continuously integrating themselves into everything we do from, you know, interacting now with our Gmail accounts to interacting with our project management systems. MCP has opened up a whole world of ways we can connect with different apps.
00:03:19
Speaker
So this is happening whether we like it or not. And trust me, like my my English loving and soul, my writing loving soul, you know, hated this at first. And anybody at Ricky Root will tell you how skeptical and just, no, we're not going to do this. We're not going to integrate this in.
00:03:35
Speaker
And then I woke up one day and i said, you know what? I have to figure this out. And starting to really learn how LLMs work, how these different wrappers work, how people are presenting what they're doing with AI, I saw a real gap between, okay, let's just run with whatever we can push out and publish at scale.
00:03:55
Speaker
And a lot less of how does that role shift to where the human element comes into play. And that's where I
AI in Content Creation Workflow
00:04:03
Speaker
started focusing. And at Rickety Root, we've built AI workflows that I have built from the ground up. This is not me looking at what other people are doing and manipulating what their custom instructions and prompts are. This is me looking at how do we build something that can take away some of the grunt work that we have from the beginning of the writing process.
00:04:24
Speaker
You know, there's a lot of people are great writers, but they're not great at ideating the ideas, right? This is a great way to use AI to research, especially now that everybody has a deep research model. So we can get even more into the weeds with the topic ideas that we come up with.
00:04:39
Speaker
And then from there, it's a matter of like, okay, what takes more time now? Okay, we move on to content research and and and brief outline, right? And i don't I don't know how many people out there have seen content briefs that I have created.
00:04:53
Speaker
um but i am like, as a person who's been writing since she was four years old and went to school for journalism and has been writing since I was a kid and have this deep appreciation and love for it.
00:05:04
Speaker
I create probably overkill briefs to hand off either whether I'm writing it or I'm handing it off to writer. So that takes a lot of time. And, you know, that's on top of the research you're doing to support that brief. So the next place I started was thinking, okay, what can we take out of the research and the outlining process that we could take, let's say, 50% to 60% of the time I spend on that and take that other 40% of the time after AI helps that heavy lifting,
00:05:32
Speaker
That's where the human eyes and the human critical thinking and the creativity comes into play. So, you know, that naturally progressed to, okay, I have great ideas.
00:05:42
Speaker
have great um outline. I'm feeling good about this brief. I'm feeling good where it is. What can I now push AI to do to create a fundamental foundational draft that a human then can run with, fact check, copy edit, integrate more of brand voice in?
Experimenting with AI Tools
00:06:01
Speaker
this is where all of the time I've spent over the last 18 months of learning how LLMs work and how prompt and custom instructions work, that I now have built a process for Rick Dewey where we can go from ideation through research to outlining to a foundational first draft.
00:06:16
Speaker
cutting more than 70% of our time of the grunt work that gets us to that place so that the human eyes can now take that and run with it and and just tweak and copy edit and play and actually get more creative than we actually normally would if we were doing this from start to finish with a time or budget constraint.
00:06:38
Speaker
And so really for me, like the whole approach, and this is what I tell everybody who's playing with AI. First off, stop reading everything out there. Get your own hands dirty. Figure out what's important to you and build that into what you're trying to build.
00:06:52
Speaker
And nothing I'm building is proprietary or something anybody out there can do with honestly any skill level. This isn't even, I'd love to see my 20 years of marketing helped me get and build this.
00:07:05
Speaker
No, just the passion for it yet. Like the the actual skillset and everything, it's, I think that's one of the greatest things about AI right now in our industry is that anybody can jump in and learn and figure out how to take their own soul and put it into what they're building.
00:07:18
Speaker
And I think that's gonna be the difference that we see in the coming six, nine, 12 months as AI, you know, Google just launched AI mode and we're seeing LLMs showing up in GA4 and other analytics tools that people are using.
00:07:31
Speaker
So how do we connect with these and overcome that robotic presence with the content we're doing. And to me, it's like, you've got to put your soul into you've got to put your passion into it.
00:07:41
Speaker
And the more we figure out how to do this with these tools, the not only the better the end product is, but the better we're all going to feel about AI's taking our jobs, right? This is how we save our jobs. This is how we save our industry.
00:07:55
Speaker
How do we take our passion and put it into what we're building and then ultimately how we're serving our clients? Fantastic. I love that. I love the the pack passion you have in talking about the the process and the creation of content.
00:08:08
Speaker
Do you have, you know, I don't want you to like, you know, you know put the, you know give away the the the whole Apple card or I don't know some sort of phrase that would capture what I'm trying to ask you, any analogy, but, you know, do you have any tips to help people understand how to properly craft prompts at each of those stages? Like anything that you would say, like,
00:08:26
Speaker
here Here are things we learned that were mistakes to avoid. And here are things that you should try to include. Even if you can speak to it more generally instead of, like said, you don't have to give away all of the sauce. But, you know, are there just some things you would say, look, if you're going to write a prompt for IDing, if you're going to write a prompt for for research, if you're going to write a prompt for any sort of like AI-assisted editing, like what are some what are some tips you might have for someone ah to effectively do that?
00:08:50
Speaker
Well, the first thing is that I hate about this whole concept of prompt engineering guides that people are putting out there. The thing is, is LLMs have a limited context window, which means that we could only give them so much input and then based on our output, continuing to feed that.
00:09:06
Speaker
And this is going to vary. I'm not going to sit here and list what the context windows are for everything. All these tools are different. Nobody's doing anything the same. um This is why you have people like me who I'm a big proponent of perplexity and chat GPT.
00:09:18
Speaker
And my boss, Blake Denman, he lives and breathes Claude. Everybody has their preference. And not all of these tools take prompts and custom instructions the same way. So this is where I say you got to get your hands dirty.
00:09:31
Speaker
My biggest tip for prompt engineering is to stop thinking about marketing and start playing with your AI tools, anything but marketing. This is one of the ways that I learned how to get AI to give me the answers I wanted, how to limit how much I had to actually give feedback and criticism and rework things.
00:09:51
Speaker
So when I first started learning the ins and outs in ChatGPT, I followed Marie Haynes' example, and she was using it to up her Fortnite game. And so she would she would lose a match of Fortnite and she's talked about this all over social media and then talk to ChatGPT about what happened.
00:10:08
Speaker
And so I started doing that because I also play Fortnite. It's my great escape and i I've been playing it since the beginning and I'm that person people laugh at because I suck at it, but I have so much fun.
00:10:18
Speaker
So that's how I first I was like, oh, this is interesting what she's doing. Let me let me see what happens. And it didn't improve my Fortnite game because I really had. But then that led to me starting to play around with it with meal prep.
00:10:32
Speaker
And this was back when chat GPT and Claude, the models were as sophisticated as they are now. So there's great improvements if I go in and try to do this now. But seeing how I could ask it things outside of something like, oh, you know, take these five yeah URLs and tell me what all these yeah URLs are doing the best so I can implement that at a content brief.
00:10:52
Speaker
That's pretty straightforward. But trying to get it to actually help me with other things outside of marketing helped me understand prompt engineering. And so that's one of the things i would really encourage people. And I encourage people on the Ricky routine to do all the time is like, stop thinking about it as just a marketing tool.
00:11:08
Speaker
What else could it help you with that you feel comfortable talking to it about and sharing with and learn how it interacts with you? Because a lot of the things um people don't understand about these LLMs outside of our industry is that they're designed to give us the answers we want.
00:11:23
Speaker
So the more context and information we can give it as simply as possible helps us get to that without it filling in gaps we don't want filled in. And so I really encourage anybody listening, like, start playing around. Like, what else are you doing in your life? You're exercising, you're trying to up your golf game, you know, you're trying to figure out... um you know What new movies do you want to go watch? Play around with it that way and learn how you're asking it questions in in a more just fun environment versus what we do sitting on our computers every day in marketing.
00:11:55
Speaker
I never thought of that and left using the tools that way. And that to me reminds me of back in the day we'd talk to businesses like, go on your phone, look on it look at your website, engage with it like a consumer, like yeah take your marketing hat off And engage with your site, your assets, your content.
00:12:14
Speaker
Now engage with these tools like ChatGPT, like a consumer would to see that how that dynamic works. That's a really good tip I never even considered to do. And i to suggest I've done it and I never thought but like this should be a suggestion for people to do.
Personal Insights and Farewell
00:12:31
Speaker
gives you a different way of looking at it. Like, of course, we're all using these tools for work, whether we're so telling clients or telling others in our industry or not, you know. And so it's like, what else, though, could you use this for? Like, I have a whole space in perplexity that I use for deep research that is literally called explain like I'm a five-year-old and I have built custom instructions.
00:12:52
Speaker
And like, If you go through that, it's like unhinged looking at the stuff I'm looking at, you know, and and and some of it's industry related, some of it's work I'm doing in therapy and like figuring out other ways and other insights from therapy sessions.
00:13:07
Speaker
You know, it's like you can use these tools in so many different ways. And we're limiting ourselves by just forcing ourselves into the marketing box when the things we're doing outside with these tools will actually help us be better prompt engineers and better marketers.
00:13:21
Speaker
Fantastic. Fantastic tip, guys. That's the big bite that you can take away to chew on for bit is start engaging with these tools like consumers and learn how how those interactions happen. Look at those outputs you're getting. See what consumers are going to see and and leverage those insights into how you can start building out prompts to get the specificity and and the the details you need in order to limit those gaps that the LLMs will try to fill. So,
00:13:48
Speaker
Melissa, thank you so much for those insights, the walkthrough of your process. And of course, the big tip I didn't think of is use it like the consumer. So thank you again for your time today. But before I let you go, i have one last question to ask you. And it's going to be the toughest question you've ever been asked on a podcast ever, I promise.
00:14:06
Speaker
So whether it's in Denver or anywhere in your travels, if you told someone, look, if you're in this city, this is a place you have to go to get something to eat.
00:14:17
Speaker
What would that place be? All right. For those of you who know me, i grew up in San Diego. um My heart belongs to San Diego. And if you're ever there and you're in the La Jolla region, you want to go to Point Loma Seafood.
00:14:31
Speaker
um They catch fish right at the restaurant. It will be the freshest, greatest seafood you've ever had in your life. um I will tell you that the the locals trip trick though You got to get there early because this place is small. It's busy. It's popular.
00:14:47
Speaker
Don't park at the lot right next to the restaurant. Park blocks away and walk. Trust me. So you'll solve a headache because you won't be able to find parking. But trust me, this is going be one of the greatest seafood experiences of your whole life. I used to go there as a child with my parents all the time.
00:15:02
Speaker
And um I try to make time every time I'm in San Diego back there, um especially now that Brighton is there. I've taken some time to go in there. Nice. And what was the name of that again? Point Loma Seafood.
00:15:13
Speaker
All right. Awesome. Well, I hope I get to go to San Diego sometime soon because I'm definitely going to check that place out. Anywhere where the fish is caught right at the restaurant is definitely going to be probably the freshest fish you're ever going have. So awesome points.
00:15:25
Speaker
um Melissa, again, thank you so much for your time here today. i really am grateful that ah you came here to talk a little bit about AI in the writing process. I'm glad to be here. I love talking about this. So anybody listening, feel free to hit me up.
00:15:38
Speaker
Tell me the weird stuff you're searching and using AI for. I'd love to hear about it. Great. All right. Awesome. Well, everyone, I'll catch you on the next More Than You Can Chew.