Crucial questions to ask your agency about their AI policies and pricing

Haydn Woods-Williams • 00:05

Hello and welcome to the Marketers of the Universe podcast. My name is Hayden Woods-Williams. I'm a digital marketing team lead here at Brew Digital. And we're here today to discuss AI. Do you know how your agency is using AI? We're gonna discuss the benefits of AI, the pitfalls of AI, and how you can spot AI if they're not telling you that they're using it. With me today to discuss the topic, I've got our Head of Digital Marketing Services, Rich Harper. Hi Hayden. And our Managing Director, Gareth Llewellyn. Hi Hayden. To give us a varied experience of how they're using AI, how they're seeing clients use AI, and how we are using AI at Brew. I'm gonna start with a very quick story of how I was using AI recently, and then kind of move that into kind of some initial thoughts of your, both of you, on how to, or how you're using AI and how you see AI usage for agencies and for in-house marketers. So for me, I was recently running a, basically sense checking a marketing strategy through Claude and Gemini. Every time I put through a certain section of the strategy, I'd get back, "Yeah, this is looking great. This is why this is great. Okay, cool." and I'd add a little bit and I'd change a little bit. Wow, that's even better than last time. Okay, cool, cool. I'm really good at this. But then as that kind of grew, it kind of felt like it was just being a bit of a yes man to me. How can marketers trust an agency if everything they're doing is going through a yes man?

Gareth Llewellyn • 01:45

And this comes to the crux of the question, doesn't it? In terms of, are your agency using AI? And if so, how are they using it and to what extent are they using it? Because I think where we are with AI right now is we're seeing lots of extremes. On the one hand, you've got influencers who are out there saying, you can create a million dollar business with one person and AI, because they think that AI can do every single role possible. And at the other end of the spectrum, you've got people who are sort of taking a bit of a Luddite approach going, no, we hate AI, don't want to use it, we'll not touch it with a barge pole. And I think what you need to do is work out, okay, what is the sort of happy medium, the middle ground that makes the best use of AI in the right way and in the right context?

Rich Harper • 02:36

I think it comes down to if the agency are using the general platforms, then it's like instructing the same, you know, junior in your team, so to speak. Here's the task, can you go and do it? Yes, here it is. OK, great. You could repeatedly ask it to do the same thing, and it will repeatedly do the same thing, and you will repeatedly get basically the same response. Let's not beat around it. I will guarantee you most agencies are utilizing AI in some form. So first and foremost, if they're not being transparent around it, I think that's a red flag. Second of all, if they're using it correctly, I think they should be using the pro versions, they should be using and training the models to think and respond how they want it to respond. So it starts to learn their tone of voice, it starts to learn how they are thinking. If they're doing it that way and they're using it in ways to create efficiencies, but they're transparent about the fact, "Hey, we've utilized they are for this, but there you can still see the general thinking. I think that's okay. I think if they're just going to chatGBT or Claude or Gemini and basically putting your brief in and then sending it to you, you can tell chatGBT has a certain format in. Claude has its own format in. You can tell a mile off now that, like I said, it's like giving it to the same junior member in your team. They're going to give you back the same format, the same style of work.

Haydn Woods-Williams • 04:18

Yeah, I think that's so interesting. Sometimes AI creates stuff that is better than I probably would have created when I was two years into my career. And I think you have to think about it in that way when you're working with it. Ultimately, you wouldn't let your two-year marketing exec in charge of your entire marketing strategy.

Rich Harper • 04:41

No.

Haydn Woods-Williams • 04:41

Right? same critical thinking when it comes to using AI. Talking on a personal level, so this can be at work, this can be at home, how are you using AI, Rich?

Rich Harper • 04:55

I don't use it from a purchasing perspective, I don't ask it for things that I want to buy, I will ask it questions like, I quite like DIY. So the other day, there was a crumbling hole in the plaster in my wall, I took a picture of it, said to JackGBT, what could be the the problem. It provided me with some suggestions. I then followed some of those suggestions, realised that actually there was a bit of damp behind the wall, and it gave me almost step-by-step instructions as a complete amateur DIYer. Take away the crumbling plaster, you know, this is how you repair it. These are the types of products that you will need. Here's the tool set that you need. Okay, so I go to B&Q, I buy the product, I go back home, I do a botched job, it's alright, it'll be fine. So I use it like that. I also use it, and no disrespect to any gym PTs or anything like that out there, but I started using Gemini as my own personal PT. So I tell it, this is the weights that I can lift, the maximum that I can lift for each different exercise. This is what I want to focus on. I want my split to look like this. I want to train for the next eight weeks to look like this. It then gives me a training program to follow. And then I program that into a training app that I've got. It's probably saved me hundreds of pounds of PT cost. Like I say, no disrespect to PTs, but.

Haydn Woods-Williams • 06:31

I tried a similar thing recently with roots. Like, I was like, I need a cycle route that's gonna take me an hour with three hill climbs. And it was crap. Yeah.

Rich Harper • 06:42

(laughing) I found it's generally quite good. I think, like I said, I use it to answer questions. I use it for things that I perhaps don't know. Like I say, I'd have to go and research or I'd have to read a bunch of stuff online and it gives me a more rounded answer. I don't take it as like, this is the truth, but I use it as a basis. I think where it will struggle, when it comes to why I'm not using it from a consumer point of view, it doesn't give you choice, it gives you an answer. I ask it something, it gives me the answer. Like you say, you ask it for a route, it gives you a route. What it doesn't do is give you suggestions or these are things, take it to, I don't know, I want to go on holiday. I want to physically look through the hotels that are available, the flights that are available. I want to then go and maybe go to TripAdvisor and read the reviews and see the real pictures from people that have visited the hotel rather than the marketing pictures that the hotel put on their website. If I went to say, chat.gpt and said, "Find me a five-star hotel in Greece," it's not going to give me all of that information. It's just going to give me,

Haydn Woods-Williams • 07:56

"Here's an answer."

Rich Harper • 07:59

I think humans need options.

Haydn Woods-Williams • 08:02

That's so interesting because when I've used it from a, I used it recently to help me scale up an ABM campaign. You know, had kind of templated structural idea of the content I was going to create, but I need to do that scale. So I put it in and when I kind of worked through client by client and I did it client by client because it meant that I had control to make sure it wasn't going wild. And I found that actually it did give me options. It was giving me A, B and C. So I think it's so interesting as well, like how differently it can behave with different people and how everyone is having this different experience with it. Obviously with the remote world now, with a significant amount of people working at part of their jobs in a remote sense. There's always been this kind of talk of those missed water cooler moments. Do you think that ChatGPT or Claude or Gemini can be those water cooler moments? No,

Gareth Llewellyn • 09:09

in short. Well, I guess for some people they can be. And I mean, an area I don't think we probably want to get into, but the whole idea of people having relationships with their AI bots and things like that, That's a whole minefield, isn't it? So I guess in some ways, because it feels more conversational, it's not like using a search engine, it feels like you're having back and forth, then I guess it can give people that feeling of sort of, yes, I'm working with someone rather than it's my computer. But no, ultimately I think the true water cooler moment is about that personal physical interaction with people. You cannot beat working with people in person. I mean, we're a largely remote first company, so we deal with this every day. And it's one of those things you cannot beat getting together with other humans. So yeah, the AI tools, they could be great in many ways, but yeah, I don't think it replaces the water cooler moment. Certainly not yet anyway.

Rich Harper • 10:16

I don't know if there was a podcast I was listening to the other day where Rory Sutherland was a guest. Obviously everyone knows Rory. And he was talking about a case study. It wasn't AI, but it was about replacing a human with effectively automation. It was in the kind of hotel industry and he was saying about the doorman that stands at the front and essentially opens the door. And this hotel replaced the doorman with a more automated system. And he said what they were doing effectively was they, in terms of value, they were looking at cutting cost, because the automation is a lot cheaper than paying a salary to a guy that stands on the door all day. What they didn't research was the perceived value that that person has at the front door. He's not just the person that opens the door and lets guests in. He is that friendly face at the front door that welcomes people, makes them feel like they're coming somewhere. He gives them that personalised experience before they walk through the door, before they get to the desk. He may open the door for their taxi or their car when they get out, he may take their luggage and put it on. He also kind of acts as that security person. So, you know, there's no homeless people or anything hanging around sleeping in the lobby or anything like that. When you take away that human element and automate it, what you're doing is cutting costs, but you're losing that perceived value that that individual brings. And I think that's a really important point to think about in in terms of the topic of this podcast today is AI personally is gonna help change the way we work. I don't believe in terms of some circles that it's going to replace people.

Gareth Llewellyn • 12:01

In the world of tech and not just AI, but in all of tech, we're so often we're really focused on how do we cut costs? How do we reduce costs through all types of automation or AI or whatever it may be. And the problem is you are losing that value that you get from the doorman at the hotel who also represents the brand of the hotel with the outfit he's wearing, the fancy hat and everything else. And the trouble is there are so many examples, I think, in the world where a human is adding value that isn't easy to measure. And I think that's a difficulty. So often we focus on what value can we measure where we can reduce the cost because we don't see the value and we need to recognise that human value.

Haydn Woods-Williams • 12:49

It's really interesting you say that as well because a lot of the time when we think about marketing we don't necessarily, we talk about advertising and we talk about social media and we talk about billboards, but one of the things that is massively impactful when it comes to marketing is real life touch points. So a good example is the banks, the high street banks. Now, one of the things, and I can't remember where I heard this, I think it may have been from Mark Ritson, is when the banks were kind of looking at, when they were looking to take away high street banks, they never even considered that actually that was probably their highest human marketing touchpoint. So that was their marketing fascia that had the most interaction. And I guess when you're talking about AI and cutting costs and using AI in some ways, you lose that element of, like you say, seeing things from your customers focus and understanding what they value and where they're seeing your brand. And we've spoken about attribution in past podcasts and the flaws in that. I think, you know, we've gone on a big tangent there to look at how people are putting AI into their lives. And we've seen different uses of AI, you know, the Coca-Cola advert. seen AI in business, potentially, as people who work in marketing and business in the agency world, we probably are exposed a little bit more. If you are a business who maybe doesn't have that much experience in AI, or they're just kind of using AI very, very ad hoc, what kind of thing do you think they should look out for when they're starting to look to see if their agency is using AI and how they're AI.

Gareth Llewellyn • 14:52

They should really be looking for their agency and actually any supplier, anyone they work with, to really make sure they're using AI in a way that kind of augments and helps them do whatever they're doing, provide the service they're providing, so that it is adding extra value and really to make sure that they're not just using it to fill gaps in their own skill set maybe that you could be doing yourself. So for example, if you're working with say a digital marketing agency and they don't currently support you with SEO and they go, "Oh, we can start sort of providing you SEO services." If all they're going to do is plug your stuff into an AI tool and feed that back to you, they're not adding any value. They're just operate in the AI tool for you. You could do that yourself. So I think you want to be using any kind of sort of agency or company or whatever that are using AI in a way that's going to add extra value but where they're still using their human expertise, their experience and things like that so that they're not just being another interface to a chatbot for you.

Haydn Woods-Williams • 16:08

You know we've acknowledged there that the human element is very very important when it comes to how your agency is using AI or how suppliers are using AI. How do you uncover that knowledge and make sure that that person does have that experience? How much more impactful can AI be in someone's hands who, is two years of experience in digital marketing and AI. How do you see the difference in that experience in AI?

Rich Harper • 16:46

Whether you've got two years marketing experience or 15 years marketing experience, the way you position the AI, we've got the same experience. If someone's joined in the marketplace now or joined in the workforce now, they're gonna have experienced AI as long as I have. Is your agency asking you the right questions? Are they interrogating your brief? When you give them a brief, are they just going, thanks. That brief is then being fed into an NRO and you get a proposal or a response back. Or are they actually, you know, reading the brief, going into detail? Are they asking questions? Are they probing you about what it is that you want from them? Are they asking you the right business questions? How does your brief relate to your business objectives? How is this going to achieve growth? You know, ultimately, we will deal with stakeholders that are working with us, but then stakeholders that we deal with will have their own stakeholders internally that maybe C-level, board level, are they interested in the same things that we're delivering? Are we making sure as an agency that the metrics of delivering to you are important to you, but also the metrics that your boss, your boss's boss is interested in? Is what we're doing driving growth? Is it driving revenue? Because ultimately any CEO or, you know, board director doesn't care how many impressions your ad got, it cares how much revenue you're generating for the business. Are they asking the right questions? Are they interrogating the brief. Now I think that comes with experience and a recently qualified grad or someone entering the marketplace doesn't have the experience to probe those questions. They will go off of what has been briefed. And that's okay because we all start somewhere. The worry for me is that we've had 15 years prior experience where we had to engage our brains and we had to do the thinking ourselves, to give us 15 years prior experience to now probe. And if we do use AI, we also are probably prompting it in a different way than a 21, 22 year old joining the workforce. Because we're probing it with all that prior knowledge. My concern is that perhaps we enter an era where people come out of university or join the workforce, whether it's in-house agency, wherever it may be from a market profession, they don't engage the thinking because AI is replacing that thinking.

Haydn Woods-Williams • 19:04

Turning you on your head there a little bit. You know, is there maybe the idea that, you know, with 15 years experience, you might be stuck in your ways. You've got your arrogance.

Rich Harper • 19:13

I do sound like an, I do sound like that old dinosaur. I don't know. Yeah. Not giving the youth their chance.

Haydn Woods-Williams • 19:20

I mean, do you think that's fair?

Rich Harper • 19:22

I mean, potentially I'm coming at it from that, yeah, old head angle. The other thing, you know, because you've kindly flipped it, is what opportunity do people now have? If they're uniquely thinking and they're creative in the way they're doing, what more can they achieve with the use of AI if they're using it properly? And I think there is vast amount of opportunity for young people to utilize the software, the technology that they've now got available. The hope is that there will be those individuals that with, as it has been through time, some great ideas and some great concepts. My concern though still is there, and again, maybe I am an old head, you can't get away from the fact now that there is a philosophy that you can, because of all of the technology and all of the opportunity, that we can all get rich quick and we can all do stuff because AI will enable us to make money and people will get sucked into schemes where there will be someone inevitably selling a course on how you can use AI to grow a business very quickly. The only person winning out of that is the person selling the course and all these people are paying their money to learn. And inevitably it's almost like it turns into a bit of a pyramid scheme of people just selling courses on how to do AI when the person selling the course doesn't know how to do AI and everyone they're training doesn't know how to do AI that out and you end up with thousands of people that don't know what they're doing, still earning their money.

Haydn Woods-Williams • 21:02

AI works best when it's in the hands of someone who is curious, someone who is creative, someone who has prior knowledge of how things work. So actually like that in itself is something to be looking out for when you're choosing an agency. I think it requires discipline as well.

Gareth Llewellyn • 21:24

Okay. In yourself. So from my own experience, I know what it's like where I can either be working with AI in such a way that I'm getting my ideas out there and it's helping me sort of shape them in some way. And I know going through that process, it's me doing most of the work and it's a tool, kind of just like if I'm having a discussion with someone and they're helping sort of up with an idea or shape something or whatever. On the other side of that, I know I can be really lazy and just tell it, "Oh, give me some ideas for this," and it will give me the very average, very generic kind of responses. So through that experience, I think you kind of need to discipline yourself to make sure you're using whichever tool you're using to help you with your own kind of creative process, whatever that is, rather than replacing your creative process. So I think that's a key part that we need to do as individuals, but also as agencies. We need to make sure we're being disciplined, that we're using the tools as a tool, not to replace our brain.

Rich Harper • 22:37

AI is not this magic tool that is going to creatively think about brand new innovative things. It's going to replicate what is put into it. Now, it's obviously a fantastic innovation in itself, and that it's making things easier. It doesn't, it's not going to think for itself. It thinks on the input that's put into it. And actually what I think we're seeing is, especially from a creative perspective, is a real sea of sameness, because you don't even have that human uniqueness that even if you're copying someone else, you've still got a bit of personality in the whoever created it. You think from a music perspective, like I like music, there's been this huge kind of explosion of hip hop songs, current songs that have been AI generated to be like 50s blues music. The first time you hear it, you're like, that's really cool. But then all of a sudden they start regenerating 10, 15 different hip hop tracks as 50s soul music, and they all sound exactly the same. It's the same voice, it's the same beat, and all of a sudden you've then got, okay, it's not actually that great, is it? And I think that is the problem from, especially from a creative perspective and you're thinking, you know, from an agency, you would hope that your agency have got that creative thinking that if they are reliant on AI, that they're just gonna, they're gonna be producing a lot of generic mundane content that's just not going to stand out.

Haydn Woods-Williams • 24:07

Moving our heads fully into marketing, digital presence, the digital world. And I want to to think of whether that's developing a website, whether that's developing a market chapter, whether that's putting out a LinkedIn post. Do you have examples of where AI can solve real world problems? And I really want to think about this from the listener's point of view, where if we're an agency working with them, what can AI do and examples that you know in the past it was never possible.

Gareth Llewellyn • 24:47

So I think a really good example of that right now is using AI to prototype apps. And I say prototype not build because there have already been plenty of cases where someone will build an app using one of the AI tools, they publish it and it's completely insecure and there's all sorts of problems with it. So I'm sure the time will come but right now, I don't think those tools are anywhere near good enough to build things that get published. But for prototyping apps, helping you work out ideas, helping you work out what you want, because normally that's quite a long process. If a client comes to us and they say we've got an idea for an app, this is what we think we want it to look like, sketch ideas out of whiteboard of stuff. That is a very long process of building something, getting it back to them and then going "no that's not quite right, that doesn't do what I want" and lots of back and forth. Whereas now, and this is something I've been doing myself, using an AI tool to prototype something, I have a really fast iteration cycle to get to the point where it's a case of "right, this now kind of does what I want, I don't know if it's good enough to launch, it probably isn't, but it's helped me work out what it is I really want, that now I can go to a development team and go, "Right, now I know really what I want." Once you've got a prototype where you want it, ask the AI tool to then generate the specification document for you.

Haydn Woods-Williams • 26:24

Okay.

Gareth Llewellyn • 26:24

Because that will document all aspects of it, including stuff that's not immediately obvious from just using the app itself. So in terms of what it's doing sort of under the covers, you can then give that to a development team and they know what you want to produce.

Haydn Woods-Williams • 26:38

Would you kind of all be sitting in a room or a zoom room kind of using AI there or would it be a kind of like them enough they pass it to you you iterate send it back? When I use AI for it it's literally just

Gareth Llewellyn • 26:51

me and the chatbot yeah so it's literally just me at my computer just trying stuff out and then once I get it to a certain state then I would share it around with other people to try out and And again, have a bit of feedback, feed that back in. And then when we're happy, then it's a case of, right, now we can build something off of the back of this.

Haydn Woods-Williams • 27:10

Yeah, that's fascinating, especially like, if you look at the project we did for Sirius a couple of years ago, how big of an impact something like that would have had on

Rich Harper • 27:19

Yeah,

Haydn Woods-Williams • 27:19

absolutely. - On delivering that project. Rich, do you have any examples?

Rich Harper • 27:24

Interpreting reports and data. When you've got a kind of large data sets and you're trying to work at kind of speed, like we do with the clients that we've got. You know, we've got multiple clients at the end of month. You've got, you know, 15, 20 end of month reports to produce in various different formats and people have got different business goals and there's different metrics that you need to report on. Find it very useful to kind of export that raw data out of a platform and then feed it into something like Claude. There's kind of already trained. We've already, you know, we've already set up what we want it to do in behind the scenes so that when we feed it in that raw data, it's like, okay, this is it. It can help spot those trends. It can help spot things that perhaps we will miss because it is just reading the black and white numbers. We can then take some of that data, put it into our own commentary, fill out the reports. Then like we said before in this podcast, it shouldn't replace the thinking. It shouldn't replace the fact that, you know, as channel specialists, as digital strategists, that we don't understand what is coming back to us and then interpret, OK, this is trending down. Therefore, our recommendation to the client is we're going to tweak this or we're going to review this or we're going to change the creative on this ad set or whatever it may be. I don't want my team to be going into Claude or GPT or whatever and saying, what are the recommendations? All I found, and I think we're using it across the team, it just speeds up that reporting process. it gets us some insights that we can then go back to the client and say, we've seen this insight. This is what we're going to do to change it. Um, and if that's up or down, I go back to, it's that efficiency. It's helping the team be more efficient with the time they've got rather than replacing what they're physically outputting.

Haydn Woods-Williams • 29:14

Obviously, particularly in B2B, market research is something that's incredibly difficult because it's really hard to get into, um, a customer's office. Unless you're already working with someone, you can ask for feedback, but often it's bias. And one of the things that I've found is really, really good recently. And there's a company called Evidenza.ai that do it and they ran a test with Deloitte that kind of showed it was almost the same level is synthetic data research. So basically feeding Claude or whoever it is, your marketing personas, and then asking for it to create qualitative initially, and then quantitative market research data that can then feed your marketing strategy. I think, especially in this kind of day and age, that market research is the easiest thing to skip when you're doing a strategy, and you then get in a position where you're kind of guessing what your customer thinks. And while, you know, synthetic data doesn't necessarily replace that, it gives you the ability to at least validate your findings and potentially uncover some new interests. As I said, like it's for me, it's something that I've only discovered in the last kind of month. So I'm really starting to play with it now. But for me, that's something that can be literally game changing for B2B.

Rich Harper • 30:43

That's where I think it's a winner, when you can find those efficiencies and you can speed stuff up. And if you're, if you do marketing properly, like you're alluding to, you know, you should be steeped in your research. It should be steeped in knowing who your customers are, knowing what they're interested in, you know, your strategy. There is no strategy if it's not following that kind of fundamentals. And if it can support some of that, what we've got to be careful is it doesn't replace the customer voice and if you're you know, is a pool of 50 people enough to make evidence-based decisions when you could be targeting, you know, your total adjustable market could be millions. In those places where you can find those efficiencies, I think it's really useful.

Gareth Llewellyn • 31:57

As long as you're going into it with your eyes open as to the weaknesses, the limitations of what you're doing. So using that synthetic marketing data, brilliant, but you go into that knowing that, okay, this doesn't replace real market research, but this can give a good stop gap or it can be something you can use a lot of the time and then verify those results from time to time with real research as well. In the same way with my app development example, yes, it's brilliant for certain aspects of it, like prototyping, it isn't going to build me the next Amazon. As long as you're going into it with your eyes open as to, okay, what am I going to stop? What are the limitations?

Haydn Woods-Williams • 32:41

It almost feels like you need to sit down and have a grown-up conversation with your potential agency about how they use AI.

Gareth Llewellyn • 32:49

You're absolutely right. It comes down to having that conversation with them and just have an open discussion about it. Ask them how they use AI, what areas are they using it in, are there any areas where they actually don't use it, and are they going to be clear with you in terms of what they're producing for you where AI is sort of being an influence. If they're producing a report, are they going to be clear sort of if it was all AI generated or if it was a mix or whatever? I think any decent agency will be open with you and will have that discussion and just be clear as to sort of yeah, where they've got that human expertise and maybe where they're leaning a bit more on AI or not.

Haydn Woods-Williams • 33:31

With that, what are the kind of, if you could have one question that the listeners should use when they're speaking to a new agency about AI, what is that question? How do you use AI? That was a shit question to ask, wasn't it? I set myself up probably for that. Okay, let me try and re-say that. How can you not be an ass, Gareth? And answer the question with something that the users can take away and speak to their agency about specifically to help them understand if they're using AI in a positive way?

Gareth Llewellyn • 34:10

I think it boils down to opening that conversation with the agency. Ask them what their policy is on AI usage. I think any decent company should have a policy now. And yeah, have that open discussion, get them to tell you about where the areas they use it, areas where they don't use it, areas where they still find it weak. What we find from lots of clients is they want help in knowing how to use AI because they're already just scratching the surface and they realise that actually we've got more knowledge than they have with it. So how can we help them with their AI journey?

Haydn Woods-Williams • 34:44

Nice, thank you. And Rich, anything to add?

Rich Harper • 34:46

I think it's just, it's about having that transparency from the off and looking at the commercials, you know, isn't an agency charging you the same price and doing it in 10 minutes because AI is doing it? Are they lowering their costs because it's taking them a third of the time? You know, interrogate that cost. What do I get? Is it project-based cost? Is it a retainer? But within that retainer, or that project cost, ask them, you know, what are the deliverables? What are we getting for this money? And if an agency is, you know, honest and transparent, they should be able to tell you exactly what they're going to deliver. This should be in the scope of work. Do they feel trustworthy? Do they have the background? Do they have the recommendations? If they're ticking the right boxes, then they should be a right fit. If they can't answer some of those questions or you feel there's something missing, call it out or move on.

Haydn Woods-Williams • 35:51

Listen to this podcast and understanding all the different things you can do with AI, that your agency can do with AI. Why shouldn't I as a business owner or a marketer in a business just go and use AI on my own.

Rich Harper • 36:03

This question gets asked all the time. Why would you use an agency in the first place? Probably because you're time poor, probably because you're resource poor, expertise poor. It's probably one of those, you know, three or four things. AI is exactly the same. Do you have the time to invest yourself to fully understand what you're doing with AI? Or are you just going to use it generically like everyone else? Think about, you know, the trades person. I could build a wall myself. It'd probably take me twice as long and I'd have to learn about it and I don't have the tools and I have to buy all the materials or I could pay a bricklayer and he'll come and do it in a day. It's the same thing. And his wall won't fall over. And the wall won't fall over. And the wall is your marketing? Yeah. Or your brand? It could be your brand. You know, is AI going to do something that could get you in hot water further down the line? Is it going to do something that you don't fully understand? Is it going to come out with something legally that you would have done, but because you're using AI, you just kind of accept it, so you push it out the door anyway? Your agency, if they're the right fit, should know all of this and should be using it what's the right word? Ethically, they should be using it in an ethical like, they will know, hopefully, what they're pushing out, if it's used in an ethical way, it's still their own work, it's just helping level it up.

Gareth Llewellyn • 37:31

If as a business owner, if you go to your AI tool and ask it to give you a marketing strategy, chances are it is going to give you something very generic, very average, very bland. And on the face of it, it would look great. It'd be like, "Yeah, it's got all the right stuff in it. Is it really going to help you as a business stand out? Probably not. And that's where going to an agency adds the real value, adds the real human touch, because they should work with you to really understand your business, really get into the nitty gritty of exactly what your ideal customer profile is, understanding what you're trying to achieve as a business, and they should be able to produce a marketing strategy for you, or whatever it is that bit of magic that is the sort of human creative spirit that is maybe going to come up with something completely original, completely new, that really works for you and your business that an AI tool isn't going to because it doesn't understand all of that.

Haydn Woods-Williams • 38:35

That is all we have time for today. Thank you so much for listening and thank you as always to the Brood Digital team for all of their expert insights. Thank you Gareth and thank you Rich. We hope you enjoyed your listen. We hope found something useful that you can take away and use in your digital strategy. If you enjoyed listening to the podcast, please go and subscribe, share. If you didn't enjoy the podcast, give it to your enemy. That way you'll both be disappointed. I'm not sure if that's exactly what Tom says, but we'll roll with it. Anyway, I've been Hayden, and these are the Marketers of the Universe.