Your Digital Marketing Coach with Neal Schaffer

From Digital First to AI First

• Neal Schaffer • Episode 402

Share episode

This episode explores the imperative shift from a digital-first mindset to an AI-first approach in marketing, emphasizing how to incorporate AI into internal processes for enhanced efficiency and creativity. Throughout, I discuss strategic alignment, process documentation, custom AI models, training with quality data, and automation for maximum impact. 

• Discussing the transition to an AI-first mindset 
• Importance of aligning AI with marketing objectives 
• Utilizing AI as a second opinion for marketing strategies 
• Documenting processes for optimized marketing activities 
• Creating custom GPTs tailored to brand needs 
• Training AI with quality data for better outputs 
• Scaling efficiencies through automation for marketing tasks

Learn More:

Speaker 1:

Today's episode will not be just another conversation about AI. It is a call to embrace a whole new mindset. Are you still scratching the surface of AI or ready to go all in? Today, I'm inviting you to shift from a digital-first approach to an AI-first approach. We'll talk big-picture strategies and day-to-day tasks that you can transform with AI's help. This might just be the most pivotal leap you'll ever take in your marketing. So stay tuned, because you will not want to miss this next episode of the your Digital Marketing Coach podcast.

Speaker 2:

Digital social media content, influencer marketing, blogging, podcasting, vlogging, tiktoking, linkedin, twitter, facebook, instagram, youtube, seo, sem, ppc, email marketing there's a lot to cover. Whether you're a marketing professional, entrepreneur or business owner, you need someone you can rely on for expert advice. Good thing you've got Neil on your side, because Neil Schaefer is your digital marketing coach. Helping you grow your business with digital-first marketing, one episode at a time. This is your digital marketing coach, and this is Neil Schaefer.

Speaker 1:

Hey everybody, this is Neil Schafer, your digital marketing coach, and welcome to my podcast, where I teach, hopefully, inspire and empower marketers, small business owners, entrepreneurs, startups, content creators, service providers, entrepreneurs. You know who you are. I just want to make sure that you know that you are in the right place, so thanks for joining me today. Now this episode. Well, let's put it this way you know every one of these solo episodes that I record. The whole idea is that I want to keep you in touch with the work that I've been doing, what I've been thinking, the books I've been writing. Really, the latest and greatest of Neil, of your digital marketing coach, is in the podcast. I release content here before I do anywhere else and when I think about what to talk about today, man, it's just about digging deeper into AI. We all know that AI is one of these central topics that we talk about, but the more I talk to the people around me and the clients, the more I realize that there are still a lot more that can be done, and today's episode is really to hopefully inspire and empower you to do more with AI. Now, it's not the first time I've talked about this. It was back in episode number 364. This wasn't too long ago, right? This episode, by the way, is 402. This is back in May 21st. So, okay, it's been nine months and in the world of AI, that's actually a pretty long time. But back then I talked about making the AI shift and it was really moving behind the hype and really integrating AI in how we work. Well, today, consider this a revision of that, and you could also consider it an updated chapter for my book, digital Threads. In Digital Threads, in part six scale. I did have a dedicated chapter on AI and marketing and really what I talked about there was pretty much generative AI and content creation. This goes well beyond that. So think of it as an updated chapter. We're taking that concept of digital first and sort of turning it on its head, exploring how to make AI the center of all of our internal marketing processes. Digital first is still critical for how we market externally, if that makes sense. So it really is fascinating how quickly AI is evolving and how much potential is still on top for many businesses, as I mentioned.

Speaker 1:

So let's start with the big idea that I hinted at. What does it mean to go from digital first to AI first? You might be thinking, neil, I've already got my social media channels. I'm pumping out content on a regular basis, I'm engaging, I'm running ads, I'm on LinkedIn, I'm blogging. Isn't that enough? Well, yes and no. Digital first was all about making sure our external marketing how we show up online was optimized for the digital platforms where our audiences live. And, as you know from those that have read my book, it's about the SES framework, or search, email, social, and making sure you have all of your bases covered. But behind the scenes, you have an internal marketing engine room, and that's where it's time to become AI first. Every piece of content you create, every campaign you plan, every piece of data that you analyze, every strategy that you want to revise, ai can turbocharge all of it. So we still keep that digital first presence on the outside, as I mentioned, but on the inside, we want to be AI, to be front and center, driving the strategies and processes that rely on, and it really begins with aligning AI with our objectives, strategies and tactics.

Speaker 1:

So if I was going to write a book on AI first marketing, I would begin with this chapter, so before we even talk, or this step. So before we even talk about automation or content generation, which I'm going to get to. It's crucial to circle back to the basics. Why do you do what you do? What are your objectives? Brand awareness, lead generation, sales growth. Who are your target personas? Have you developed them fully? Have you confirmed that they are the right personas? Maybe AI has other ideas.

Speaker 1:

Right, if you feed all of this context into your AI tools, you can use AI as your very own marketing consultant. I like to think of it as meta consulting, and I don't mean Facebook, meta AI advising a marketing consultant like me on fresh angles and new perspectives. It might sound wild, but, trust me, it is incredibly powerful. And the key is after you put in all that information. You ask ChatGPT or whatever tool you're using is there any other question you have for me or any other additional information that I could provide you, to allow you to provide me the best answer possible, or something along those lines, and that will give you even more advice on what additional information you want to train the AI with and that's another topic I'm going to get to in a second. But whether you're brainstorming a campaign or analyzing audience data, ai can just give you insights you might never have thought of on your own.

Speaker 1:

It's really the same concept of user-generated content, right, that your customers will take pictures and or write a review or whatever it might be, snap a video and often the way that they portray your product or service or talk about it in a review, it's just very different than the way that you talk about it, and sometimes you're going to learn a heck of a lot. You're going to be able to leverage that for your own content creation, right? It's the same thing with AI. It's going to give you a different opinion. Now, honestly, I love how it gives the second opinion. Sometimes it's spot on, other times it sparks a new direction.

Speaker 1:

So, if you're feeling stuck, start by telling ChatGPT here's my audience. Here's what I'm trying to achieve. What do you suggest? You might be surprised at how specific and creative AI can get, and I actually did this today with a I won't say the name of the person, but someone from my digital first group coaching community. We had our one-on-one call and this was the advice I gave them Like this is the industry they want to be in. These are the marketing activities that they've been doing. What haven't they been doing that they should be doing? How can they improve what they're?

Speaker 1:

already doing and do you have any other advice? And the more we fed it, the better and better and better the answers became, and now my member has this whole to-do list, but they feel really empowered by what, in this case, chatgpt provided them. So that's step number one. Right, aligning AI with your objective, strategies and tactics and really seeing if there's any optimization or second opinions or development of those ideas that you can do to improve your work. Right, improve your strategies. So the next thing is we want to take it down a notch. We want to go from strategy now to tactics. So step number two is going to be reviewing your marketing activities and documenting processes and documenting them so that they can be reviewed. So one of the keys to leveraging AI in your marketing is having your core activities spelled out in clear, repeatable processes. Look at every major task content creation, scheduling, community management, analytics. Ask yourself, do I have this documented somewhere? The reason is simple Once you have these steps written out, it's way easier to see where AI might step in to handle a portion or all of this workload. It also gives you the ability to train and teach the AI on what exactly they should do to replace you or compliment you in completing that task. So this is where I do a lot of experimenting with ChatGPT and other tools. Can AI draft an email sequence for me? I don't know, let's give it a try. Can AI set up a social content calendar? I don't know, let's give it a try. These are tasks that I have to do. Why not have an AI first mentality and see how far I can get with AI? So if it can help me, fantastic, and if not, that's cool too. But you won't know unless you map it out and try. And this is one of the hallmarks of having an AI first mindset is trying everything AI first, even if it's things you've already been doing, even if it's things that you might not think take up a lot of time. Ai might not just do it faster but also better. Really depends on the task. But this is really where the rubber begins to meet the road, in my opinion. All right.

Speaker 1:

Step number three creating custom GPTs or custom AI models. So, after you've documented your processes, you've experimented a little with ChatGPT and you've realized that for certain tasks, you know it might make your life easier or you might get better outputs by leveraging ChachiBT in part or in full, and I mentioned you know in part as well you don't have to do everything in AI. Ai first means you go to AI first to do it, and I'll give you a very, very good example of AI assisted content creation, where it is partially using AI, not fully automated, but it still provides great value and saves me a lot of time and it's really easy right. I have gotten into the to introduce the concept of custom GPTs. I have gotten into basically going 360 degrees, full circle. So I started with AI for those that remember with the tool called Jasper, and I even had a podcast episode about going to the Gen AI conference. This was back February 14th I'll never forget, because it was Valentine's Day, 2023.

Speaker 1:

Just literally, you know, less than three months after ChatGPT came out and there was a vibe, there was a buzz in that room, let me tell you and I was using the different recipes that Jasper had. So, hey, you want to create a Twitter post, use this recipe. If you want to create a Facebook post, use this recipe. And they were all very, very specialized recipes or templates that allowed me to easily do certain tasks right. This is very much generative AI. So what's interesting and I talked about this before on this podcast, for those that remember. But Jasper then changed their user interface. They went more enterprise. So, instead of having all these recipes, you just had an open chat, just like ChatGPT, and you had these open conversations and it would develop things for you. So that got me thinking, and it was actually a webinar that Jasper did where they described the prompts that they put into Jasper to get output, and I ended up just using the same prompts in ChatGPT and I was able to get the same output. And that's where I stopped using a lot of third party tools and I started really doubling down and tripling down on ChatGPT, which is still my advice for you today, for things where you can use it.

Speaker 1:

Now I want to give you an example of this, partially using AI, and I'm trying to replace basically all these tools I used before. I've already replaced Jasper. In all honesty, jasper is a great tool don't get me wrong and if you're an enterprise that has certain enterprise features that are great for enterprises, just they're not that appropriate for me and the size of my business right, and perhaps the size of your business. So one AI tool that I've used before I no longer use because it's also been replaced by ChachiBT is a tool called CastMagic. Now, castmagic is also a great tool. Nothing wrong with the tool, right, it is similar to Jasper. It has these pre-made recipes so that you can plug something in and get all sorts of different types of output.

Speaker 1:

Now I found out about CastMagic because a lot of people at PodFest and this is another episode you can listen to. This was PodFest 2024. I probably recorded it back in January, february of last year, but this was a tool that was on the tongue of many of the speakers Just feed your podcast or, you know, live stream into CastMagic and it'll create all these assets for you. And it did. And I used it for a while and then I realized you know what? I could just do it on my own. I can just create a custom GPT that mimics what I was doing in CastMagic and I can fine do it on my own. I can just create a custom GPT that mimics what I was doing in Cast Magic and I can fine tune it for my own voice, my own tone, customize the prompts and at any point, I can add or delete from that, so I can customize what I do. So this is where I will take a transcript, which I use otteria for, and I will feed it into the custom GPT that I created and it will create all these different assets that I need for this podcast. So another one, and that's an example where the AI is not like going into Buzzsprout, you know, and actually filling everything. It is not completely automated. I'm going to talk a little bit about that at the end, but, on a similar note, this is something where ChatGPT could provide me all the outputs.

Speaker 1:

Another tool that I've recently replaced although I'm still subscribed to because it still does a value is a tool called Opus Clip or Opus Pro, and this is a tool that, once again. I had VidSummit last year and I have another podcast episode about visiting there. It was what everybody was talking about. They were sponsored, they were speaking on stage, they were doing workshops, and it is a fantastic tool for taking a live stream video, like the interviews that I do for this podcast, plugging it in and it'll give you 30 different based on the filters. You give it 30 different, basically short form videos where you customize the captions, you check the captions, you can reframe, you can add, you know AI B-roll and you can create, you know, a short form video that you can publish in social media on all these sites very, very easily.

Speaker 1:

But what I realized was it didn't give me all the editing features that I wanted. It wouldn't allow me to bring in a logo of the your Digital Marketing Coach podcast or a logo of my book or perhaps a sound effect or just different Marketing Coach podcast or a logo of my book or perhaps a sound effect or just different things. That it felt a little bit limited. Also, when I was given these 30 different episodes or 30 different snippets to go through, I wondered how it decided on what they thought would be the most engaging, because I often thought that most of the time it wasn't the best of the content. So I experimented with both Otterai and then ChatGPT, because Otterai also has a chat feature. So I basically now take the transcripts from Otterai, put it in ChatGPT.

Speaker 1:

This is a different custom GPT because for this custom GPT I wanted to go through and look for snippets between 30 to 60 seconds that have an educational point, where something is explained so that the listener can learn something concrete. And I've realized that it's not about a quote unquote engaging snippet. It's more about that. I'm talking about a specific topic, that I can create a video around, that someone can watch it from beginning to end and learn something, just as if I was going to record a short form video myself and I realized by doing this I'm not asking for 30, because even with an hour interview I rarely find 30 snippets I want to use I'm asking for, you know, 10, and of those 10, maybe I only use three or five or seven. The most I've ever used is eight, but it helps me better, you know, be in direct contact with my own video, with my own interview, with my own podcast, whereas if I'm using the tool like Opus Pro, I feel a little bit distant because it's already coming up with this stuff for me With ChatGPT.

Speaker 1:

It's coming up with 10 ideas that now I'm going in and listening again and making sure that's what I want and then I outsource it. I have someone on Fiverr who is in South Asia who does it very inexpensively and I basically give them the timestamps, the title, the transcript, and to me that is a great way of having AI partially or AI-assisted content creation right. The AI is doing the heavy lifting and then obviously hiring someone to do the rest of the work for me instead of using what everybody else is using the same tool, the same sort of captions, and I know you can customize it but this way I have a completely customized output right, and going through the process with ChachiBT, I think the quality is going to be much better. I mean, you tell me if you were to go into my TikTok, my Instagram Reels, my YouTube Shorts, if you look at the short-form videos that I've been publishing over the last two or three weeks, versus what I published a few months ago, there's just a big difference. It's like night and day, right. So this is an example of a task that I was using an AI tool for that I decided to bring in the ChachiBT to increase the quality, which I believe it has, and it is only assisting me but because I am able to hire someone very inexpensively to finish creating the videos, and it's not something I have to. You know, even within Opus Pro, there's still a lot of little things you need to do to customize and edit what have you. This is just a much better use of my time and, I think, a much higher quality product at the end.

Speaker 1:

So, everything I talked about here and I'm sorry for digressing, but everything I talked about here is about custom GPTs. It's about once you document the process and you experiment in a conversation with a large language model, the next step is you want to codify that. So think of these specialized AIs such as custom GPTs. I think in Cloud they're called projects. They're specialized AIs that have been trained specifically on your brand voice, your messaging and your best practice. They are the recipes that Jasper used to have best practice. They are the recipes that Jasper used to have, right, the templates that CastMagic has, but you are developing them and fully fleshing them out and personalizing them, and you can have one GPT that focuses on, say, blog outlines and SEO optimization, another that's specialized in email marketing copy. Whatever tasks you do most frequently, at least my approach right now, is to have very, very specific custom GPTs for each one of those things.

Speaker 1:

Now, some folks prefer bundling multiple processes into a single custom GPT. Right, like you have an email marketing expert. And some people name their custom GPTs with, like human names, like I'm going to ask Mary what she thinks about this email sequence. Right, so you could do it by function like that. I just prefer to do it by task, right, so others like to keep them separate. Like me, it just depends on how you define the work and all the different tasks that you have, but the bottom line is that these custom GPTs will get smarter and more accurate the more you train them with your past successes, feed them your best content, your highest performing campaigns, your brand guidelines, even your style quirks, and watch the magic happen.

Speaker 1:

So that was a long section, as you can imagine, but now we move on to what I would say step four, or that was a long step. So step four and I sort of hinted at this at the end of step three step four is training your AI with quality data right. The custom GPTs mean nothing unless you upload the right training data, you give it context, you give it examples, then it could create something that might just be better output than you can get. So this can be a little painful at first. I'm not going to lie, digging up old documents, hunting for top performing ads or emails or pulling metrics from your successful campaigns, but trust me, it is worth it. It is you do it. Once you're done Like my LinkedIn profile analogy right Once and done, maybe go back and look at it every 90 days or, in the case of custom GPTs, when you see the output, you might want to edit them slightly.

Speaker 1:

You know fairly frequently, but you're not going to have to go through the process of giving it all that initial training data again. So, as I hinted at, ai works best when it has something solid to emulate. If your brand voice is casual and friendly, or maybe more formal and B2B, give your AI examples of that exact tone so that it can replicate it right. Look for, for instance, if it's going to be social media posts, find your top performing posts. If it's emails, find your top performing emails right. It's not just nilly willy upload any and everything. It's finding your top performers, doing a little bit of analysis right, finding the things that best represent your brand, your tone, your voice, and applying analytics to it to say, okay, well, these were the top 10 performing subject lines over the past two years of sending out weekly newsletters. I'm pretty sure this is the formula I want to stick with. And then uploading those and then allowing ChatGPT or whatever LLMUs to give you options right. Then you're going to get a very, very good quality output.

Speaker 1:

So a lot of people said, hey, think of it like training your intern. Think of it like training a new employee right, you want them to learn from your best work, not from random stuff lying around. And then we get to step five. I'm only going to go five steps today and I think over time this will be divided into further steps, but I wanted to provide somewhat of a framework today so that you can work off of it and you don't have to do all the steps immediately, because this last step, in all honesty, is something I've yet to do. I know about it. I know enough about it to talk about it. I have not gone there yet, but I hope to go there in the not so distant future. So step number five is scaling efficiencies through automation. So let's talk about automations, because this is where AI first really starts to shine.

Speaker 1:

If you're familiar with tools like Zapier or makecom I think most marketers already have subscriptions to one of them, if not both of them you know that you can link different apps and tasks together. Now imagine chaining multiple GPTs or other AI tools together with other tools in your marketing stack to create entire automated workflows. So it's the concept of automating workflows with a zapiermakecom is not new. What's new is including AI as part of that solution, including your custom GPTs, in everything you do when it comes to automation. So, for instance, maybe a new lead comes in via your website form, immediately, an AI takes that data, personalizes a welcome email, shoots it off automatically. Or maybe record a quick voice note about a new product launch and it goes through a transcription AI, like Otter, which then feeds into a GPT that drafts a blog post, which then gets auto-uploaded to your CMS and then your editor can get notified, and to your CMS, and then your editor can get notified and proofread it, and then you're done.

Speaker 1:

So this isn't just saving time, it's really accelerating your marketing. Now, I'll be honest, some of these setups can get quite intricate, but even small wins add up quickly If you can offload just one repetitive task and automate the AI that is attached with it. That's time you can spend on higher level strategy or nurturing client relationships or developing relationships with social media users or influencers or content creators. Right, it's all about freeing up this time, and I should have mentioned earlier on in the podcast I had this aha moment a few weeks ago, where it's like man, I have a lot of free time today, and it was the first time I'd felt it in a while, and it's because I think I'm just using this AI first approach and finding ways to not only make my work better but also save me time, and I hope that you, too, will get there with whatever tasks that you have.

Speaker 1:

So here's my question for you how AI first is your marketing team right now? How AI first is your business right now? Are you all in, maybe already chaining together multiple GPTs and calling them and automating everything? Or are you still on the fence, unsure if you're really getting the most out of AI, or maybe you're scared of AI privacy issues? Is it going to replace you? Whatever it might be?

Speaker 1:

Wherever you are on your AI journey, I'd love to hear about it. I was very upfront, saying, hey, I'm not at the automation stage yet, I'd like to get there. I am, you know, paving a path to get there, but a path to get there. But send me an email, neil, at neilschafercom, or drop a comment. When I do publish this podcast, it goes to podcastneilschafercom. You can comment there, or just tag me on social right or just DM me on social. I think you know where I'm at. I'd love to hear where you're at and I'll do my best to point you to resources or strategies that I think might be able to help you.

Speaker 1:

So that wraps it up for today's episode on moving from digital first to AI first. If you found value in this episode, go ahead and subscribe so you don't miss what's next. Speaking of which, I have an incredible interview lined up next week or I should say next episode with James Herman, and we talk about advertising, the science behind advertising, but we also talk a lot about branding and some of the short form videos that I created from our interview on YouTube. You might have seen them on on Tik TOK or LinkedIn or Instagram. There's some of my higher performing ones, so I learned a heck of a lot. It is one episode that you will make sure you don't want to miss. As always, I am here to help you build, grow and transform your business using digital marketing and, now more than ever, ai first marketing. So all that's left to say is this is your digital marketing coach, neil Schaefer, signing off. You've been listening to your digital marketing coach, neal Schaefer, signing off.

Speaker 2:

You've been listening to your digital marketing coach Questions, comments, requests, links. Go to podcastnealschaefercom. Get the show notes to this and 200 plus podcast episodes at nealschaefercom to tap into the 400 plus blog posts that Neal has published to support your business. While you're there, check out Neil's digital first group coaching membership community if you or your business needs a little helping hand. See you next time on your Digital Marketing Coach.