This is the Collective Content How to choose a content agency podcast series.

In our second episode, hosts Daine Lindsay and Collective Content MD Tony Hallett are joined by two more content marketing experts to help companies choose their next agency wisely. Collective Content senior editor Shirley Siluk and writer Eve Michell join us to start our ‘What does AI mean for your content marketing?’ conversation. Check out our full How to choose a content agency guide, available here.

You can also find this episode here;

       

Episode Transcript

Tony: A lot of different AI engines – I mean we come back often to ChatGPT, we don’t have to, but obviously there’s a time that is drawing on all the data from around the Web and all sorts of billions of data points. A real live expert, their views are so up to date and changing day to day. So you’re likely to get something different and evolving in a slightly different way. And you know, the art of the follow up question – you can keep on asking AI to do follow ups. And iterate very, very quickly. There’s something about a human conversation that can also evolve in a very human way. And it’s not just something we’ve been used to for thousands of years. There’s something which I think people can tell the difference maybe. And I think that that’s going to be something that we notice.

Daine: So, so far my job as a host is safe for now.

Tony: Never say never, Daine.  We’ll definitely be seeing you for the next series. Put it that way.

Daine: Good to know!

[Music intro]

One, two, three, four

Daine: Hello and welcome to the Collective Content Podcast. This is our AI episode in our How to choose a content agency series. Today, I’m once again joined by two great guests to help companies choose their next agency wisely. We have a Collective Content senior editor Shirley Siluk, and a Collective Content writer Eve Michell, to dive into the ones and zeros. And if you want to get your hands on more expert advice on the subject, then please check out the full Collective Content How to choose a content agency guide, available on our website. Link in the description.

But right now, also joining us is Collective Content managing director Tony Hallett. Hi, Tony.

Tony: Hey there Daine, hey everyone, good to be back again and talking about this very interesting subject. I think the thing I said to you beforehand was on a meta level, we’ve talked about this being a discrete series of seven episodes all around choosing your content agency, and AI as a subject found that into our guide on this subject because it’s another option next to what we covered in our very first episode around doing things in-house, working with agencies, working with freelancers. We can definitely add on AI, maybe there’s an argument to say working through people like publishers who have their own commercial offerings, AI is a really interesting area, and we kind of talked about whether this is going to go out of date as soon as we, we record it. So we’re not really even sure where this is going to appear in the series, and I think there’s every argument for doing more episodes on this subject. So, yeah, for now, I mean, I guess the context with AI is that actually, believe it or not, it’s nothing new. It’s been around for many decades. Lots of companies have been using different aspects of AI for a while to do all kinds of clever things. In fact, Shirley, your Tech Quarterly editorship, one of the subjects which is close to your heart is AI, and that’s about how all kinds of companies are using it. This today, though, is specifically about for your content as any kind of company out there, what benefits are you going to get from it? What’s interesting, what’s risky, you name it.

[Music intro swells and fades]

Three, two, one

Daine: What do we mean when we talk about AI in content marketing, Shirley?

Shirley: I think the obvious one is generative AI, which is on everybody’s minds since the rollout of ChatGPT this year and using it for creating content. Obviously that’s what content agencies do and it seems like a natural, but there’s a lot of considerations there and there are many other applications that we can consider using AI for, whether it’s mining data, improving business intelligence, exploring automated stories or content curation, all sorts of different things that it could be used for. So it’s not just content creation.

Daine: Thanks Shirley, Eve, anything to add to that?

Eve: I think the big thing that people are using it for at the moment is idea generation as well, and it certainly depends on how you prompt it, but you can get quite good starting points before you then go to write your content. And that is definitely a way that we’ve experimented with it. Saving for actually writing, well, in quotation marks, writing content because we don’t do that.

Daine: Excellent, thanks both, for the definition there. So, Tony, how should clients talk to agencies about use of AI?

Tony: So I think the first thing to say is the any agency out there should be really transparent about what they’re doing. That’s, that’s something which is very close to our hearts and we’re not the only ones that are transparent. But you will find people increasingly either aren’t talking openly about how they’re using AI because they see it as some kind of advantage. Maybe they’re cutting costs or maybe they’re feeling their way and not quite sure of all the issues. So I would say always ask your agency to explain what they’re doing with it and they should be doing something, I’d argue, and then in which areas they’re using it and what they see as the sort of the pros and cons of that.

Eve: Just building from Tony’s point, I think it’s worth thinking about contracts and whether you’ll make it very explicit what the terms are for using AI, because that’s not to say nobody should be using it for content, but I think an agency and a client should agree on exactly how and if it will be used, and if that’s in a legal document, then that protects everybody involved.

Tony: We’ve also talked, haven’t we, over the last few months about how a lot of the time you have paperwork in place with existing client agency relationships. And some of those things go back a few years and people weren’t having these conversations then. And there are contracts in play right now where technically speaking, some of the ways that AI comes in to help could be falling foul of those contracts. And people aren’t necessarily thinking of that which, which no one’s going to thank me for mentioning, but I think is important.

Daine: Great. Thank you. And so following on from that, what are the positives and negatives, would you say, of using A.I. then? Shirley

Shirley: Well, there’s quite a bit. I mean, there’s obviously the ease of using it and the brainstorming that it helps with, like Eve said, it’s really good for ideation and exploring areas that you might not have been able to explore otherwise or did not know to explore. But the positives and the negatives – you have to be aware of both of them definitely. And I think the most important thing, like Tony said, is that transparency as a content agency, being very transparent about what we use it for and most importantly, making sure that clients understand that there’s always a human being in the loop, that we are not putting out, you know, proprietary information from our clients onto these generative AI tools that we’re not farming out our content creation and what they understand humans to be making to machines, that we are always checking the accuracy of the content that is created because that’s a really important one. You’ve got that hallucination problem with AI. So yeah, there’s a lot of risks that you do need to be aware of, but we can’t not use it either. You know, it’s the tool that everybody is going to be using it. It makes no sense to not use it, and you’ll fall behind if you don’t, because everybody else will be.

Eve: I think another risk there as well is that AI can create very samey content, very kind of bland, dry, SEO-driven content, which then makes everything feel quite similar and boring. And then it creates an issue with SEO, if everything is kind of very similar and pumped full of these key words. Then perhaps SEO won’t be as important as it has been before and Google might change its methods of ranking things if everything becomes quite so diluted with these similar articles.

Shirley: Well, and I think to add to that too, is just that the ease of using AI means that there’s going to be so much more content on any subject, with a lot of people generating content using these generative AI engines that, you know, what might have 10,000 results in a web search today, could be a million next year. So you’re competing with much more content that’s out there and you have to be aware of that. And, yeah, exactly what Eve said too, the risk of that is that the content value will be diluted because there’s just going to be a lot of garbage out there. Not everybody will be as conscientious about checking what they’ve generated and making sure that what’s produced is accurate.

Daine: Is there anything to be worried about then when considering AI use? Tony?

Tony: Yeah, I think there are several. I mean, Eve and Shirley were both talking about quality issues. There are issues around the risks and exposure that you have potentially on a legal front, even on a reputational front where you get this issue of being very close to or even using material that your competitors are using or vice versa. And that’s kind of embarrassing in either direction. We witnessed up close a situation in the last couple of years. It wasn’t actually to do with AI, but you had a very senior stakeholder in a company who was actually cutting and pasting from a competitor and not telling us. So they were they were plagiarising essentially, and we had to fix that situation. And that thankfully worked out in a good way. But you can very easily see a situation where output that looks actually pretty good, shouldn’t be used for those kinds of reasons. It’s coming from a source somewhere which shouldn’t be touched. And that’s, that’s never cool.

Shirley: Yeah. And I’ll add to that too. I think, you know, the thing like I said before, is that it’s so important to make sure there’s always a human being in the loop when you use these kinds of technologies. And the two key things that a human will be needing to do is, one, checking the validity of those sources, checking whatever AI might output and going back to primary sources and making sure that anything that’s generated is actually substantiated and has valid real-life sources that you can point to and say, yes, this is correct.

Eve: Yeah. And I think that human element is a really good point too, and that we can’t really forget that while AI is really useful, it can’t replicate an in depth talk with an SME, a subject matter expert or an interview with somebody who’s really close to the subject, like, you just don’t get that same kind of passion or involvement with AI.

Tony: We talked about how there’s so individuality in the voice of those experts. And I think part also of our thinking is with this is that, a lot of different AI engines, I mean, we come back often to ChatGPT, but we don’t have to. But obviously there’s, a time that is drawing on all the data from around the web and all sorts of billions of data points. A real, live expert, if I can use that expression. Their views are so up to date and changing day to day, so you’re likely to get something different and evolving in a slightly different way. And you know, the art of the follow up question, you can keep on asking AI to do follow up, and iterate very, very quickly that there’s something about a human conversation that can also evolve in, this sounds awful, but in a very human way. And it’s not just something we’ve been used to for thousands of years. There’s something which I think people can tell the difference. Maybe. And I think that that’s going to be something that we notice or are surprised if we don’t notice in the coming months and years.

Daine: So, so far, my job as a host is safe for now, we’re saying basically?

Tony: Never say never Daine, we’ll definitely be seeing you for the next series. Put it that way. Yeah. I’m not I’m not trusting this anyone that’s not on the team at this point.

Daine: Good to know, good to know. So how should clients use AI going forward? If we should even use that at all, then Eve.

Eve: Well, I’ve been thinking about this in discussing how you use AI. One thing I don’t think the client should do is come to either an agency or a freelancer and say, “Can you drop your rates? Now there’s AI”. I think that will leave both parties with a bad taste in their mouth. And there are ways to have a much more considered conversation about that, about how you might collaborate with AI but don’t lead with costs. I think as we mentioned earlier, it’s all about communication and having an in-person meeting or a Zoom call where you brainstorm about using AI is a good point to start with.

Shirley: And I’ll add that I think it’s really important, like Tony had mentioned before, the transparency just having an open conversation with clients at the very start of any sort of project or relationship and asking, you know, are you using AI? How are you using AI? And having that conversation about some of the risks and benefits and, you know, just having, making sure that they’re aware of the risks and using it in a way that we would use it as well – carefully.

Tony: Yeah, maybe a question for an agency or writer is, is there a way that this can be better, this being like an item of content or a whole project by using AI somehow? And I think expressly that’s not a question saying we want you to use AI to write something. I think it’s showing an openness that AI can be used for different things – for automations, for ideas, to come back to how you guys were describing it earlier. I talk about it as being a sparring partner, when especially when members of the team aren’t around, and it’s the weekend, or it’s the middle of the night or something, and you want to like knock an idea back and forth. That can be a really great use of AI.

Daine: How can clients keep on top of the evolving world of AI, Shirley?

Shirley: There’s lots of different ways to do that. Obviously, just following the news developments, subscribing to newsletters and sources that cover this Tech Quarterly, like Tony mentioned, is our quarterly newsletter that does have a special focus on AI, among other issues, and we try to cover sort of the most recent research that might be of interest to companies and clients that are considering using AI. And yeah, like I said, just being open and discussing it with clients and then talking about it and asking if they have any questions because we do try to stay up to date with everything. So if anybody has a question, isn’t certain about how they want to use AI or is using it in a way that we might want to offer some input, it’s important to have that conversation and bring that up.

Tony: And Shirley, because we have tech clients as well, we’ve had AI conversations with them for many years, but it’s been about how they’re using AI in all sorts of different fields. So the difference is just it’s now how about us as a content agency is using it? So we’ve often had open lines of communication. It’s just in a different way.

Daine: Eve, anything to add?

Eve: I would encourage people to go beyond LinkedIn for their source of information. You will see lots of carousels by people who are experimenting with AI or researching it, but if you can go to primary sources, that’s always best case scenario, in my opinion.

Daine: Thanks for that Eve. So lastly, Shirley, will AI essentially change the way we all work?

Shirley: Yeah, I don’t think there’s any question it’s going to be a very different world, you know, two, five, ten years from now. We’ve just seen with ChatGPT this year how really it exploded in everybody’s consciousness and people were generating interesting content, interesting images, all sorts of strange stuff, interesting stuff, creative stuff, eye-opening stuff. So, yeah, it’s going to change things. It’s going to change how we work in ways good and bad, so we can’t ignore it. We do have to use it responsibly and be aware of the risks and benefits, but I think it is out there. Other companies are going to be using it. We need to be aware of it. You can’t, you know, put your head in the sand and ignore it. It’s out there.

Daine: Eve you have anything to add to that?

Eve: I think as a writer it initially was quite a daunting subject to see everywhere. “Oh, all the writers are going to be sacked because AI is doing it for us now.” But I think the more productive way to think about AI is as a tool for us to harness and almost like a team mate rather than a competitor and to try and carry on that attitude going forward on both the client and the agency side is probably a good way to approach it.

Shirley: And I’d add too, one more thing. I think, you know, again, the personal aspect is so important and that is always going to come from a human content creator and a human expert on subjects. It’s not going to be from machine-generated content. There was a really interesting discussion at the Frontiers Forum about AI and the future of humanity, and Yuval Noah Harari was talking about the mastery of language and how AI could actually begin to form relationships and people could form relationships with AI, because it is so effective. Understanding context and feeding back information that you’re interested in hearing. But that person is always going to be a far more valuable connection. And like Tony said, people change, people evolve, and that will always give you better, more connected, intelligent content than a machine.

Daine: Anything to round off from you, Tony?

Tony: To round it off, I mean, this sounds obvious in a 20-minute podcast, but we’re only just scratching the surface and I mean that more generally, both in the sort of writing agency marketing world, in broader business, in society, we are absolutely scratching the surface. There’s so much we haven’t really got into here even very specifically to our world. I think we have to remember there’s going to be lots of different platforms, lots of applications. We’ve mainly been in our mind here, I think thinking about the written word, we were talking a lot a couple of months ago about one of the things we do a lot is case studies. One of the things it’s very often very hard to do is to get good photography that goes with a particular case study. And we were like, “Wouldn’t it be cool if we’ve got a story about a food delivery service in a country in the Middle East? It’s nighttime, but it’s hot and you want to show a driver who’s just got off of a scooter and walking down the road with a bag full of three pizzas.” You can give that instruction, right? That’s something that can be pulled together. You try to get that either through stock photography, probably not going to happen, or commission someone to take that picture. I mean, that’s just one little, tiny example that this whole field is made up with thousands of niche applications like that. And as I say, we’ve scratched the surface and we should do this podcast again.

Daine: I’ll look forward to that one then.

Multiple voices: This is the Collective Content podcast.

Daine: Thanks to both our guests today, really, Eve and Shirley, and thank to you as well. Tony.

Tony: Thanks everyone.

Eve: Thanks for having me.

Shirley: Thanks.

Daine: Thanks, guys. It’s been great chatting with you and looking at the bigger impact AI is making right now. And thank you to the listeners, we’ll return to the planned schedule in the next episode as we discuss reputation. So please join us again then. But in the meantime, don’t forget you can always get more quality insights in Collective Content’s How to choose a content agency guide on our website. Links in the description. Check out and let us know your thoughts on Twitter/X and LinkedIn. So until next time, take care.