Paul Avery: Hello everyone. Welcome to episode 22 of Artificially Intelligent Marketing. It's Paul Avery here, one of your hosts. I'm joined by my very good friend, Martin Broadhurst. Hi, I'm Martin. How you doing? Martin Broadhurst: Um, really good. Thank you. Paul. Uh, Darby County are on losing ways at the moment. We're, we're having an awful start to the season. Uh, so if I look past the footballing side of things in my life, uh, generally, I. Things going pretty well. How about you? Paul Avery: I'm good, thanks. I think I'm gonna be even better after our discussion today because I know that we've got a lot of really interesting news stories and a fantastic interview coming up at the end of the program today. So yeah, loads of AI news for marketers that you've all come to know and love. And then an interview with Robert Rose from the Content Marketing Institute. Very, very well known. Marketing thought leader. Absolute pleasure to have him on the podcast this week talking about the Content Marketing Institute's new market research report on the impact of AI on current and future roles of marketers. That was a great interviewer. Mine, I. Martin Broadhurst: Yeah, it was a really fascinating discussion looking at their careers outlook report from 2024. Um, and obviously the story of the year is generative ai. Uh, so the report really delves into what content marketers I. Think and feel about the impact of AI on what they do day to day. Paul Avery: Well, there you have it. That's what we're gonna turn to later. But first, we're gonna whip you through these stories of the week. Right Martin, you've got the first story for us about, uh, the hype cycle of ai. Tell us a bit more. Martin Broadhurst: Well, the hype cycle has reached its peak, so it would seem, yeah, generative ai. According to Gartner, at least, we've now reached the peak of the hype cycle, and this, uh, comes off the back of generative AI startups securing over $14 billion, uh, in investment in the first half of 2023 alone, uh, which is more than five times last year's. Total, right? So massive, massive, uh, increase in investment. So despite the height, the adoption rate of generative AI has been faster than pretty much any other, uh, technology, including smartphones. And we saw that reflected in the monthly active users on chat, G p T, it was the first, uh, the first tool to reach a hundred million users or. Fastest tool to reach a hundred million users in history. I think that was superseded by threads. Although I dare say chat, G p t was somewhat stickier for its user base than, uh, Meta's Threads. Amazon, c e o, Andy Jassy revealed that every team in the company is actively working on generative AI projects, which, uh, goes to show how important this is in the operations of Amazon at the moment, even though we don't see a great deal from Amazon in the generative AI space, although they have got their Titan model and they've, uh, they are making it. Accessible for, for developers to build on various different Gen AI models, uh, on. However, so with, uh, reaching this peak hype cycle, it also means expectations are currently at the max. So while we can expect to see more gen AI enter the enterprise, there are still significant uncertainties. In widespread adoption at the enterprise level due to those common things that we hear about, like security and copyright and hallucinations and data privacy, and the all important looming regulation from the EU and the us. what does this mean, uh, for us as day-to-day marketers trying to navigate this field? Well, the rapid adoption of generative AI is giving us Lots of opportunities and making us more efficient. However, we have to be careful not to get caught up in the hype completely. Right? We do have to, uh, work out ways that we can integrate it into our existing workflows, not fear that it's gonna replace what we're doing completely take a more cautious approach. Uh, and, uh, yeah, just make sure that we're, we're, we're staying informed. We are at the peak at the moment. Everyone needs to just. A little bit of steam out of the, the pressure cooker, uh, and then take a more pragmatic approach for integrating into their workflows. Paul Avery: Yeah. It's a fascinating piece of news. We've talked about it previously, um, whilst not on air, where we are in the, in the hype cycle. And I believe the next step would be the, the, the trough of, uh, disillusionment. Right? And if that's true, then all of this promise, all of this excitement will actually lead to not much. Sort of commercial gain, commercial output, improvement in work, improvement in how we do things. It's weird. Depending on your expectations. I'm not sure that we're gonna see a big trough, honestly, because I think people are already using a lot of generative AI tools to help improve their work day to day. And certainly one of the things that could trigger a dropdown into that trough of disillusionment would be things like hallucinations and needing to fact check everything. I think certainly AI generated video. Has a fairly big traffic it could fall into, if it turns out to be quite hard to create exactly what you want. Um, and that actually takes longer and more effort to try and generate things with AI than it would be to just go shoot stuff or produce animations in a, in a more manual way. But I think it's gonna be interesting to see how it plays out. Martin Broadhurst: I feel like in some respects the trough of Disillusionment was, was reached already by a lot of users that were just using Chat G p t with G P T 3.5. They dived in. They created a few funny poems. They wrote a few email subject lines, and then found that it wasn't great at writing long form content or, you know, doing rational creative outputs, that kind of thing. Uh, but when people dive back into G P T four, they soon realize that all is good once again. Paul Avery: You could, you could definitely get better outputs from it. And so it's, it's almost like if you try and group the whole generative AI field all together and then slap them on the hype curve, I feel like you're missing some of the important nuances of different areas of where AI is enabling marketers. And I think we're probably in different areas of the curve for different areas. Um, for different areas of generative ai, but um, hey, this is what Gartner reported. We're here to report the news and tell you what we think good people, and that's what we think, which is, uh, we gotta pay attention to where we're at. But some things are actually starting to deliver value today versus possible value in the future. And I think that's only a good thing for us who are getting to grips with using the tools. Let's move on to our next story, which, uh, is about global management consulting firm McKinsey, who have developed an internal chatbot named Lilly. It's been trained on the company's entire knowledge base, spanning over 100,000 pages of documents, and has the ability to provide expert advice on any project related query, pulling up research data, relevant case studies, et cetera, from McKinsey's Vault of information. What this basically does is give every McKinsey employee the ability to think like the most brilliant consultant in the company's history, uh, quote unquote, the um, McKinsey staff have apparently experienced a significant reduction in research time from weeks down to just hours using the tool. Um, and it's quite an interesting one for us on the podcast 'cause it mimics some of the, um, studies we've talked about over the last couple of months where Different people in different roles have been leveraging internal tools to improve how they do their work. For example, junior customers, service agents who could perform at the level of senior agents by having AI that had access to all of the customer service information that they would need, and in essence, porting customer queries into the AI to help them answer them, so they didn't have to do the thinking. The AI did it. So this is pretty cool and interesting. What does it mean for us as marketers? Well, to create a chat point, a chatbot like Lilly, a company's knowledge must be easily accessible and stored in one central location. So, Certainly as a business, how are we collating and collecting the information about how we do what we do? If you are thinking about marketing, expanding out to things like customer service and knowledge bases, how can you leverage those to create perhaps customer facing bots that provide that information against something we've talked about on the podcast before? The other thing, uh, I think we even mentioned this last week, Martin was How can you improve the effectiveness and the efficiency of your marketing operation by codifying a number of things you do in marketing to enable members of staff to make decisions based on recommendations and information that a chat bot gives them, rather than necessarily having to go speak to teams in sales or more senior marketers. And so there might be some possibilities there. I don't think this will be the last time we see a large organization that has a lot of information. Collected in knowledge bases in different forms, turning them into chatbots that you can interrogate for information. What's your thoughts on this one, Mike? Martin Broadhurst: I think companies need to start thinking about how they document everything and how they store that documentation. If you're going to plug your whole company's knowledge base and customer records into something that a large language model can access, um, you need to have that well-documented, well-defined, uh, being able to, you know, create the, the necessary. Kind of vector embeddings and sticking that into the, the databases using things like Pine Cone as a, a, a method for connecting to a large language model. People haven't talked about this before. it. It's going to require a new way of thinking about how, how do we. Talk about how do we store, how do we manage all of that data? So yeah, if it's not on your, uh, horizon, put it on your, uh, horizon. I think it's gonna be different as well for SMEs, isn't it as well? SMEs are gonna have, uh, much less. Data and documentation. Then, you know, McKinsey, uh, they have huge amounts of, of knowledge and data that they can tap into. But actually, you know, we were having a conversation earlier about training of new staff within SMEs. If you've got your processes well documented, if you've got your customer data and your c r M connected into your, your, your own version of. Then you can train and onboard new members of staff more efficiently, which is gonna work brilliantly for more remote workforces and more remote teams and SMEs. Paul Avery: Absolutely reminds me of, uh, when we talked about Ray Dalio at Bridgewater and how they recorded all of their meetings and conversations going back like 20 years. I. And how they could use that data to train an l l, you know, a large language model chat bot to basically make decisions and think like a senior leader of Bridgewater. Um, that would've been quite unusual for the time required technologies that maybe weren't prevalent in all board rooms and meeting rooms. Um, but with lots of us remote now we are We're recording a lot of our calls like the one that we're doing now, Martin. So actually I think it's easier than ever for businesses to start collecting a lot of that information. Personally, I would love to get to a point certainly within Bio Strata, where perhaps we're recording all of our calls for internal uses, um, just so that we've got a data repository that hopefully a really smart AI at some point can go in. Understand what those conversations were, what the actions were, what the insights were from the discussions, and codify that so that the bot can, can learn from them. But, um, I'm not sure we have a. An AI that can do that yet, but I'm sure we will in the future. So maybe that's a recommendation to organizations, which is if you're using a lot of virtual based meeting tools, maybe start recording your calls as standard, right? Because that could end up being quite a wealth of repository of information that you could train an AI on. Right? Next story then, Martin. Martin Broadhurst: This is a story that is, Very much focused on search engine optimization and AI generated content. So if you've been paying attention to search engine optimization at all in recent years, you will know of a guy named John Mueller. Um, he is. The, the kind of search guru over at, uh, at Google, and he's often giving, uh, advice to, to people on Twitter or on Reddit about how to think about publishing and get found within search. And, uh, he said something interesting on this very topic about AI on Reddit recently, and it's definitely worth paying attention to. So in a thread about using AI generated content to get found in search, he responded and said by definition, and he said, I'm simplifying if you are using AI to write your content, it's going to be rehashed from other sites. So he's saying AI generated content. It's basically just spun content from elsewhere. You are saying things that are already known in the world, and what he's highlighting here is that, um, it's existing information available on the web and if you have ever read the, uh, the Google's search content guidelines, which available to read on on Google, you'll know that it talks about creating Fresh, new content that is contributing to the existing body of work. If I try to rank a number one, uh, for the topic of search engine optimization, I can't just take a blog post from HubSpot and basically say the same thing as they're saying. I have to be adding to the body of work. I have to be finding, uh, my, bringing my own research to the table. And by definition, if you are using purely AI written content, you are not saying anything new or novel because the AI. Is trained on an existing corpus of data and is saying what's the next most likely thing? Um, so to highlight this, they go on to say that Google isn't completely against using ai. They're quite okay with you using it. You can rank. AI generated content. It's just that it isn't going to meet some of the criteria that Google uses in its indexing and ranking algorithm. The idea of adding to the body of work, saying something new, bringing new research to the table, you won't achieve that with ai. Alone. So this is a real con uh, wake up call for, for content creators. Uh, great content isn't automatically going to rank well, but making terrible content rank well is much harder. He added in the Reddit discussion. Um, so having great content isn't the golden ticket that it, that it once was. It has to be. Great content. Saying something new, saying something that's fresh, getting people linking back to your webpage is always going to be important as well. Um, so yeah, it's definitely worth having a, a read of the full post. We'll share it in the, uh, in the show notes. Uh, ultimately what Mueller is doing is, uh, Saying you should start with a strong foundation, have a clear purpose for your content, have a unique value proposition focused on answering the questions that users are genuinely looking for, um, and that will basically give you a great anchor and a great foundation for your search engine optimization strategy. Paul Avery: Yeah. Martin Broadhurst: What are your thoughts on that? Paul Avery: yeah, it was an interesting one. Um, I think, and we've talked about it a bit, but it's such a temptation to go, oh my goodness, I can create so much content at scale now. Go, go, go, go, go pump it out. But you know, just to repeat things we've said before, but it is important, this is another version for me of don't just crank content out. Say something interesting, say something new. I think people are gonna crave authentic new ideas from humans. By all means, record a quick audio note, transcribe it, and whisper and push that transcription into chat G p t and say, turn this into a really well written blog post. That's a great use of ai, but it starts with your novel ideas, your novel human experience, sharing insights that, that the machines can't provide. So I think that's gonna be valuable. Market research reports like the one we'll talk about from Content Marketing Institute, um, towards the back end of this podcast. Again, do the research, find out novel things about the market that no one else knows and sure have chat, G P T or other tools help you write it up. That's a great use of ai, but literally saying, produce a blog for post for me the 10 most important things about digital marketing. It's just gonna churn out what's already been said, and you're gonna be very lucky to rank at all with that type of content without providing some sort of human additional value and interest and, and information based on your experience. I. Martin Broadhurst: Yeah, generative AI can definitely help with getting found in search, but it's finding those elements. Again, going back to what we said, uh, the start talking about how you can introduce it into your workflow. And I love, um, we've spoken about this so many times, I. Using Whisper to transcribe audio. Take that transcription, extract some highlights, turn that into a blog post. That is a beautiful use case of it. And you can do this with client interviews if you've working in a B two B environment where maybe it's a particularly technical product that you are selling and your technical team that know all of the nuances and intricacies, they're not great writers. They don't need to be. Now you can sit down with them, have your phone recording, extract the knowledge, and then use AI to turn their knowledge into a well structured, well written blog post. That's where generative AI is gonna help content marketers to, to really shine. Paul Avery: Agreed. I'd add one extra step to that, which is I still think humans need to know what that written output. They need to know what good looks like. Um, we find when in our tests, um, and we're not using AI driven tools massively in the agency at the moment, to be honest, because it's our technical aspects of what we write about are so Tricky for chat G B T. It doesn't really know enough about those subjects that it can't. We're finding it can't help us a huge amount, but when it does, we're finding we have to edit its outputs because it's still slightly formulaic in terms of how it writes. So I still think having someone edit the work 'cause they know what really good writing looks like. Like I would say a really good human writer might be able to hit nine or 10 outta 10 on the writing scale. I think chat G P T hits seven. Or six. So I do think there's value to be add added to the piece after it's been written, but you know, G P T five is gonna be a lot be if, if the jump from G P T 3.5 to four was anything to go by the next step, again, is probably gonna improve to the point where maybe we don't need those human editors. But I definitely think we need them now. Right. Next story we're gonna talk about Microsoft. So Microsoft has unveiled Azure Chat, G P T, which in essence is a new enterprise solution for businesses that want to leverage AI in their operations, but haven't felt comfortable using chat, G P T or Claude or some of these tools that have emerged. In essence, it is To, uh, touted as being secure and private operating through a dedicated Azure account. So whilst it has a similar or will have a similar interface to chat G p T, it's running on Microsoft servers and it's in essence as a business, it's like having your own Personal deployment of chat, g p t, um, that's private, that you feel like you can share proprietary information with, and it's not floating off and being used by open AI or anthropic to, to train their models or that your data's at lower risk. So why is this important for marketers? Um, Biggest thing to know really is if you're working in an enterprise and you're not allowed to use these tools, Microsoft and others are scrambling to make sure that you can purchase and access a tool that the IT bots in your big enterprise would deem safe in terms of Controlling and managing your data, your ip, where it goes, who can see it, and all that good stuff. There's a caveat here. There's a little asterisk, which is the first time that this came out as news. It was because someone had spotted that the, there'd been an entry on GitHub about Azure Chat G P T, and that has now been removed. So whether it was done by accident or they weren't quite ready to release this news, how soon will we actually see this tool? They're all question marks for me, especially given on the grapevine, we hear that OpenAI is also scrambling to try and create an enterprise safe chatt p t for business. And it's another great example of Microsoft being the biggest investor in OpenAI, but at times they seem like they're quite proactively competing. With each other. Martin Broadhurst: Yeah. And that's, um, where that relationship is such a, such a funny one like, it seems like there is going to be a clash there. At some point. The the kind of has to be, but hey, that's not a problem for me or you to solve. Paul Avery: no, we've just gotta figure out what it all means for marketers. Um, and, uh, I think it's a good time to move to our next story. Martin, tell us about, uh, Google, what Google are up to. Martin Broadhurst: So following on really from Google's search and how you can use AI content to get found in search. Well now, uh, I want to focus on how Google is using generative AI in search with some updates to. Search generative experience, which we still do not have access to in the uk. Generative AI built into the search results directly, and they've made a, a bunch of updates. So one of the headline features of the latest update is that search generative experience, uh, can be used while browsing. So it doesn't just summarize the search results. Now it can summarize key points and answers directly from web pages as you browse. So the goal of this, they say, is about researching complex topics. A snip, it'll just make it a breeze. You go to a page with 10,000 words on it, and it will give you a, uh, neat summary of the article and get you on your way, browsing more pages more quickly. Of course leading to more ad impressions for Google's AdSense network. I'm sure that definitely wasn't the, uh, the, the incentive there in, in short, this really sounds a lot like the perplexity Chrome plugin, so we had perplexity as tool of the week recently, and the Chrome plugin is brilliant. You go on a webpage and you just hit hit. Summarize, and it will summarize that page for you, and then you can interrogate it with follow up questions and what have you. It sounds like Chrome is getting this baked in as a, as a main feature. Um, so it's, uh, the whole. Search generative experience concept began as a kind of G P T chat, g p t esque version of search. Uh, but they're really expanding that out into the full web browsing experience. You can test these features through Google Search Labs. Um, as I mentioned, those of us in the UK don't have access to this, at present. So why is this important for marketers? If you're a business that's been reliant on. Search engine optimization and organic traffic. Well, Google's search generative experience is really coming to eat your lunch, right? There is now whole blocks of text summarizing full articles before people have even looked at your webpage. And unlike the featured snippet, which would at least have a link back to your website. The search generative experience often doesn't even have a link to the, to the source material. It will have lots of ad units around it, but the organic traffic has been pushed way down the page. This is an evolution in the way that Google thinks about surfacing, highlighting, and, and bringing content to the end user. So as a marketer, you've got to be aware of what these changes are because if you are still in the mindset of publishing content, From, you know, 2020. Well fast forward to 2024 and we're gonna be living in a very different landscape. Um, so you definitely want to have this on your horizon and be aware of the changes that are coming, uh, through Google search, generative experience. Paul Avery: Yeah, it's um, it's an interesting one. I think you're absolutely right. Ss e o is gonna be in such a fluxx while we figure out how this impacts how our content is ranked. And one of the things that's on my mind a lot here is if one of your goals is to produce educational informational content that gets indexed on search and then drives traffic to your website, how viable is that gonna be? Is Google just going to understand these webpages as they go live, synthesize them, and then provide the relevant information that a user might need within the search experience. Right? So how valuable is that as an activity then as a content marketer becomes a bit of a question. The other thing is, for purchase driven queries actually might still be worth doing, right? Because one assumes Google's goal would be to surface relevant products and solutions for a, for a, uh, a searcher. And that ultimately those people are gonna buy from businesses. They're not gonna buy from Google, at least not yet. Um, but then you look at things like Google Shopping and maybe that's actually where Google will start nudging its users to actually show, you know, products through that way. So, You could imagine a world where as a marketer, Google is mostly a paid channel, right? You're paying to have your ads shown in relevant areas of this, um, source of search, generative experience, and or paying to have your ads shown through Google shopping for relevant product or solution based queries, and that actually the organic impact of any content that you create might be limited in terms of driving traffic to your website. Martin Broadhurst: Yeah, and we see the Google Shopping ad units are one of the main ad units that are appearing in search generative experience. Um, so yeah, you can absolutely see this being a, a, a big or almost exclusively pay to play channel in the, in the near future. Paul Avery: That will really massively change a lot of how marketing is done. I think. Martin Broadhurst: Well, if you think about it from, from Google's perspective, right, they've spoken about trying to organize all of the world's information as being like their, their kind of mission, and they've always been about wanting to deliver information as quickly in the best user experience as possible for users. They've never once said. We are really big on driving traffic to your site, right, for free. They, they've never made a claim about organic search being like hugely important for them. I, it is hugely important for them because it's what brings people to their site. But if they can surface the same information, if they can deliver what the end user is looking for by going to Google and they can do that. Presenting the answers still within the Google ecosystem while serving up many, many ads and keeping the number of queries coming on Google, relatively stable because users are ultimately finding the answers that they're looking for organic traffic. It's not, it's not the driver for them, is it, it sending people off. Google is not their driver. So yeah, I think it, the incentives for Google to get this right are, are, are pretty big. Paul Avery: Yeah. And it's a, it's quite a clever mission from that perspective because We've been as commercial entities producing content to drive website traffic and then to try and drive commercial outcomes. Like I know subscribers or people buying stuff. We've been surfing, Google's providing information to people. Um, and we've been a, a lucky byproduct of the fact that one of the ways Google did that was Sent us traffic, but as you said, they are actually far more incentivized to keep people on their platform as we are all incentivized to try and keep people on ours and engaged with us. So we, we shouldn't really be surprised if it takes a turn this way and that in essence becomes borderline or paid channel. But yeah, I think if you look at the emergence of inbound and content marketing over the last, you know, 10, 15 years, something we've been very active in, Martin, it could trigger the next mindset shift really in terms of how do we go to market, right? Is content marketing in its current form, inbound marketing in its current form, gonna withstand the emergence of generative search and other AI driven sort of search chat bot style tools. Martin Broadhurst: I guess the only counterpoint. To, to what we're saying here is the fact that as long as we can put blockers on the bots that crawl our site, we still have some control and there's still, uh, Google has to keep the incentive there enough, to, to send us enough traffic to, so that we are not, uh, we're not blocking their bots from crawling and indexing their content to train their new models. Paul Avery: Yeah, I wonder if you'll see the emergence. I mean, these already exist, but the emergence of search engines that do the opposite of what Google's trying to do, and that actually we'll be generating content more for those benefits of those other search engines. It's just whether or not those search engines can offer up a valuable user experience that's better. Than what Google will offer through its new mechanism, and I think that's a, who knows at this moment. Um, while we're talking about big tech companies and, uh, monopolies on, uh, people searches and products purchases, let's talk as Amazon for a bit. Um, so Amazon is leveraging generative AI to create concise highlights for, from their customer reviews, which will allow shoppers to quickly understand product sentiment. So these new AI highlights provide a snapshot of product features and opinions, uh, extracted from all the reviews on the platform. And the AI only uses trusted reviews from verified purchases. Um, and of course, Amazon continues to try and combat fake. Reviews, submissions and all that stuff. Now, seeing as Amazon's empire was kind of built on customer reviews and being able to use that as a mechanism for selecting the products that you want, it kind of makes sense that they would try and improve that experience for people. Like it's not fun, um, surfing through a hundred reviews to try and figure out whether you should buy a product or not. Try and figure out which reviews are real, et cetera. So I can see a lot of value here. For us all as users. I think it just further highlights for marketers the importance of reviews, but also understanding how AI tools are going to summarize the sentiment and content of those reviews to offer them up to users. Because if I'm honest, when I'm browsing reviews, I usually don't pay any attention to the positive reviews. I'm looking for the number and severity of negative reviews. At this point, um, and how will the ais take those into account? Who knows? But it could definitely mean that the way that certainly e-commerce marketers, Amazon sellers leverage or rely on reviews to drive purchases for their products. Martin Broadhurst: It goes to show a really strong use case for LLMs as well, doesn't it? That this is something that they're really strong at. Categorization, classification, sentiment analysis. They're really, really good at that. Um, so I imagine that Google, uh, sorry, Amazon are going to get a really. Uh, effective model together. Personally, as someone who, you know, buys a lot of things on Amazon, like many of us do, I'm really pleased that there's gonna be a system that will summarize verified purchases for me. That's the key point in this. The amount of times you look at a product and it's got like 5,000 reviews. I think how, how many of those are real? Is it 500 verified purchases? I don't have time to go through and, you know, read the reviews and see who's verified, actual purchaser and who's, just someone that's, you know, a bot from a review farm. So, uh, I, I'm, I'm pleased for this as a consumer, I think it gives a really interesting use case for marketers to think about how they, uh, what, what they do with their customer reviews and how they can summarize and present them in a new way using ai. Paul Avery: Yeah. It certainly incentivizes people who sell on Amazon to work even harder to get verified reviews. Right. Because you can certainly imagine users defaulting to just reading Amazon summary and if Amazon summary is only verified purchases, the other reviews, whether they had much value at this point is questionable based on, you know, certainly any equals to of our user behavior. Um, but they're can have even less value if, if what people use is this AI summary that only uses verified purchases. Right. Last story of the week. We source, uh, an interesting email from Ethan Molik that you were gonna summarize for us. Martin, tell us a bit more about this. Martin Broadhurst: Yeah, more great content from Ethan Molik, uh, highlighting some, uh, research in the field of ai. Uh, and this is specifically focused on creativity with chatbot like chat, G p t versus humans. So there was a study done, uh, that found that G P T four outperformed 91% of humans in a version of the alternative uses test for creativity. That it beat 99% of humans in the Torrance tests of creative thinking. So standard human creativity tests are being aced by the state-of-the-art AI like Ja, G P t and g p t four. So a study from Wharton School, put chat G P T with G PT four in a direct idea generation contest with students from an innovation class. You. These might have some creativity, uh, juices that would give them, uh, an advantage over ai. Well, the outcome chat, g p t with G P T four produced a greater quantity of ideas that were also judged to be of a higher quality. And this is a really interesting point, more economically viable. Stick that in your pipe and smoke it. Wharton School of Innovation. Paul Avery: Yeah. Humans, Oh no, we are humans. Martin? Martin Broadhurst: I, this is a particularly interesting piece because, um, Molik is the, the professor that teaches the school on entrepreneurship, uh, teaches the course on entrepreneurship and innovation. So this presumably was uh, was actually his class. Paul Avery: Yes. Yeah. Martin Broadhurst: in another study centered on crowdsourcing business ideas for a circular economy. G P T four and human ideas were judged to be of a similar quality, but G P T four scored higher in feasibility and impact while humans generated more novel concepts. So it is creative and it sounds like the creative ideas that generates are by and large judge to be more. Well, more feasible as it says there. Bottom line, AI is proficient at generating both creative and practical ideas, surpassing humans' ability to ideate. Yet top tier creative individuals did still outperform the ai. There was a subset of these groups, the top 10% of creative individuals within the group that would actually, outperform. The ai so I guess the key takeaway is much like we've seen in the benchmarking tests with G PT four, G P T four performs better than the vast majority of humans for things like passing the bar exam. Yet there is a subset of humans, the very elite people that that sit these exams, that outperform the ai. And it's the same thing here, by and large G pt, four more creative and more practical than most humans, but there is a subset of humans that outperform the ai. Paul Avery: Yeah, I think it's really an interesting one because you think of creativity as the being the absolute domain of humans and not of ai. So I think it's quite an interesting thing to see that CHATT four is, is able to be so capable. I think if you're a marketer, This means that your AI tools are not just content production tools. You can really see them as brainstorming collaborators, right? Especially if you yourself don't feel like creativity is perhaps your strongest skill. So making your uh, chat G p T or brainstorm buddy for things like campaign themes, content strategies, new ideas for customer engagement, and those types of things is actually a very viable. Um, approach and those of you who've been listening to all 22 episodes of the podcast since the very beginning will know that Martin and I actually used G P T for, to help name the podcast. 'cause we had a brainstorm with it. Um, and I've used it in workshops with, with clients to show them how you can use the tool to brainstorm campaign concepts, um, for new product launches and all those types of things. And in general, I found My personal experience as a marketer is that I would never use any of the ideas that come out of G P T for as is, but there's usually a really good route to them that you can then expand upon, make your own, and then that's a good idea. And the artificially intelligent marketing was not what I. G P T four gave us, it gave us something else that was slightly different from this, and we were like, oh no, but that's the magic bit, and took the bit out of it that we thought was really great. So I think as marketers, how can you leverage chat G p T as part of your brainstorms And ideation processes across different aspects of what you're doing? Martin Broadhurst: And the way that you describe how we used it, there isn't that exactly the same in a team brainstorming session, right? It's very rare that if you're in a, in a group ideation session or whatever you wanna call it, somebody says something and everybody turns. That's it. That's, we are, we are running with that as the final nailed thing. Job done. Somebody will say something and you go down that rabbit hole and you refine and you iterate until you get the the final polished version. And that's what it's great at. It will give you those threads, uh, as much or even more so than a team full of humans. Paul Avery: Yeah. I, I, I think you're absolutely right and it touches on a point. Um, when I was traveling a couple of weeks ago, I was, um, having a conversation about all of this with my friend Damo, who, uh, is a listener of the podcast. Hello Damo. And we were talking about the unusual ways that these chatbots could replace human interaction, right? Because you are absolutely right. I do think that's what, how a good brainstorm works, but in essence, maybe now it's one human and a chatbot, and you can have the brainstorm and you don't need the other humans, right? I think there's something magic that happens when humans are riffing on ideas with each other to create better ideas. But what I think this shows is that you. Don't need the other parties to be humans. Right? They could be chatbots and dmo and, and my conversation was more on line management, right? A lot of people do their best thinking out loud, speaking with other people, and getting feedback on their ideas in real time. Well, chatbots might be able to perform. A first line of conversation to help people refine their ideas before they share them with others, maybe in a pitch meeting or when they go to their line manager, a particular challenge they're looking to solve. So there was a new story we decided not to feature this week 'cause it wasn't very marketing focused about. Google DeepMind developing chatbots for mental health support, similar to some of the ways that, um, PI can help you solve problems beyond just work problems and coach you through life in essence. And it's where these chatbots are emerging as, as assistance to help with bizarrely human aspects of life, right? Like dealing with stress, brainstorming ideas, but actually it looks like being quite good at it. So there you have it folks. Um, soon enough AI will truly be your friends that help you come up with new ideas and get through your life, not just something to crank out blog posts that Google probably won't rank very well anyway because they're all a remix of other people's ideas. Um, right. I think with that, we'll call it on the news for this week and segue into our interview with Robert Rose. Tell us a bit more about this interview, Martin. Martin Broadhurst: Yeah, so we were fortunate enough to be given an advanced copy of the career Outlook report from the Content Marketing Institute in which generative AI was a key focus. It's hard to get away from the impact that generative AI, specifically chat g p t has had on the role of content marketing, as we've discussed on today's episode. Um, and the discussion is fascinating. If I can just give kind of one little insight. On the whole respondents had mixed feelings about the use of AI in content marketing, uh, to the point where one of the interesting takeaways is that, uh, a certain number of respondents are using AI to write content, but they feel bad about doing so. I actually feel guilty. Paul Avery: Interesting. Well, that's coming up next for you, dear listeners. And, um, other than that, I just wanna say thanks for your time again, Martin, and we'll dive into the interview Martin Broadhurst: Our guest is a true pioneer in the world of content marketing who has spent two decades helping global brands develop intelligent content strategies to tell their stories more effectively. He's the author of many influential books on content marketing and he Is a true thought leader to marketers on everything from strategy to hiring talent and continues to push the industry forward. Most recently through his work, researching how AI is transforming content careers, which is the topic of today's discussion. Welcome Robert Rose, Chief Strategy Advisor at the Content Marketing Institute. Robert Rose: Thank you very much. It's great to be here. I have, uh, I'm a fan of the show and so I'm, I'm super happy to be here. So thanks for the invitation. Martin Broadhurst: Yeah. Delighted to have you on today specifically because there is some new content from the Content Marketing Institute in the form of the 2024 Career and Salary Outlook Report. Could you give us a bit of a brief overview as to what you were really setting out to find out and how you went about putting the report together this year? Robert Rose: Sure. So, uh, first of all, thank you for that. Um, yeah, we this is this is our second year, uh, to do this research and the original intention behind the research. And this is going back to last year. Um, and also as a, as a through line for this year was really to look at the career opportunities and current state of the industry when it comes to content marketing for salary, career path, and generally the things that practitioners of content marketing in business we're facing. And so it's a way to look at, where current salaries are, mostly us, even though we opened up the it was a global research, we really started trying to limit it down to the U S cause it's really, it gets very complex to try and do it internationally. And we didn't want to sort of get out in front of our skis too much on that, but basically looking at salary and career pathing in this year, we just, we couldn't ignore the, the huge artificially intelligent elephant in the room as it were. And so we specifically designed this year's, uh, research to. Look at the impact of what was going on in the world of AI, especially as it pertains to content marketers. And so, like last year, we had just a little bit more than 1000 content marketers respond to the research, and then we've put together this research study as a result of that. And really, the story of AI is the story of 2023 for sure. Martin Broadhurst: Yeah, it's hard to avoid, uh, generative AI after November 30th last year and the launch of chat GPT, which, uh, yeah, it, uh, it shook things up a little bit. I think that's a fair assessment. Robert Rose: I think, yes, it's a, yes, said in the most English way ever. Yes. It definitely through. Uh, a little bit of a spanner in the works, as it were. And, and yes, it has disrupted. I mean, I think that's the, the overall headline is that it has been an extraordinarily disruptive new evolution, and I think we're still working out whether or not it's going to be a force for good or a force for evil or something in between. And I think that's ultimately what our research found specifically with content marketers. Yeah. So in Martin Broadhurst: terms of how it's changing the workflows of content marketers, then what was some of the key findings in that area? Robert Rose: Well, it's, you know, it's a mixed bag really. Um, but I think probably the headline or the sort of major takeaway would be that content marketers are a little freaked out right now. They are. Looking at the opportunities experimenting with the technology and doing the best they can to get better at it. But at the same time, there is an overwhelming sense that it's going to take jobs that it's going to replace writers that it's going to, uh. Cause all kinds of disruption from an efficiency standpoint and that there may be a real threat to their careers as, as a result of this. And I think, you know, on the positive side, they're finding lots of good, interesting uses for the, you know, the main one. And I'm, this is one finding that I was very heartened to see, and we can talk a little bit more about, you know, sort of best practices, certainly, but. Um, the one that I was really heartened to see was that the main value that they seem to be getting out of it now is research and actually working with it to look at topics and look at the ability to find, you know, what's not being talked about and all those kinds of things. Um, and that writing assistance was really sort of down the list in terms of the current value that they're getting out of it. But I think overall, I think a lot of that is just. It's a nature of the current sort of ad hoc experimentation that's going on right now, because a lot of companies are still not really sure how this all integrates into their existing marketing workflows and how they're doing it. And there's, of course, concerns around intellectual property and, and all of those kinds of things. So I think it's, you know, the, the research found that it was a mixed bag, but if I had to sort of move the needle one way, it's a little more on the freaked out side than on the positive side. Martin Broadhurst: Yeah, there was one takeaway that I found, uh, well, two actually that stood out in the stats, uh, more than one in three. So 36% are using AI for writing, even though many of them feel conflicted about that. And I thought that was, was quite curious that people are seeing value in it, even though they know it's a potential threat to them. My personal take on that is it's a tool. And if you're a writer, you use it as a tool and you shouldn't feel conflicted. Good writers using it will. Get good outputs from it and it will assist you, but can't use it on its own. I think it will give you trash a lot of the time. It will feel like a bot. The other thing that I found interesting was, um, there was the question, do you use AI tools like chat GPT for these content production tasks and 25% said that they do not use AI tools. And that based on my experience alone from recent conferences and, workshops that I've ran, that's a high figure. Robert Rose: Yeah, I think and, and I think both of those things are indicative of, of what we were just talking about, which is in many cases, we're finding that there are a lot of businesses out there that have said, no, we're, we're not using. You know, we, you know, yeah, you can use it on your phone and you can sort of route around the institution of our current I. T. Policy and those kinds of things. But especially in things like health care and financial services and those kinds of things, we're finding that there are a lot of companies that have said not yet, right? So you're seeing a lot of those marketers, I think, basically Experimenting at home to the extent that they are or on their phone or or doing things that aren't part of their normal workday. And I also think and I put this honestly at the feet of many of the vendors. Um, that are out there sort of hawking AI software, um, and a lot of it is, you know, it reminds me so very much of, you know, I mean, I'm old enough to remember the early dot com days and some of the early tools that were out there for for businesses. And it reminds me very much of that. This sort of get rich quick idea with many of these software vendors that are basically No The headline of their software is replace your writers and make it auto magical. And basically you don't need to think about writing anymore. AI will do it. You know, our tool will do it all for you. And I think a lot of that marketing hype and a lot of that sales hype around the actual technology. Isn't going to the place that you're talking about, which is such the more nuanced, interesting conversation, which is, it's just a tool that helps you expand your capabilities, but rather it's being marketed as a replacement. And so I think a lot of the content marketers out there and marketers in general are going. I believe it. Right. I, and so they're trying to get good at it, using it to write sort of almost like, you know, almost like you'd hate watch a show that you were told to watch, but you don't really want to watch. It's like, they're, they're trying to get good at it, they feel kind of icky about it. Martin Broadhurst: Yeah, I think that's really reflected in the data. I've just got some of the stats from that. 62% say writers and editors will earn less respect. 55% writing will become commoditized. 46% fear lower compensation and 45% predict fewer jobs for content marketers. So yeah, there is, it, it's a pessimistic outlook. Robert Rose: It is, but I think it's going to turn, um, and I do believe it's going to turn and we're even starting now just beginning to see there's some limitations here, right? There's some real limitations in the technology. There's some real accuracy problems that are only becoming a growing challenge, not necessarily a declining challenge. And yes, the technology will get better and yes, the technology will get smarter. And there's, there's a whole other conversation I'm sure you have on other shows about. You know, sort of the, you know, general intelligence and AGI and when it, you know, when it finally emerges and all those kinds of things. And those are certainly all concerning, but when it comes to just the current state of generative AI and content, I think we're starting to see a little bit of a turn where. Marketers are realizing that there are ways to use the tool that are expansive to their own talents and bring in the ability to have human empathy and contextual information and wisdom and the things and creativity and the things that humans can bring to bear that AI can't and it can free you up. To do more of those things rather than some of the derivative things that we have to do basically just take a lot of time and research is a great, you know, is a is a great example of that. Martin Broadhurst: Yes, certainly within my own workflow. I find it's great for that summarization of content, which is, you know, you take a long form piece and distill the key points and then decide whether you need to dive in and do a kind of a thorough read of it. I think it's great for that kind of thing as a research aid. I think it is brilliant. I noticed in this report, actually, that there is, uh, there are some, some positives, certainly. Uh, I think nearly half of marketers said that they were actively interested in developing the skills required to work with these new technologies. On that point, what skills do you think content marketers need in order to compliment their creative abilities as AI usage inevitably grows within the industry? Robert Rose: Yeah, I think this is this is probably the heart of the most difficult question to answer right now. Um, because I'll tell you what it's not. It's not getting better at prompting. Um, and and this seems to be, you know, when people say, you know, because I often see that sort of and again, I put this at the feet of many of the software vendors that are out there. And a number of the other companies that have sort of, you know, claim to actually be masters of this, right? I always, I always giggle a little bit when I see these companies out there saying, we've mastered AI, we've mastered generative AI, and you can take our class or you can hire our consulting firm. We'll do that. You've been good at this exactly since November of last year. So. Nobody's good at this yet. We're all figuring this out as we go. And the interesting thing to me is that I always ask what, what does getting good at AI really mean in 2023 and probably into 2024 and beyond, and it's not going to be prompting, right? If there's anything that's going to get easier. Or more, you know, more better interface. It's going to be the actual prompt itself. I mean, you think that there's improvement to make in generative AI. That's probably the biggest and most obvious place to get improvement. I think the hardest part of it right now. Is exactly what you said, which is getting good at it means figuring out how it integrates into your daily workflow, right? So what is it useful for for you for you as a practitioner or you as a business? Where does it provide the most value? For example, I think creating content is the least interesting thing that chat GPT does. I think I take it at its name, right? I like chatting with it. I like actually talking with it and working out an idea and thinking about things and saying, help, help me organize this thought or help me write this summary or help me think through this topic where Google can't really provide me that kind of interactive feedback. If I find it to be really interesting in that sort of brainstorming, ideating, and working through topics and research that I'm doing for a particular piece, which accelerates my, my ability to be more creative and add my own spin to it. And so I think that's really the key is that getting good at it right now. Is figuring out, which is a very unsatisfactory answer, by the way, to, you know, figure out how it works best for you. It means go work with it, right? Go work with it and find out where it really works best for you. And just know that nobody knows more than you at this point, right? There, there's just people who've used it a little longer. Martin Broadhurst: Yeah, I think that's sound advice. I was, um, I was speaking to Jim Stern at the Digital Analytics Association recently. We were talking about what. I love Jim. Yeah. Yeah, he's a great guy. And we were talking about prompting and he, it has to be something he said at his conference talk at Macon as well, where he, he said to. To be a good prompter, in any particular subject area, you need to be a subject matter expert, right? You've got to know your brief in order to know how to ask the right questions to get the good results. And I think that that works with any field, whatever you're trying to do. If you're trying to write good copy, then you need to know how to write good copy in the first place in order to. Craft the prompt to get good copy out of it. If I'm looking for it to give me good advice on a particular topic, it helps if I have some understanding of that topic in order to be able to get the best outputs and understand where it might be going off piece a little bit, or where it's given me a bit of a thin answer, so to speak. Robert Rose: I think that's, I think that's really great advice but what's implicit in that advice is that it's an efficiency, not a, not a specific skill. In other words, knowing how to ask great questions. Is and it is a more efficient way to getting to your end goal. Uh, you know, in a Socratic method or whatever you want to, you know, whatever, however you want to think about it. And it does require knowledge of the subject to ask a great question in order to get insightful answers in return. What I find so Interesting is that when prompting is promoted as some sort of weird secret hieroglyphic syntax creation, you know, very much by the way, like html was in the early days, right? You know, it's like, Ooh, you know how to do html. You're some sort of secret code God, right? You know? And of course, the coders would, you know, sit back and giggle at that. The idea that we're not going to make that process Easier so that you don't have to have a prompt that's, you know, 500 words to get a good image out of mid journey or something like that is, is, you know, is it's just wrong is they will improve that. And it's already happening. Some of the software that I've seen recently is like, literally fill in the blanks and the AI will write a great prompt for you. Martin Broadhurst: Yeah. And you, you actually see this to a certain extent within mid journey. As an example, you can ask it to basically. Shorten your prompt and it will tell you which text is superfluous to the prompt and it will strip that out back right back to its bare bones or you can put an image in and it will describe the scene to tell you what you should put in in order to get a similar image next time around. Yeah, I think the next iteration of development from an AI perspective are going to be on the. The UI, UX side and how it makes it easier on that note then. So in terms of the kind of toolkit for content marketers on a, an AI tip, obviously we have chat GPT and most people have at least dabbled with that. And to some extent, are there any other tools that you would recommend a content marketer play with, even from a content strategy perspective, anything with an AI, maybe NLP bent or anything like that. That's Robert Rose: interesting. Yeah, I think, you know, the most interesting work there is a lot of, so a lot of the companies that are coming out now, I think the most interesting layer that they add is the ability for the AI to save and learn from the business from the institution or from me right in other words to save and to learn my content so that it becomes part of the corpus in the learning model that is used for my ongoing work. So without naming any vendors in particular, basically what I'm finding is, is that those and they're not terribly high priced products, but they're more than just the 20 a month, you know, subscription to chat GPT, but they, but they do offer things like. Having a specific centralized repository for all of our brand guidelines and for our terminology and for the way that we speak about things and our brand architecture and tone and editorial guidelines and those kinds of things and then doing things like, you know, there's one case study in particular that I that I love where a company made an acquisition. Of a smaller company, and they wanted to transform all 800 blog posts of the smaller company into the editorial guidelines of the newer company. And so they had a I do that basically rewrite the 800 blog posts so that would reflect the right ideas there. And I think that's a really interesting use case for how generative AI can be used to create derivative content, but then basically improve it so that it's actually matching the tone, the guidelines and the editorial strategy of of what it is we're doing. So those are the tools that I. I think our most interesting, the ones that actually help us speak more clearly or speak in a more structured way and help with our SEO and do all those kinds of things basically unburden our teams from some of the derivative things that we have to do as part of our jobs, writing that webinar abstract, writing the summary of the podcast episode, writing the, you know, and then freeing us up to create hopefully more impactful, creative, Uh, Content on the, on the backside of things. Those are the tools that I'm most interested in that, you know, and then you start to get into some of the workflow automation, like, you know, automatically responding to emails and some of those things, which I'm a little on the fence about, but, but, you know, I certainly see them from an efficiency standpoint, from a CRM standpoint, you know, some of the Salesforce. Types of ideas and, and how that's getting used to, to, to create, you know, drip campaigns and some of those, some of those more, I guess they're derivative content, but really just the, the idea of, you know, how do you create content automatically and personalized on the fly is sort of that kind of idea. So I would explore those kinds of tools if I were in a bigger marketing team and that's where our focus is right now with the clients that we deal with every day. Yeah, I think Martin Broadhurst: that's a good tip on what to look for when, when selecting any of these tools. Um, but just back to the report, on the skills as a follow on to that. I know one of the questions was asking about career advancement and I think about a quarter of marketers believe that their company provides a clear career advancement path. Yeah. So in light of. AI disruption, why is this such a problem and how can companies adapt to make sure that they're retaining the experienced talent that they've got in house? Robert Rose: You know, this is, this is my favorite part of the research and it's, it's the original reason we got into this research in the begin with, but was because of the. What we, what we believed anecdotally, and it turned out to be true was that there was really no place for content marketers to go within most large organizations, right? Because you basically had one of three choice. Once you reached sort of senior manager status, sort of just beyond sort of the single contributor and into sort of managing a small team, there was really no place to go in most businesses from a career pathing standpoint. You either one found another corollary job at another company. You know, another content marketing job doing the exact same thing in another company to you matriculated up to some other job in marketing, right? Some other, basically non content marketing role. So if you wanted to go to director, you got to go to product or you got to go to comms or you got to go somewhere else or three jump out at, you know, and they sort of abandoned ship and start their own consultancy. Right. So those are the kind of the three options that they had. And as it turns out last year, we, we found that exactly to be true. And I think I has only complicated that over the course of the year because now we're seeing that even more so is that companies don't have career paths for content marketing or content practitioners full stop and arguably it's because they've. Really created a confusing career path in marketing more broadly, right? It's because it used to be back when I didn't have so much gray hair that you could, you joined a company and you could, you know, easily matriculate up through, you know, I become a manager, then I become a director, then I become a senior director, then I become a VP. And then hopefully I become part of the C suite. Well, that doesn't really exist anymore because marketing has basically siloed off. You know, there's now social media and there's brand teams and web teams and content teams and now AI teams and even digital marketing versus regular old marketing, you know, and so it becomes very confusing for marketers generally, but content marketers specifically to find their way. And so what we see and reflected is that many of them just have sort of. They've given up right in terms of trying to find a career pathing that that makes sense for them. And so when you ask, how can companies really retain the talent that they have in their content marketers? It's make it a thing, right? Create, create a career path so that I don't have to jettison after I become a single level contributor and I can actually build my strategic career off the back of saying, I can matriculate all the way up to the C suite, you know, I can become CMO, you've got to create that belief in the content marketing team in order to make them not feel like they have to jettison out at some point, because there's no, there's no room for them there. You Martin Broadhurst: think there's a, well, those people in the C suite that aren't looking at content marketing as its standalone thing, what, what, what do you think that's going to take to get that shift to make them see, to see the value in, in, in doing that? Do you, you know, I've heard when we've spoken within the industry for years about. Companies thinking more like publishers and, and you, you know, you see some companies taking that very literally and buying, you know, HubSpot buying the hustle and various, not that they weren't thinking like a publisher anyway, I think they were already on that tip. Um, but you know, what's it going to take for organizations to, to make that leap, to, to give that career path, if they're not already in that mindset. Robert Rose: Yeah. I think it's going to come to the realization that. Really getting good at content. Uh, and I mean this from not just acting like a publisher or acting like a media company, but getting good at content means acting in many ways, like a media company. So in other words, I think marketing generally speaking is becoming more like content marketing than content. Marketing is, is integrating into marketing, right? Businesses are slowly realizing that their ability to create multi channel content driven experiences. Is the way forward, right? That's the way forward for advertising. It's the way forward, you know, which we would largely call paid media. It's the way forward for our owned media experiences, our websites, our blogs, our resource centers, everything we're doing, our email newsletters, our podcasts, all of those kinds of things are becoming more and more important, not less and less. And so our ability to speak clearly with clarity and with an impact that gets beyond all of the other things that are vying for our audience's attention is a core fundamental part of modern marketing full stop. And so. Making content a strategic function in the business and not just an aspect of everybody else's job, right? You know, my little tweetable moment that I often get quoted on is that content is everybody's job and nobody's strategy. And so for most businesses, they have to come to that realization To make content as functional, a part of the business as accounting or legal or marketing or sales or any other part of the thing that we care deeply about. And that includes the use of AI as a fundamental platform by which we build content as a strategic function. And I think it takes that level of realization to go, ah. It's not marketing ourselves like a publishing company. It's not selling ourselves like a media company. It's operating like a media company does and how they operate and creating content is a strategic function that is the key to modern marketing. And without that realization, I think they, you know, it will always be working with one hand tied behind their back. And that means at some level making content a true career level kind of position within the organization. Martin Broadhurst: And I dare say that anyone wanting to know more about that subject could find out in the upcoming book due out in September. Well, Robert Rose: thank you very much for that. That's, uh, that's, that's very kind of you to say. Yes, indeed. That is really the theme and the topic of my upcoming book. Yeah. Martin Broadhurst: So to flip that, um, and not looking at people who are, um, you know, that maybe reach the ceiling at content. What about people at the start of their career? Content marketing careers, maybe early career. They've just come into a team that they're doing the content thing, but they're concerned about AI, what career advice would you give to those in the marketplace now? Robert Rose: It's really understanding your role as content practitioner isn't typing words or making pictures. The differentiator that you have is that you are you and that you bring a certain level of ideas and thinking to the table. AI just simply can't bring. Um, and so what that means is, is getting good at something other than typing out words. And what I mean by that is that whether that's for you getting really deep. Into a particular vertical and becoming a subject matter expert in that vertical and so that you know what the trends are or, you know. How to position things where you know how to strategically engage as a more expert level audience, or you can address things, you know, becoming really good. You know, think of it like becoming a really good journalist, right? You'd be great. Journalists are also typically subject matter experts in the thing that they really cover, whether that's politics or legal or finance or health care. They become strategic subject matter experts in those topics. That's key or becoming more horizontally strategic. In other words, becoming a business strategist or becoming a marketing strategist, or becoming a media strategist, or becoming a overall sort of strategic part of providing counsel to. Whoever it is you work for, that's the key layer, finding a layer that goes beyond just simply writing or creating pretty pictures and saying, yes, I can do that. I'm a great writer. I'm an amazing writer, but I also offer strategic thinking on X or strategic capabilities on Y or all of those things. Adding that to your sort of quiver full of arrows in terms of where you are marketing yourselves to what kinds of companies and where you want to work, or if you want to go out on your own, that's the critical piece. It's really at the broadest sense, adding wisdom, adding wisdom to your skill sets. And by wisdom, what I mean is the combination of experience and insight. And the ability to express the, the sort of fusion of that experience and insight into something that can provide judgment on a particular topic and bringing that wisdom is going to be a differentiator to AI always. Martin Broadhurst: Yeah, I think the entry point for so many people is, oh, I enjoy writing, I, I can write well, and that will take you so far, but that next level is so important. On that, that point about judgment, though, that's such an interesting one. Uh, Wired magazine's, uh, Article that they published a few months back about how they will and won't use generative. I touched on that very issue where it spoke about they won't use generative. I for writing articles except where the fact that it was written by I. Is the very point of the article, but also that they won't use generative AI for editing because editing requires judgment and human judgment about what's important, what angle is what so yeah, I think that's such a, an important takeaway for people is that I, you know, editing isn't going to be replaced by AI anytime Robert Rose: soon. It's a great point. Yes, it's a and and as we all know, those of us who had any level of experience to writing, writing is rewriting, right? Writing is editing, you know, just getting your idea out on the screen or on a piece of paper is, you know, that first draft. Anybody can do that. It then takes the special skill of being able to mold that and shape it into something that is truly valuable. That's the real skill. And to your point, I is not terribly good at that. Um, it's good at organizing patterns and recognizing patterns. So when you think of things like. You know, writing a summary of the last football game or looking at the, you know, financial results of an analyst that says, here's what the company did, you know, in the last quarter, basically organizing facts and figures into a readable paragraph is great at that. What it's not great at is saying, what does that mean? Like, what does it mean that Metta had a great quarter? What does that mean to the future of social media? You know, that's the, that's the real key of becoming a true wisdom creator. Yeah, Martin Broadhurst: why is it important that's Robert Rose: why is it important? Exactly right. Martin Broadhurst: Yeah, looking ahead. Um, what do you think of the big priorities and for organizations are looking to integrate AI into their content marketing teams? Robert Rose: It's the same priorities that I would put for any technology, which is and this is one of the challenges that we've seen with so many marketing technologies. Over the last 20 years, which is many businesses tend to put the cart before the horse. And what they do is they say, ah, there's a new technology that can help me with X. How do we do X right? And how do we figure that out? And so they add X and don't have any clue as to how it will actually help them further their strategy. And so what we say is, you know, it seems like yeah. But, but it's rare that it happens is you've actually got to figure out what your desired workflow, your desired process, your desired activities for your strategy are, then go, okay, where are the gaps that we have in terms of our ability to scale that with a tool and figuring that out. Then you go great. A. I can fit in this gap in this gap in this gap. Now you're prepared to actually go out and look for a solution that can fill those gaps because you know what you're looking to actually, you know, you're looking for the tool that actually helps you scale the process, which is all technology is is the tool to help you scale a human driven process. Martin Broadhurst: Yeah. Start by identifying the use case, like where that bottleneck might be in your content production or something like that. Yeah. I think that's a, um, yeah, entirely pragmatic, but also an important step for people to, to undertake. Just looking at the report, then I'm interested in anything that you thought was particularly surprising. Robert Rose: I'll point to two things, one of them having to do with AI, which was, and I mentioned it earlier in the show, which is I'm heartened and was a little surprised actually by the top uses the top values that Content marketers are getting out of it. I expected it to be, you know, it helps me write blogs easier. It helps me write, you know, an ebook faster, you know, and all of that. And I was really heartened by how low writing assistance actually came down on the list and how high things like research and looking at topics and using it. As a research assistant kind of thing, um, I was really heartened by how high that was on the list. And that gives me hope that content marketers are actually really figuring this out. Like, are they're really starting to identify where the real value of a tool like this can be and not be so scared of it, which is a great thing in my mind. The second thing that I would mention, uh, is on the, on the career pathing side, How high actually, and it was just over half, I believe, if I'm remembering the number correctly of content marketers that feel good, that are enjoying their job, that are basically that generally speaking, the interesting tension is that most content marketers now in their work are happy with their work. So that's a big shift over the last 10 years, uh, certainly anecdotally. And then, and then from our research last year is how many of them are actually enjoying the work, enjoying, you know, and finding it meaningful, which is a huge piece, by the way, of what most content marketers are looking for in a position is the meaningful work part of it. And so I'm really heartened again by that, about the future of content marketing, by knowing how many content marketers are actually feeling good about it, but then you sort of add that tension of, yeah, but. If a new job comes along, I'm I'm ready to jump out the window, right? I am ready to go. Um, so it basically tells me that the work needs to be done on the business side. We need to make content marketing a much more strategic function because they're finding great value there. Meaningful work their success to be had. We're just Providing a career dead end that can easily be overcome and it can actually become a differentiator for talent acquisition in the years to come. So, those were actually my two, two biggest surprises in there in the research and both of them on the positive side. Martin Broadhurst: Yeah, I think that's a lovely place to wrap it up. Actually, there is an optimism, even though we started off somewhat pessimistically looking at it, there is generally a sense of, uh, of enjoyment in the work that people are doing. Uh, if I remember rightly, salaries, uh, were quite healthy. People Robert Rose: seem to be. Yeah, a hundred percent. Yeah. It was, it's, you know, it's, it's a little, it gets a little funky with, with trying to do salaries because of urban versus suburban and regions of the, you know, the U S is just a very large country. So it gets a little weird, but yes, generally speaking across the board. Yeah. Salaries are pretty healthy for content. So businesses are valuing it, but it's interesting. They're valuing it only at the execution level. They're not valuing it all the way up the chain. Martin Broadhurst: Yeah. And that is where the, uh, the real work is yet Robert Rose: to be done. Martin Broadhurst: Exactly. So, uh, leadership teams pay attention and make it a strategic function. So, um, where can our listeners download the report? Robert Rose: They can head on over to content marketing institute. com. So, um, you can download the report and, uh, and hopefully get a ton of value out of it. Martin Broadhurst: Thank you very much. Well, really enjoyed this discussion. So thank you very much for joining us, Robert Rose: Robert. Absolutely. My pleasure. Thanks so much for having me.