Tom Rudnai (00:13)
Hello and welcome to another episode of Demand Geniuses. I'm Tom Rudnay, same as I was last time. And with me today, I'm going to get straight into introducing, rather than do a long preamble, and introduce Araminta Robertson. Araminta is the founder and CEO of Mint Copywriting Studios. First of all, hello.
Araminta Robertson (00:32)
Hi Tom, lovely to be here.
Tom Rudnai (00:34)
Thank you for joining. guess rather than me butcher an introduction into you and your background, do you want to join a movie? Just give us the headlines and tell us a bit about Mint as well and what you do there.
Araminta Robertson (00:43)
Yeah, definitely. So yes, I'm the founder. call myself a managing director rather than CEO because we don't have a C-suite. it's managing director of Mids Studios. We're a content marketing agency for fintech and financial services. And we help these companies turn content into measurable pipeline. And we've worked with companies like Persona in the US, GBG, ClearBank. These are some of the names in the fintech and financial services space.
Tom Rudnai (01:14)
Nice, and why did you pick when you kind of started that then? Why specifically FinTech and financial services? What was kind of uniquely interesting about that for you?
Araminta Robertson (01:22)
Yeah, I mean, there's a whole backstory to it. think it all started when at the age of 18, I finished high school and I realized that I didn't know anything about finance as many 18 year olds do because we're not taught about taxes and mortgages and how to invest in high school. And I realized that and I also realized if I did learn about these things, I would probably set myself well for the future. So I started a blog.
And I would read a lot of these typical financial books, you know, the Intelligent Investor, Richard Perda, all of those books. And then I'd read them and then I'd write about them in this blog. And it was kind of aimed at other kind of young adults, like, let's figure this out together kind of thing. I was living in Barcelona at the time, which is where I kind of grew up. And I met a fintech company at a networking event. That's a kind of a business event. And they saw the blog and they liked it. So they hired me as their kind of...
content marketing person. And that's kind of how it all got started. So it really started from a passion project. And then as I did more and more of it, I just loved it. I just loved how when you learn about personal finance or financial topics, it gives you control because you start to understand how the world works. And I also love that now obviously we do more B2B rather than personal finance, but I just love how payments and finance are kind of like the arteries or the blood vessels of the world, right? Like without that, you've got nothing.
Tom Rudnai (02:31)
Hmm.
Araminta Robertson (02:44)
you can't, the world doesn't work right without payments. So I just love how it's such a unsexy and yet so important part of our economy. And also the people are fantastic in this industry. They're so generous and lovely. And I think it's because it's an unsexy industry. It attracts people who genuinely want to do this or genuinely are interested in this topic. So, yeah, that's a bit of a backstory into how I got into it.
Tom Rudnai (03:10)
So it's interesting though, because it is unsexy, there's nothing like it, it doesn't sound fun, FinTech, but it's like the operating system of the world and I can see that going into the wheeze on how all of that stitches together must be quite interesting, revealing and probably quite helpful just from a personal perspective. I'm terrible at things like that. Nice, and then one question I always love asking early on, because I think it's really interesting hearing different people's answers, it's like...
Araminta Robertson (03:15)
Mm-hmm. Yeah, exactly.
Mm-hmm. It's exciting.
Yeah.
Yeah, exactly.
Tom Rudnai (03:36)
When you look back at the stuff that you've done so far, is there a step in your career that you would say was the most formative or like a moment that you felt that you really like clicked and you were like, I know what I want to do and I know what I'm good at.
Araminta Robertson (03:48)
Yeah,
I guess there were multiple, but I mean, just because it was the beginning of it all, guess, you when I was 18 and I read a book, I started reading personal finance books and I think that's, at that moment I realized, okay, this is what I want to do. I think another one was potentially when I went from freelancer to agency, it was an interesting moment where I had a current client as a freelancer who was saying like, okay, where are the conversions? Like traffic is great, but we want customers.
I was reading a book called They Ask You Answer, which is fantastic. Like I call it like the Bible of content marketing. And at the same time, I was doing a course by growing Convert, a fantastic content marketing agency on how to generate conversions with content. I did all those things, like all those things were happening at the same time. And that's when I, it all clicked and I was like, perfect. Now I know what I'm going to do with this agency. We're going to focus on that because I can see the problem there. So yeah, multiple, but maybe those two are the biggest ones.
Tom Rudnai (04:42)
Okay, and then I guess I think it's always one thing I always want to do with this podcast is make sure that I frame any advice that we give because I think there's advice out there everywhere and the really hard thing is understanding which advice to listen to. So before we get into some of the stuff that we want to talk about, which is that content, AI search, things like that. What's unique about FinTech specifically from from like the content marketing conditions.
that would influence, let's say if I'm someone listening to this podcast, whether or not I should follow your advice. In what conditions should I follow every bit of advice that you're about to give us?
Araminta Robertson (05:14)
Well, to be honest, first of all, I'm not a huge fan of, like advice is so contextual. It's very hard for me to give good advice without knowing the situation of each person. So I don't like giving personal advice from business advice. would, that's why I take it with a grain of salt. It depends a lot. I can just share what's worked for us and what we've seen. So that's just one thing to say. So what is unique about content marketing and fintech and financial services? The first one is that the stakes are higher. So, you know, when you're opening,
Opening a social media account, opening a buffer account like a SaaS tool is not the same as opening a business bank account or even choosing a banking as a service partner. The stakes are way higher. And so from a content marketing perspective, means especially SEO, that means you're in the YMYL bucket, which means your money or your life. So Google classifies anything around those topics as your money or your life topics. And the reason is because if you give the wrong advice or the content isn't correct, the consequences are much higher.
Right? Telling someone to invest in crypto coin and lose all their life savings is a much bigger deal than open this SaaS tool that whatever costs $5 a month or whatever, you know, so the stakes are higher. And so that means that your content needs to be more rigorous and more like you have to be more careful about what you're writing about. think that's number one. Number two is that it's more complex. Whether you're B2C or B2B, writing about payment infrastructure is more complicated than talking about
a recipe, would say. mean, I don't want to put down people who write recipes, but it's definitely simpler than giving instructions on how to build the table or whatever it is. So it's more complex, which means that, again, the content, you have to be a very strong writer to know how to write in a way that can make that complexity easy to understand. And then the last thing I would say is, and this is kind of linked to the two others as well, is like expertise and trust.
especially in the B2B sector, know, people are looking for content that is more advanced, more technical, as I said, complexity, but they also want it to come from a place of expertise, right? And often you're writing about topics that, you know, maybe haven't been written about before, are, again, the stakes are high, so you want to make sure you're writing or you're reading from someone who knows what they're talking about. And all of that combines together to build trust, because in financial services, trust is super important.
way more than, as I said, other industries potentially because the stakes are higher. So I think this is why content marketing comes in so hand to hand with financial services is because trust is so important, expertise is so important. And the best way to build that trust is to create really like high quality content that explains that complexity that comes from a place of expertise.
And yeah, it's one of the key distribution channels for a financial services company. That's what I say is unique about it.
Tom Rudnai (08:15)
Super interesting though, because what you're describing is to me, a space like, it's the hardest possible or the most extreme conditions in which to market, right? And I think of like, what is it, like diamonds, like they're built under extreme pressure and that's how you form a diamond. That's how you, it's probably a very good breeding ground for a good content marketer because what you've described is you have to be able to boil complexity down to something that people can understand. And there is a...
Araminta Robertson (08:38)
Yeah.
Tom Rudnai (08:42)
even higher bar of expertise and trust at a time where that's quite a hard thing to develop because I think there's a shortage of that in digital spaces at the moment. I want to come back to the your money your life thing is that a is that an official is that that's a google recognized thing that they have a category of this and their algorithm works in a different way.
Araminta Robertson (08:48)
Mmm.
Yep. Mm-hmm. Yep.
Yeah, and it's not just finance. So healthcare is in that bucket. You know, other sectors where, again, the stakes are higher. If you give someone bad advice on how to resolve, I don't know, specific disease, that's bad. It could affect their life. Same with financial services. And I always find myself attracted to those industries because, again, like the stakes are higher, means your work is more important. It's more exciting. It's like, can really, the impact that you can make on the person who's reading or a customer is so much higher.
And that's really exciting to me. yeah, YMYL is a term that Google uses, yes. And I wouldn't say, does the algorithm change? mean, I'm not a technical SEO expert, but it's more rigorous. They'll say, know, EEAT, the framework that Google, yeah, they'll say that it's even more important for those sectors, basically. That's what they'll say.
Tom Rudnai (09:53)
does that impact when you're kind of building a content strategy? I'd imagine that has quite a big impact on how you approach it because I assume it makes it little bit more of a slow burn to kind of start getting things ranking and building up reputation because there is a higher bar for them to kind of consider you credible. Is that something that factors into the way you build a strategy or?
Araminta Robertson (10:09)
Yeah.
Well, I would just say that it just, it means that your content is going to be harder to produce and your standards are going to have to be a lot higher. You also like have compliance and everything that you have to deal with. So, for us, I mean, we, especially we only work in financial service fintech. So we know this really well. And, and, but, but even, guess, if we working with a non-fintech, we would still have high standards. And because, you know, especially in B2B, the people you're targeting, you want them to take you seriously. And if you just write a generic.
lame, boring piece, like they're not going to take you seriously. one part of our process, for example, that's a key part of this is that we interview experts, right? So especially when we first start working with a new client, we'll spend a lot of time interviewing the experts on their team, maybe for each article, but then eventually as, you know, as we repeat certain topics and stuff, that's no longer necessary. But we spend a lot of time interviewing experts, understanding the product, understanding the pain points to make sure we're writing about the right topics.
but also that the content itself is written for their level of knowledge. It's expert based. It's as complex as necessary. And it's always a nice one. A client's like, wow, this is really good. And I'm like, yes, because we were able to do an expert. That wouldn't be possible with just desk research. And it's funny because a few years ago, sometimes we'd get pushback from clients. They'd be like, come on, can you not just do desk research? You're taking up time. We no longer get that pushback. And I think it's because of AI. Like people can see that.
If you don't interview an expert, it's just going to be AI slop. Like what's the point? Why read it from you? If you're to say the same as everyone else. And so I think introducing the expert into that process is just so important, especially for, for a FinTech and financial services.
Tom Rudnai (11:55)
interesting because you've preempted my question a little bit there which was going to be what you've described is that essentially creating content in your space is necessarily slower right you have to be more diligent you have to be more detailed and I was like is there a bit of a conflict there between what it seems like everyone is kind of obsessed with at the moment which is speed efficiency cutting headcount doing more with less all of those phrases that every marketer probably wakes up at night terrified of but you but you found the opposite it sounds like
Araminta Robertson (12:06)
Yeah.
Well, mean, it depends again on the product and on the client. So for example, for clients where the product is more straightforward, think of a business bank account or a money transfer service. Definitely, and even for the complex clients, there is a place for AI. So we'll talk about later, I guess, our AEO framework, GPT articles. Those are AI assisted. And we're right now developing a tool to help us with one part of the process, even for content creation. So there's definitely a place for that. But I think that...
you know, as long as you include the experts and make sure that the quality is high. I mean, you do what you want, right? Whether it's AI or not, like that's ultimately up to you. And I think it has just because it's more complex. Yes, usually, you know, that means it takes a little bit longer, but we're a lot more focused because we're outcome focused. So X number of leads, MQLs, SQLs. It doesn't really matter how many pieces of content you produce as long as you're hitting
as long as that's growing and you're hitting your goals, ultimately it doesn't really matter, right? So we might produce less or fewer pieces of content than another agency, but because we're tracking the metrics that really matter, ultimately it doesn't really matter, right?
Tom Rudnai (13:35)
Yeah, it's interesting. it makes me, I guess there's a bit of a selection bias here as well because obviously people who are talking to and engaging with an agency like yours are the people who are interested in doing good quality content and investing in that. But it doesn't contradict. So we did a big study recently where we basically used our AI to analyze like 50,000 pieces of.
Araminta Robertson (13:46)
Mm-hmm.
Tom Rudnai (13:54)
content in B2B Fintech specifically, because we were quite interested in understanding how that had evolved over the last five years. It was really cool because actually one of the things you could do is like tell a story of the industry and how the industry as a whole is evolving just from how the narratives and the storylines change. But one thing that we did notice was there was a huge increase in content velocity. So the speed of output had increased massively, which obviously makes you fear that AI slop. But we actually one of the cool things about the way our product works and the way that we were doing the study was we were looking for quality markers as well.
Araminta Robertson (14:13)
you
Tom Rudnai (14:24)
and they were also on the rise, which kind of reinforces what you're saying, is that as people have increased that velocity, I'd imagine there are some notable exceptions to this, but generally speaking, quality was still improving.
Araminta Robertson (14:34)
Yeah, there are ways to do that.
And we're doing them. there are, like for example, if you've got already a large body of content published on your website, like that's huge. Like it's so much easier to create good quality AI generated or AI assisted always. Don't do AI generated, AI assisted. If you already have lots of content, it will be so much easier. And that's why brand new clients that don't have a lot of content, that's where a human writer is a lot more involved because there's not much to go from.
But if you already have lots, we do a lot more AI assisted with the client. We're like, OK, cool. You've got some really good stuff here. So it also depends on where you are on the journey.
Tom Rudnai (15:11)
Yeah, well, because that's something we've been playing around a lot with is so we have our kind of information game framework, which I won't go into now. But trying to experiment with how we can use AI to create the highest possible level of information game within a piece of content. And we found that if you just use AI from scratch, it's rubbish. It's not going to get it done. If you feed it with your perspective, you can get our kind of level one, which is interpretive game, so slightly new slot on that. You can just give it your brand voice and do it. Anything bigger?
Araminta Robertson (15:29)
Yeah.
Tom Rudnai (15:40)
It just will not get to but what you can do is if you've got great research pieces and you're giving it quality sources that it can infuse in with other things then it can do a fantastic job and I guess you're saying for a lot of brands who already have that it's a big bonus
Araminta Robertson (15:55)
Yeah, yeah, I mean, I always say that your output will be as good as your input. So if your input's really good, so will your output. If we take on a new client and maybe they've got a lot of content, but it's not great, we'll still need to do a bit of human-written content for a bit, and then we'll do more AI-assisted. But yeah, if your input's really strong, then that's where AI can really be fantastic.
Tom Rudnai (15:58)
Hmm.
Yeah, that makes sense. Let's talk about the content itself a little bit then. So, I'll start with what in your experience makes a piece of content NLM optimised?
Araminta Robertson (16:27)
Okay, so to answer this question, I'd have to backtrack a little bit. So around October, September, 2025, we developed a framework to help companies basically influence LLM visibility, get recommended more often by LLMs. And so the way this GPT, we call it the GPT framework, and the way this works is you start off by deciding the bottom of the funnel topics. Like this is our whole thesis, start with the bottom of the funnel, because there are,
the people who are ready to convert, these are people who are aware they have a problem and are actively looking for a solution. already have money in their hands, they're ready to close. Start with those people and then work your way up the funnel. That's what we do for SEO and content. We have taken that also for LLM visibility, optimization, and all of that. The way that framework works is you start off by deciding the topics that are again bottom of the funnel, and it can be the same ones as SEO ones as well. Then once you've decided on the topics,
put together kind of prompts that you think people might be typing into Chachapi, TGEM, and I, cetera. So you can do, for example, for one topic, maybe 10 prompts. How do we come up with a prompts? Honestly, it's by understanding your target audience, putting yourself in their shoes. What do think they'd be typing? So it is a guess. Yes. Because I mean, there's no way that you can, there's like millions of variations. You're never going to get it exactly right. So you put together kind of maybe prompts. And then after that you.
create short articles, maybe 500 to 800 words, answering each of those prompts. So the way I think of it, like a bit of a hub and spoke model. At the center, you've got your SEO piece that's very in depth, that goes into detail about X topic. Let's just say open banking APIs, for example. And then you create your GPT articles. These are short articles, and they answer different questions about open banking APIs. So for example,
tell me an Open Mac API that does CVRP or what's an Open Mac API that works across Europe and not just the UK. So all these different variations of that topic and then you answer them and make sure they're bottom of funnel. That's the main thing, right? So I think the number one way to ensure that your piece of content is LLM optimized is to first go after the right prompts. I think that's 80 % of the work. If you go after the wrong ones, then there's no point because say you go after a topic like
What is open banking? ChachiBT will just give you a straight up answer, but it's not going to mention any brands. So there's no point in really targeting it. It's going to use your content. It probably won't even source you. So start with the bottom of the funnel. And then after that, honestly, we've been experimenting and testing. But I think the main thing is make sure you answer the question in the article. It likes a few different things like tables, headers, all stuff that is good SEO.
practice as well. It likes long URLs rather than short ones. Maybe that's a bit of a differentiator. We used to think that, you you had to answer the question in one sentence immediately at the beginning of the article. You had to always add a table, but we've been playing around with it and we've realized actually, no, it can just be a normal piece of content that's really good. yeah, make sure that it's, you know, good quality for a person to read as well, not just for edit lems. So.
So the way to answer your question is it really starts with picking the right props. Everything else after that is kind of similar to SEO in that way.
Tom Rudnai (19:56)
Yeah, and I like that my thing is always like if you algorithms or whether you say if you say that search and then AI now is a kind of incremental improvement on search and it's very incremental. But if you want to kind of draw that line, they're just getting smarter and smarter at reflecting what humans want from content. So start with that is a pretty good rule. As they get smarter, you have to think less about optimization and more about quality.
Araminta Robertson (20:11)
Mm.
Tom Rudnai (20:19)
The two things I want to come back to in what you said. One is, so your hub and spoke model, which I think makes a lot of sense. You're kind of splitting off splinter pieces that answer very specific questions. One thing a lot of people talk about is FAQs. Have you found that it works dramatically, like significantly better to split those off as separate pieces of content rather than FAQs at the bottom of an article? And do you have a hypothesis on why?
Araminta Robertson (20:31)
Mm.
Mm.
Yes, we have found that answering, having multiple pieces does a better job than adding FAQs. So each piece will have FAQs as well. We like adding them because ultimately there's like millions of ways that you can ask a question. So might as well just add them. I think it's because, and this is me not being like an expert on this, know, but my theory is that just LLMs, when you have exactly the question, when you're answering exactly within a piece and providing some context, adding that table,
it just, it's easier for it to know, okay, you're answering the question here. When it's a separate page, it's like a lot easier for it to, crawl, find the URL, which is exactly what that question is, find the answer. It's just directly there. Whereas in FAQ, right, there's a lot more information. The URL is maybe shorter, different, all that kind of stuff. So we have found that, yeah, creating those shorter GPT articles makes a difference. And then the important part that I maybe didn't mention is you then want to track those prompts that you're.
targeting, we use peak, but profound does this. There's a lot of other tools that do this as well. And you can see like it's often within days, maybe even hours. Once you publish your GPT articles, you'll see visibility shoot up. And then the next question is, okay, visibility, great. Does it bring in conversions so far? Yes, we have, I have an article in this that I can share with you. And we have seen that when you start publishing more GPT articles and going after the right topics, you'll start to see an increase in inbounds as well.
which proves the theory that if you target one prompt, you will automatically start appearing for related prompts. You know what I mean? Like even if it's not exactly worded the same way. So that's why I don't worry too much about how you word your prompt. If you're going after the right sentiment, you'll start appearing for related questions as well. And we have found that client or companies that are quite early stage, maybe even too early for SEO, they're the ones who benefit the most from this because
LLMs are so much faster than Google at showing brands and all that. So yeah, it's a really exciting space.
Tom Rudnai (22:50)
Yeah, that does make sense. And I've got to put it in there. Demand Genius also does that as well as Peak and Profound, which do a fantastic job, but it's fine. No offence taken. No offence taken. The other things that I wanted to come back to, you mentioned starting at the bottom of the funnel, which I think makes all the sense of the world. And my one thing that we've known from our research where we were looking at top middle bottom of funnel and how AI behaves at each stage of this is that
Araminta Robertson (22:56)
Okay, perfect.
Tom Rudnai (23:16)
Retrieval only gets invoked at the bottom of the funnel. Really, a little bit at the middle of funnel, but mainly at the bottom. Like, have you had customers where you started at the bottom and got to the point of now we're working up and what have you found works then and how do you measure it? And because if retrieval isn't getting invoked, I think it's fair to say none of this is going to work at that point, right?
Araminta Robertson (23:34)
Yeah, I don't think so. I think people always underestimate. There's always that statistic out there. Oh, only 3 % of your customers are at the bottom of the funnel. I think first of all, OK, 3 % sounds very small, but it's so much bigger than we think it is. And then there's also a few things you can do in terms of targeting people at the middle of the funnel, but who are closer to the bottom. So for example, I call this damning demand.
Tom Rudnai (23:47)
Mm.
Yeah, they've crossed over.
Araminta Robertson (23:59)
So for example, say your product is like a credit card for a business credit card. Someone who types in AMEX fees, sorry, AMEX business, AMEX card, no fees. That is kind of, where are they in the funnel? I don't know, they want an AMEX card, but they want an alternative that's without fees, right? What they're writing is a pain point, really. And so there are ways to be creative.
and figure out the topics and the keywords that people are using to indicate they're looking for a solution like yours. And so I think people underestimate that section of their target market. It's a lot bigger than they think it is. So with GPT articles, have definitely, we're still doing lots of the bottom and like more middle, but aware. We haven't run out of that. For SEO, because the keywords are shorter, there's not like, it's a more straightforward process.
We have reached a point where you have to go more middle of the funnel and all of that. That definitely has happened. But for GPT articles, think because there's so many variations, you just never... And what's really cool with prompts as well is that you can almost do it the other way around from STO. But because we're able to do so many more and there's so many shorter articles, often we have like a hunch. Like, I feel like people looking up for this topic might like... But we're not sure. We'll go after it with GPT articles.
And then we'll see that it starts to work and people start to come in. Maybe even some of the GPT articles are ranking. And then it's like, okay, let's do an SEO article now targeting. So sometimes you get more data just by doing these shorter articles because it quickly tells you maybe which topics people are actually looking for. And you'd be surprised at how many, like search volume zero or NA on SEMrush, know, like barely any. And actually it's people are searching for it. I think a big reason is also potentially
people are starting to use Google more like Gemini already. So they're already starting, the long tail's already getting longer and longer as people start to personalize their searches. yeah, anyway, there's a lot to talk about here, as you can see. It's an interesting space.
Tom Rudnai (26:07)
That's very interesting
though, because one of the things that I get quite interested in is how SEO and AEO or GEO overlap, because I do think that, I love the opinion that they're quite different. The kind of SEO side is one component of GEO, but there are a lot of differences between the two. And I do think at times those things actually conflict. But that's a really interesting use case or example where the two go hand in hand, because I one of the big...
kind of edges that you can achieve in SEO is if you're better at finding the long tail keywords that are getting relevant traffic or high intent traffic for your product. That's huge. So if you can use GEO as a testing ground for that, that's cool.
Araminta Robertson (26:45)
Yeah, exactly.
Tom Rudnai (26:47)
The other thing that have to come back to that you said in yours, sounds ridiculous to me is long URLs Why on earth does AI prefer that?
Araminta Robertson (26:56)
I think it's just because it's more specific. there, again, I'm not the right person to break this down. There's lots of articles that do it better than I will right now. But anyway.
Tom Rudnai (27:05)
It
signals the content better.
Araminta Robertson (27:08)
Well, when it is rag, right? When it like researches on online, let's say chat GPT, what it does is that, and you can see it when it does the fan out queries, right? Is that it's going to take your prompt, it's going to break it down into keywords, right? And then it's going to search and see which URLs match that. So if your URL has all those words in it, in that prompt, it's going to go immediately there because it's like, okay, boom, you've, literally got all the words that,
Tom Rudnai (27:12)
Yeah.
Araminta Robertson (27:35)
that this person put into the prompt. So I'm going to go there first. And then it might go to some other ones as well. But because you're giving it exactly what it wants, and also in the meta description, always make sure to answer the question exactly as well, then it's going to, that's why it prefers that. It will prefer, I mean, it's an algorithm.
Tom Rudnai (27:55)
Yeah, we like to we like to personify the algorithm for people who spend a lot of time doing this you get to know it Okay, awesome, that's super interesting. Let's look at the other side of the coin Are there I always think it's a difficult space to wade through at the moment and there's a lot of kind myths and misinformation out there Are there anything that you see very commonly? Kind of recommended or talked about that when you've tried it. You simply haven't been able to see that that has an
Araminta Robertson (27:56)
Yeah. Yeah, exactly. Yeah.
Yeah.
Yeah.
I think the biggest thing is that there's just so much vague advice out there. People will be like, yeah, do chunking or make your headers like this or make it more conversational. And I'm like, people are already doing this. This is not actionable. I like a framework, like something repeatable that I can scale over time, which is why I love our GPT framework. So I think that's the biggest issue. It's not that it's wrong. It's just like unhelpful and people who are calling themselves GEO experts and then they just, you know.
give out this kind of generic advice, it's not very helpful. I don't know if there's anything that's specifically wrong. I mean, I'm always going to be against mass AI content. So doing like hundreds of pages a month, all AI generated, that's a very good way to get penalized by Google. So definitely don't do that. And any of the, what do they call it? Ranked to tank or something like that? Something like that. Oh yeah, ranked to tank. basically it's like these...
SEO experts or not experts actually, just people who are mass producing pages to rank very fast and then it tanks because they're doing too many. So you have to be very careful with that. I think there's a lot of stuff out there about Reddit, which I find it's just so hard to manufacture. Like I think either it's authentic or you're going to struggle. I've been testing a bit like LinkedIn articles and Medium articles haven't seen a big impact. We did do a test of digital PR.
versus GPT articles. And I think digital PR is good for the long term, just in general, because you're building authority. But I still haven't found a repeatable framework that works as well as GPT articles, honestly. Like this is one that we have seen work best. Maybe we haven't been as rigorous as, we're not like a research organization, so we haven't done like rigorous testing with others. And maybe there are others out there. But the answer is simple.
create more content that answers very specific questions. And I find it funny because like when LLMs came around everyone was like, my God, content's gonna die. What's the point of becoming content marketer? And I'm like, it's become even more important now content because that's the only way to appear in basically prompts. So yeah, anyway. I...
Tom Rudnai (30:34)
Yeah,
no, I say often we're in the kind of black cat a phase and some of it will work in the in the short term. think there'll be a cost. I'd add is I'm just so much echo to anyone listening. Don't go down the route of just spamming out AI content. You will create yourself and you'll get short term benefit and create a massive headache pretty soon because the thing that you have to do now is you have to have a plan for how you maintain all of the content because as soon as you start getting kind of clarity, consistency issues that causes a big, big problem.
Araminta Robertson (30:39)
Mm-hmm.
Yeah.
Yeah, mm-mm.
Yeah.
Tom Rudnai (31:04)
Good luck in picking that. The thing that you just kind of came on to, which was the kind of importance of content, I also completely agree with, I think the value of human input into content as we come out of the black cat AEO phase will be actually greater than it probably was before. But I guess with that in mind, like, and when you're hiring at Mint, how does the skills that you look for from a content marketer
change in today's world.
Araminta Robertson (31:37)
Yeah, it's interesting because, so first of all, we hire writers mainly and then eventually they become strategists. it always, everyone starts out as a writer because having, you need really strong writing skills to edit an AI. So first of all, we never do fully AI. We always do AI assisted and I don't think that will ever change because if not, it's just very, it's still not reached that point where you can fully rely on it. And also I think you always want an editor just to make sure that it sounds good.
So when we hire, we're mainly looking for writers, we're looking for very strong writing skills. And the main thing that we're looking for is clarity of thought, a lot of empathy, putting yourself constantly in the position of the reader, and curiosity. These are the three things that we're looking for. Someone's got these three things, then usually they'll be a strong writer. And also all the other things like a communicator, but that's like, you know, of anyone in a white-collar job, that's important.
So for content marketers in general, think an interesting shift that I've seen is that it's no longer enough to be a great writer. And this is why we are, people hire us and why we're getting more work now is because they're like, okay, great writing is great, but like, how do we connect it to revenue? And I am seeing even freelance writers who used to be booked up constantly and always with a lot of work are now like coming to us and saying, can we work with you? Because
I'm good at writing, but I don't know how to do the strategy side of things. And so I think that's the key thing that's changed is that being a great writer isn't enough yet now because a lot of companies will use AI. And instead, what is still going to be difficult, and I think for a long time because it's a strategy, is, OK, you're great at writing, but how does this turn into customers? And that's a skill that's more important now than ever. And it's also what we love to do, so good timing for us.
Yeah.
Tom Rudnai (33:36)
Yeah, I think it's interesting because I speak to a lot of freelancers as well in a similar position, but there I think a lot of them are struggling to make the transition into strategy. They bump up against a challenge where in-house marketers don't want that from them. They want to do it. But I guess it's again, it's probably that that's a lot of selection bias there, right? Because when you want just content written, maybe that's when you go to a freelancer and when you're looking for strategy, you're looking more for an agency or a...
Araminta Robertson (33:46)
Hmm. Yeah.
really?
potentially.
But I mean, there's still like definitely freelance content. I definitely think if you are a very strong content marketer, you can be freelance for sure. You can if you're good at the growth side, there's still so much demand for that. So I think there's definitely it's just like they might not think of that first. They might think, yeah, agency. But I think a freelancer. The thing is that if you're going to be a freelance content trash, you'll need a writer to help you out. It's a lot of work for one person.
And that's why I set up my agency. was like, I want us to do this, but I can't do this by myself. I need a team because writing is a lot of work or editing or whatever it is. So yeah, I think you can be a strong freelance content marketer, but you'd have to be very good at marketing yourself as that because if not, yeah, you're right. They will not, an in-house marketer might not think of that immediately.
Tom Rudnai (34:51)
Yeah, no for sure. And what, so, because you keep kind of actually preempting where I wanna go, which is gonna make everyone think that we've planned this much more than we have, I promise, we're not that competent. In terms of in-house marketers, so one of my kind of views, I think one thing that lot of in-house marketers I speak to struggle with at the moment is,
AEO is this massive new thing and they don't like it's something that isn't just a project It's a muscle that you have to build up a new skill that you have to learn but it's very difficult for them to hire at the moment because the general pressure is in the other direction in terms of headcount so I think a lot of them are caught between that like How much of the AEO job do you think there is a benefit to having in-house or
versus using an agency. And I know obviously you have a pretty strong bias on that one, are there elements that you think are best kept internal?
Araminta Robertson (35:47)
Hmm. It depends so much on the agency. Like if it's us, I would say, well, we only take on clients where we do SEO content and AEO. We wouldn't do AEO only. And we have had people reach out and say, can you just do the AEO side? And we'll say no, because it's such a small thing that like, it just doesn't make sense to take on a full on agency just for AEO. I think it only makes sense if you hire them to do SEO and other stuff as well. So yeah, exactly. It just doesn't.
Tom Rudnai (36:11)
Yeah. It's so intrinsically linked to content. I don't think you can separate that
Araminta Robertson (36:17)
No, no, no,
Tom Rudnai (36:17)
because that's a, it's not an optimization job in quite the same way as I think people want it to be.
Araminta Robertson (36:21)
Yeah.
Yeah, yeah. It requires producing content. And if we're to produce content, might as well do SEO as well. It just doesn't make sense to separate the two. So it depends so much on the stage, on the agency they're going to hire. It depends so much on where they are at in their SEO and AEO journey. So what I would say is if the agency, because we developed our own framework, if the agency uses that framework, then I would say pick the agency because they know what they're talking about. They're doing the right thing.
If it's an agency that is telling you to create tons of AI spammy content, I would say don't pick an agency. So this is why it depends so much on that strategy that that agency is going to implement. Having said that, whenever someone comes to us and says, can we just do AO, I just send them to our article that explains our process. And I'm like, you can do this internally. It's not that complicated. And I think that if you, and we actually did have this with a client of ours where the work
had reached a natural point where results were really good, there wasn't much left for us to do. And we were like, honestly, you guys are doing really well. We're just going to pass on our framework to you and you can implement it internally. And it's literally happened. a few months ago, well, actually a month ago, we checked in with them and they were like, yeah, AI, LLMs bring in 20 to 30 % of our amount now and the best quality as well. So that's why it's, again, based on where they are in their journey, it's hard to answer that question. I think if they're very early stage again, maybe agency helps with strategy.
framework, but if you're quite, you have quite a mature contact program doing well with SEO, I think you can do AO internally for sure. And, and I can send the article that I talked about and implement, implement the GPT framework and you're good. You're going to do well. So yeah, I kind of try to answer your question, but as you can see, it really depends on the situation.
Tom Rudnai (38:07)
Yeah,
it's a difficult question. mean, maybe a better way to frame it is like, what do you need from a client alongside what you do in order for to do in order for what you do to be as effective as possible? So if I'm an in-house marketer thinking of bringing in an agency, what do I also need to think about in order to maximize that investment?
Araminta Robertson (38:28)
Yes, so I'll talk about specifically for us rather than any agency because I don't know every agency will be different. I think the first thing is, it sounds really like simple, but having a good personal working relationship. So they're willing to trust us. They're willing to listen to what we have to say and rely also on our experience. I think if they want to decide on all the topics, for example, this will make it hard. understand where we're coming from. And that's why I always say like, you've heard us to bring us customers. Let us do that.
That's what we're going to do. So I think inherent willingness to trust and listen, I think that's a big one. I think the other one would be access to subject matter experts. I always say this before we start anything. I'm like, if your experts are super busy, no chance to talk to them, then I don't think it makes sense for us to work together because we really need access to the experts. And in the first two or three months, we'll need a bit more. And then after that, it tails off and not so much, but they do need to set aside a little bit of time.
And I think maybe ideally an internal person who is going to be dedicated to either content or who can be our point of contact and has the capacity to work with us. So especially the first two months, there's more reviews necessary to make sure that we're getting the product correctly, we're describing it correctly, and to make sure that we're getting the tone of voice correctly, all of that. Someone who can spend a little bit of time the first two or three months just to educate us and train us as well. And then just the last one, this is just in general.
I think it depends on the maturity of the company as well, but strong positioning, strong USPs really helps us. So if it's definitely true that earlier stage companies, might be changing their positioning every few months, that can make things complicated because then we're not sure, like, what do we say in the article? What are your USPs? If those keep changing, that's not good. Also for positioning, like the keywords we're going to go after, that keeps changing. That's going to make things tricky. So think having a clear, like positioning, branding and all that, that's just, that's not just for us, for anything in general.
any agency or freelancer you work with that will make life easier.
Tom Rudnai (40:31)
was just
the job of content is to communicate and kind of amplify the message or positioning a product, right? If the product is crap, then there's only so much you can do. You can't put lipstick on it. Yeah.
Araminta Robertson (40:37)
Yeah.
Or if it keeps changing as
well, or if the priorities keep changing. But more mature companies, doesn't tend to happen. Like that's once per year, that's fine. Once every six months even, that's okay. every month it's like, okay, well, makes things tricky.
Tom Rudnai (40:55)
or you'd hate working with me. Final question I want to ask you, just because it's something that I've been thinking about lot recently and I'm really interested in, Chad GPT launching ads, do you have any kind of take or hypothesis on what impact that might have on the kind of organic, AEO side of what we both do?
Araminta Robertson (40:58)
Okay.
Yeah.
Well, I haven't really, to be honest, looked into this too much. I think because first of all, they have said that it shouldn't change responses. That's what they've said. In reality, we'll see what happens. But if what they say is correct and true, then truly it doesn't change anything. So we kind of to see what happens. See what happens there. And then also only free users will have ads, right? So paid.
Tom Rudnai (41:21)
Hmm.
Araminta Robertson (41:42)
head users won't. So that's also something to think about. think if you're targeting people who, because the reason we primarily focus on ChatGPT is because it still is the tool that has the highest number of users, but that is decreasing. I know Gemini is catching up, so that's also something to keep into account. yeah, so far I'll just say it hasn't changed anything because it seems that it shouldn't bias or change the answers. If it starts to change the answers, then that's when I'll, you
then we maybe need to change something or look into it. But I think it's too early to say. I haven't really got any thoughts on it. Have you got any?
Tom Rudnai (42:18)
I have thoughts but not conclusions. I think I'm very interested in how it potentially opens up another lever for an AEO strategy, right? So how can you find ways to use it? Yeah, if you have a budget that you can invest in ads, then maybe you can find a clever way of finding, I'm going to use the horrible word that I hate using, you'd find a synergy between your organic and kind of paid. The other thing I'm just curious to see,
Araminta Robertson (42:21)
Mm-hmm, yeah.
Yeah, that's true. With ads.
Mm-hmm. Yeah.
Tom Rudnai (42:48)
In order to run ads, ChurchyBT has to give you a way to explicitly communicate who you are, who you're for to it, right? Because it needs to make sure that it's nailing the ad content versus what you want. Now, most of what we both do day in and day out is help brands communicate, inexplicitly, implicitly, who they are and who they're for. So I'm curious whether that presents an opportunity to educate the model much more explicitly in a way that actually...
Araminta Robertson (42:55)
Mm.
Yeah, that's interesting.
Tom Rudnai (43:16)
carries over beyond what they're doing in Adverts, it's going to be interesting to see how they
Araminta Robertson (43:19)
That's a point.
Tom Rudnai (43:20)
implement that. But I don't know. Yeah, I think that's
Araminta Robertson (43:22)
Yeah, that's possible. I don't know, it's very
interesting. We'll see. think definitely once there's more, I think it's only live in the US at the moment, but once it's more worldwide and stuff, I think it'll be really interesting to experiment and test with that because yeah, you're right. It could have other benefits.
Tom Rudnai (43:40)
Yeah, for sure. that's the, I don't know about you, but that's what I love so much about working in this space. It's so much fun because it just like changes every day. And you just get to spend so much of your day figuring stuff out.
Araminta Robertson (43:51)
Yeah, definitely.
Tom Rudnai (43:52)
Cool, right, before I let you go, let's do a couple of quick fires that I like to do at the end. So the first one which I'm going to give you is an AI use case or an AI tool that you absolutely love.
Araminta Robertson (44:05)
What I've been getting into recently, anyone who follows me on LinkedIn will have seen is building little AI tools with lovable. I'm completely addicted. My Sunday afternoons are spent building these tools. And I want to say specifically not for public use, for internal use. So for me, my team, and our clients, the whole doing it public, you need so many other things in place, compliance, all these things.
start charging people money. No, I think that's too complicated, especially for us. We just use it for internal. So I built an internal linking tool that's actually useful that we actually use. And we've built a bunch of other tools for clients that are really exciting. that's, I've been addicted to that lately.
Tom Rudnai (44:47)
I'm exactly the same I've got two groups with Claude code and the stuff that I can build with it in like a weekend is I mean it's awful with my weekends now I just sit in like a dark room coding but I like yeah I keep making fun of our engineers like I'm so good at this they're building me little playgrounds where I can like do stuff and like right just stay over there but don't touch the actual product but I can kind of mess around in staging and stuff like that
Araminta Robertson (44:57)
I know, it's not good.
Mmm.
That's cool. Yeah, that's true.
Tom Rudnai (45:12)
Yeah, cool. I love that. And then for you personally, is there a skill or a trait that you have that you think has really helped you in your career?
Araminta Robertson (45:23)
I think this could be, it's a positive and a negative, which is being very stubborn. Which I think that just when things aren't going well, just being like, I don't care, I'm going to keep trying. Like even if everything is telling me to stop or everything is like, I think that's what's allowed me to keep going when things can get tough. But there's also downside to being stubborn and my husband will always say that. Which means that sometimes it can be...
You have to also know when to quit and when to stop. But yeah, I think it's benefited me a bit as well. bit of the two sides of it.
Tom Rudnai (45:59)
Yeah, I think you could probably find some people who would say that I have a similar trait. And if I, maybe, how we can rephrase this one for you, but I always find it quite fun. If I were to approve your plan A budget request and maybe, you can go at this in two ways. The one that your board would never be stupid enough to let you have, let's pretend, I don't know if you have a board, pretend that you have a board. Or the thing that just you could never justify, but think would be really fun to do.
Araminta Robertson (46:03)
Yeah.
Tom Rudnai (46:26)
or maybe that a client would never give you the budget to do if that's more fun for you, what would you do?
Araminta Robertson (46:33)
So if we're talking about Mint Studio specifically, I think I am so bullish or excited about events. If I could, I would hire a bunch of fantastic event marketers and just do lots of events. I love being one of the organizers and attending them. I don't like being the only organizer because they're very stressful. But hiring a bunch of event marketers and just doing lots of things like dinners, workshops, side events, everything, I just love it.
Tom Rudnai (46:46)
Hmm.
Araminta Robertson (47:00)
It's so much fun, especially when we spend so much time in front of a laptop, being able to do in-person stuff with pretty smart, interesting people. I, yeah, I love it.
Tom Rudnai (47:09)
Yeah, I've had the same thought actually quite recently because I think it's something that works really well at the moment, but I'm just not like an event organisery person. I've been really bad at organising stuff for my birthdays. I barely ever invite my friends over. I'm not the host. But I think it's a very effective thing to do at the moment. And then final one is always a bit more fun. I always think on podcasts people sound really smart and I think it's nice for people to realise that they're often actually not.
Araminta Robertson (47:17)
Yeah.
Tom Rudnai (47:39)
Like what was the biggest screw up that you have or that you remember in your career?
Araminta Robertson (47:45)
I'll just say the most embarrassing one because I haven't really, I'd have to think about it for a lot more longer to think about it, to really have an answer. I think two embarrassing moments stand out. One is on stage at one of our conferences, I think. I accidentally called someone by their company name rather than their actual first name. I was exhausted and it was the end of the day. Luckily, everyone laughed, but I felt mortified. And then I think at a different event as well, I...
Tom Rudnai (48:09)
You
Araminta Robertson (48:14)
completely forgot someone's name. I knew this person well, I knew her name, not a problem, but on the stage I actually said, I'm so sorry, can you remind me your That was very embarrassing as well. So it happens, yeah, when you do a lot of events, things are gonna go wrong and yeah, that's happened, unfortunately.
Tom Rudnai (48:30)
Yeah, it's that thing.
Do you have like the lights? And they're so disorienting. It's so easy to just have like a mental block in that moment. Can you share the name of the company? What did you think this person's name was? Okay, yeah, fair enough.
Araminta Robertson (48:36)
Mm-hmm.
No, I don't want to embarrass them even further or embarrass myself even further. On that stage,
that was enough. No, I don't need to do it anymore. But luckily they took it very well, really chill. And I'm still in touch with her and everything and very nice people. to me, I'm always like, especially someone's first name is so important. People always mess up my name. Like, I know how important it is to get that right. So it was it was a weak moment for me. But I try to not let that happen again.
Tom Rudnai (49:11)
Yeah, I mean, luckily that's not one that I empathize with as a Tom. don't have to I don't really have to come across one too much. That's not a battle I have to fight my son and that one I get. And then just before I you go one recommendation for everyone. So a book, a podcast, a thought leader, something you love that people should check out.
Araminta Robertson (49:16)
Yeah.
Yeah.
Yes, so I love April Dunford, huge fan of everything she creates, especially her book, Sales Pitch. I think if you're selling anything, whether you're a freelancer, agency, or if you're an in-house marketer trying to, maybe if you're product marketer especially, and you're trying to understand sales more. What a great book. If you're in the agency world, Blair Enns, David Baker, love them. Anything they do is fantastic. And then solely marketing, I'm getting into Emily Kramer. I think that's how you say her last name, Kramer, I'm not sure.
Tom Rudnai (49:54)
Yeah, market one.
Araminta Robertson (49:56)
Yeah, so good. Wow. It's so nice to have like a prolific creator who's also just so great and really actually genuinely useful stuff. I loved her sub stack on how to run dinner dinners, like executive dinners or founder dinners, called them so good. So I think, yeah, those three or four resources there are really great.
Tom Rudnai (50:19)
Great, and then I normally at this point say anything that you want to promote that you're doing, but I guess I can do that one for you probably and say go you go check out Mint Copywriting, but I'll give you the space to do that yourself. Sorry, anything that you'd like to promote that you're doing.
Araminta Robertson (50:28)
Studios, yes. Sorry, what's question?
promote?
Well, we just relaunched our podcast. It was under the Fintech Marketing Hub brand. Now it's under our own brand. So if you'd like to check that out, under, it's called Fintech Marketing Lab. And we've got some fantastic guests, including yourself, who are coming on. So I'd love if you could check that out.
Tom Rudnai (50:54)
feel free to skip the me episode. I will not be offended. I don't blame you. Cool, Armenta, I'm gonna let you move on with your day, but thank you very much for joining us. think one, like, I always love, I think anyone who's trying to like sort through who to listen to in the space that we're in, the AI search space, ask them questions. And if they don't ever say they're not sure, then just leave, don't trust them. And you said you're not sure like five times in this switch.
Araminta Robertson (51:18)
Yeah,
Tom Rudnai (51:19)
You could
Araminta Robertson (51:19)
definitely.
Tom Rudnai (51:19)
frame as a negative, but I think that's a fantastic thing. It's a reason for people to listen. So thank you for joining. been fun.
Araminta Robertson (51:29)
Thanks a ton for having me, it's been a pleasure.
Tom Rudnai (51:31)
Bye bye.
Hello and welcome to another episode of Demand Geniuses. I'm Tom Rudnay, same as I was last time. And with me today, I'm going to get straight into introducing, rather than do a long preamble, and introduce Araminta Robertson. Araminta is the founder and CEO of Mint Copywriting Studios. First of all, hello.
Araminta Robertson (00:32)
Hi Tom, lovely to be here.
Tom Rudnai (00:34)
Thank you for joining. guess rather than me butcher an introduction into you and your background, do you want to join a movie? Just give us the headlines and tell us a bit about Mint as well and what you do there.
Araminta Robertson (00:43)
Yeah, definitely. So yes, I'm the founder. call myself a managing director rather than CEO because we don't have a C-suite. it's managing director of Mids Studios. We're a content marketing agency for fintech and financial services. And we help these companies turn content into measurable pipeline. And we've worked with companies like Persona in the US, GBG, ClearBank. These are some of the names in the fintech and financial services space.
Tom Rudnai (01:14)
Nice, and why did you pick when you kind of started that then? Why specifically FinTech and financial services? What was kind of uniquely interesting about that for you?
Araminta Robertson (01:22)
Yeah, I mean, there's a whole backstory to it. think it all started when at the age of 18, I finished high school and I realized that I didn't know anything about finance as many 18 year olds do because we're not taught about taxes and mortgages and how to invest in high school. And I realized that and I also realized if I did learn about these things, I would probably set myself well for the future. So I started a blog.
And I would read a lot of these typical financial books, you know, the Intelligent Investor, Richard Perda, all of those books. And then I'd read them and then I'd write about them in this blog. And it was kind of aimed at other kind of young adults, like, let's figure this out together kind of thing. I was living in Barcelona at the time, which is where I kind of grew up. And I met a fintech company at a networking event. That's a kind of a business event. And they saw the blog and they liked it. So they hired me as their kind of...
content marketing person. And that's kind of how it all got started. So it really started from a passion project. And then as I did more and more of it, I just loved it. I just loved how when you learn about personal finance or financial topics, it gives you control because you start to understand how the world works. And I also love that now obviously we do more B2B rather than personal finance, but I just love how payments and finance are kind of like the arteries or the blood vessels of the world, right? Like without that, you've got nothing.
Tom Rudnai (02:31)
Hmm.
Araminta Robertson (02:44)
you can't, the world doesn't work right without payments. So I just love how it's such a unsexy and yet so important part of our economy. And also the people are fantastic in this industry. They're so generous and lovely. And I think it's because it's an unsexy industry. It attracts people who genuinely want to do this or genuinely are interested in this topic. So, yeah, that's a bit of a backstory into how I got into it.
Tom Rudnai (03:10)
So it's interesting though, because it is unsexy, there's nothing like it, it doesn't sound fun, FinTech, but it's like the operating system of the world and I can see that going into the wheeze on how all of that stitches together must be quite interesting, revealing and probably quite helpful just from a personal perspective. I'm terrible at things like that. Nice, and then one question I always love asking early on, because I think it's really interesting hearing different people's answers, it's like...
Araminta Robertson (03:15)
Mm-hmm. Yeah, exactly.
Mm-hmm. It's exciting.
Yeah.
Yeah, exactly.
Tom Rudnai (03:36)
When you look back at the stuff that you've done so far, is there a step in your career that you would say was the most formative or like a moment that you felt that you really like clicked and you were like, I know what I want to do and I know what I'm good at.
Araminta Robertson (03:48)
Yeah,
I guess there were multiple, but I mean, just because it was the beginning of it all, guess, you when I was 18 and I read a book, I started reading personal finance books and I think that's, at that moment I realized, okay, this is what I want to do. I think another one was potentially when I went from freelancer to agency, it was an interesting moment where I had a current client as a freelancer who was saying like, okay, where are the conversions? Like traffic is great, but we want customers.
I was reading a book called They Ask You Answer, which is fantastic. Like I call it like the Bible of content marketing. And at the same time, I was doing a course by growing Convert, a fantastic content marketing agency on how to generate conversions with content. I did all those things, like all those things were happening at the same time. And that's when I, it all clicked and I was like, perfect. Now I know what I'm going to do with this agency. We're going to focus on that because I can see the problem there. So yeah, multiple, but maybe those two are the biggest ones.
Tom Rudnai (04:42)
Okay, and then I guess I think it's always one thing I always want to do with this podcast is make sure that I frame any advice that we give because I think there's advice out there everywhere and the really hard thing is understanding which advice to listen to. So before we get into some of the stuff that we want to talk about, which is that content, AI search, things like that. What's unique about FinTech specifically from from like the content marketing conditions.
that would influence, let's say if I'm someone listening to this podcast, whether or not I should follow your advice. In what conditions should I follow every bit of advice that you're about to give us?
Araminta Robertson (05:14)
Well, to be honest, first of all, I'm not a huge fan of, like advice is so contextual. It's very hard for me to give good advice without knowing the situation of each person. So I don't like giving personal advice from business advice. would, that's why I take it with a grain of salt. It depends a lot. I can just share what's worked for us and what we've seen. So that's just one thing to say. So what is unique about content marketing and fintech and financial services? The first one is that the stakes are higher. So, you know, when you're opening,
Opening a social media account, opening a buffer account like a SaaS tool is not the same as opening a business bank account or even choosing a banking as a service partner. The stakes are way higher. And so from a content marketing perspective, means especially SEO, that means you're in the YMYL bucket, which means your money or your life. So Google classifies anything around those topics as your money or your life topics. And the reason is because if you give the wrong advice or the content isn't correct, the consequences are much higher.
Right? Telling someone to invest in crypto coin and lose all their life savings is a much bigger deal than open this SaaS tool that whatever costs $5 a month or whatever, you know, so the stakes are higher. And so that means that your content needs to be more rigorous and more like you have to be more careful about what you're writing about. think that's number one. Number two is that it's more complex. Whether you're B2C or B2B, writing about payment infrastructure is more complicated than talking about
a recipe, would say. mean, I don't want to put down people who write recipes, but it's definitely simpler than giving instructions on how to build the table or whatever it is. So it's more complex, which means that, again, the content, you have to be a very strong writer to know how to write in a way that can make that complexity easy to understand. And then the last thing I would say is, and this is kind of linked to the two others as well, is like expertise and trust.
especially in the B2B sector, know, people are looking for content that is more advanced, more technical, as I said, complexity, but they also want it to come from a place of expertise, right? And often you're writing about topics that, you know, maybe haven't been written about before, are, again, the stakes are high, so you want to make sure you're writing or you're reading from someone who knows what they're talking about. And all of that combines together to build trust, because in financial services, trust is super important.
way more than, as I said, other industries potentially because the stakes are higher. So I think this is why content marketing comes in so hand to hand with financial services is because trust is so important, expertise is so important. And the best way to build that trust is to create really like high quality content that explains that complexity that comes from a place of expertise.
And yeah, it's one of the key distribution channels for a financial services company. That's what I say is unique about it.
Tom Rudnai (08:15)
Super interesting though, because what you're describing is to me, a space like, it's the hardest possible or the most extreme conditions in which to market, right? And I think of like, what is it, like diamonds, like they're built under extreme pressure and that's how you form a diamond. That's how you, it's probably a very good breeding ground for a good content marketer because what you've described is you have to be able to boil complexity down to something that people can understand. And there is a...
Araminta Robertson (08:38)
Yeah.
Tom Rudnai (08:42)
even higher bar of expertise and trust at a time where that's quite a hard thing to develop because I think there's a shortage of that in digital spaces at the moment. I want to come back to the your money your life thing is that a is that an official is that that's a google recognized thing that they have a category of this and their algorithm works in a different way.
Araminta Robertson (08:48)
Mmm.
Yep. Mm-hmm. Yep.
Yeah, and it's not just finance. So healthcare is in that bucket. You know, other sectors where, again, the stakes are higher. If you give someone bad advice on how to resolve, I don't know, specific disease, that's bad. It could affect their life. Same with financial services. And I always find myself attracted to those industries because, again, like the stakes are higher, means your work is more important. It's more exciting. It's like, can really, the impact that you can make on the person who's reading or a customer is so much higher.
And that's really exciting to me. yeah, YMYL is a term that Google uses, yes. And I wouldn't say, does the algorithm change? mean, I'm not a technical SEO expert, but it's more rigorous. They'll say, know, EEAT, the framework that Google, yeah, they'll say that it's even more important for those sectors, basically. That's what they'll say.
Tom Rudnai (09:53)
does that impact when you're kind of building a content strategy? I'd imagine that has quite a big impact on how you approach it because I assume it makes it little bit more of a slow burn to kind of start getting things ranking and building up reputation because there is a higher bar for them to kind of consider you credible. Is that something that factors into the way you build a strategy or?
Araminta Robertson (10:09)
Yeah.
Well, I would just say that it just, it means that your content is going to be harder to produce and your standards are going to have to be a lot higher. You also like have compliance and everything that you have to deal with. So, for us, I mean, we, especially we only work in financial service fintech. So we know this really well. And, and, but, but even, guess, if we working with a non-fintech, we would still have high standards. And because, you know, especially in B2B, the people you're targeting, you want them to take you seriously. And if you just write a generic.
lame, boring piece, like they're not going to take you seriously. one part of our process, for example, that's a key part of this is that we interview experts, right? So especially when we first start working with a new client, we'll spend a lot of time interviewing the experts on their team, maybe for each article, but then eventually as, you know, as we repeat certain topics and stuff, that's no longer necessary. But we spend a lot of time interviewing experts, understanding the product, understanding the pain points to make sure we're writing about the right topics.
but also that the content itself is written for their level of knowledge. It's expert based. It's as complex as necessary. And it's always a nice one. A client's like, wow, this is really good. And I'm like, yes, because we were able to do an expert. That wouldn't be possible with just desk research. And it's funny because a few years ago, sometimes we'd get pushback from clients. They'd be like, come on, can you not just do desk research? You're taking up time. We no longer get that pushback. And I think it's because of AI. Like people can see that.
If you don't interview an expert, it's just going to be AI slop. Like what's the point? Why read it from you? If you're to say the same as everyone else. And so I think introducing the expert into that process is just so important, especially for, for a FinTech and financial services.
Tom Rudnai (11:55)
interesting because you've preempted my question a little bit there which was going to be what you've described is that essentially creating content in your space is necessarily slower right you have to be more diligent you have to be more detailed and I was like is there a bit of a conflict there between what it seems like everyone is kind of obsessed with at the moment which is speed efficiency cutting headcount doing more with less all of those phrases that every marketer probably wakes up at night terrified of but you but you found the opposite it sounds like
Araminta Robertson (12:06)
Yeah.
Well, mean, it depends again on the product and on the client. So for example, for clients where the product is more straightforward, think of a business bank account or a money transfer service. Definitely, and even for the complex clients, there is a place for AI. So we'll talk about later, I guess, our AEO framework, GPT articles. Those are AI assisted. And we're right now developing a tool to help us with one part of the process, even for content creation. So there's definitely a place for that. But I think that...
you know, as long as you include the experts and make sure that the quality is high. I mean, you do what you want, right? Whether it's AI or not, like that's ultimately up to you. And I think it has just because it's more complex. Yes, usually, you know, that means it takes a little bit longer, but we're a lot more focused because we're outcome focused. So X number of leads, MQLs, SQLs. It doesn't really matter how many pieces of content you produce as long as you're hitting
as long as that's growing and you're hitting your goals, ultimately it doesn't really matter, right? So we might produce less or fewer pieces of content than another agency, but because we're tracking the metrics that really matter, ultimately it doesn't really matter, right?
Tom Rudnai (13:35)
Yeah, it's interesting. it makes me, I guess there's a bit of a selection bias here as well because obviously people who are talking to and engaging with an agency like yours are the people who are interested in doing good quality content and investing in that. But it doesn't contradict. So we did a big study recently where we basically used our AI to analyze like 50,000 pieces of.
Araminta Robertson (13:46)
Mm-hmm.
Tom Rudnai (13:54)
content in B2B Fintech specifically, because we were quite interested in understanding how that had evolved over the last five years. It was really cool because actually one of the things you could do is like tell a story of the industry and how the industry as a whole is evolving just from how the narratives and the storylines change. But one thing that we did notice was there was a huge increase in content velocity. So the speed of output had increased massively, which obviously makes you fear that AI slop. But we actually one of the cool things about the way our product works and the way that we were doing the study was we were looking for quality markers as well.
Araminta Robertson (14:13)
you
Tom Rudnai (14:24)
and they were also on the rise, which kind of reinforces what you're saying, is that as people have increased that velocity, I'd imagine there are some notable exceptions to this, but generally speaking, quality was still improving.
Araminta Robertson (14:34)
Yeah, there are ways to do that.
And we're doing them. there are, like for example, if you've got already a large body of content published on your website, like that's huge. Like it's so much easier to create good quality AI generated or AI assisted always. Don't do AI generated, AI assisted. If you already have lots of content, it will be so much easier. And that's why brand new clients that don't have a lot of content, that's where a human writer is a lot more involved because there's not much to go from.
But if you already have lots, we do a lot more AI assisted with the client. We're like, OK, cool. You've got some really good stuff here. So it also depends on where you are on the journey.
Tom Rudnai (15:11)
Yeah, well, because that's something we've been playing around a lot with is so we have our kind of information game framework, which I won't go into now. But trying to experiment with how we can use AI to create the highest possible level of information game within a piece of content. And we found that if you just use AI from scratch, it's rubbish. It's not going to get it done. If you feed it with your perspective, you can get our kind of level one, which is interpretive game, so slightly new slot on that. You can just give it your brand voice and do it. Anything bigger?
Araminta Robertson (15:29)
Yeah.
Tom Rudnai (15:40)
It just will not get to but what you can do is if you've got great research pieces and you're giving it quality sources that it can infuse in with other things then it can do a fantastic job and I guess you're saying for a lot of brands who already have that it's a big bonus
Araminta Robertson (15:55)
Yeah, yeah, I mean, I always say that your output will be as good as your input. So if your input's really good, so will your output. If we take on a new client and maybe they've got a lot of content, but it's not great, we'll still need to do a bit of human-written content for a bit, and then we'll do more AI-assisted. But yeah, if your input's really strong, then that's where AI can really be fantastic.
Tom Rudnai (15:58)
Hmm.
Yeah, that makes sense. Let's talk about the content itself a little bit then. So, I'll start with what in your experience makes a piece of content NLM optimised?
Araminta Robertson (16:27)
Okay, so to answer this question, I'd have to backtrack a little bit. So around October, September, 2025, we developed a framework to help companies basically influence LLM visibility, get recommended more often by LLMs. And so the way this GPT, we call it the GPT framework, and the way this works is you start off by deciding the bottom of the funnel topics. Like this is our whole thesis, start with the bottom of the funnel, because there are,
the people who are ready to convert, these are people who are aware they have a problem and are actively looking for a solution. already have money in their hands, they're ready to close. Start with those people and then work your way up the funnel. That's what we do for SEO and content. We have taken that also for LLM visibility, optimization, and all of that. The way that framework works is you start off by deciding the topics that are again bottom of the funnel, and it can be the same ones as SEO ones as well. Then once you've decided on the topics,
put together kind of prompts that you think people might be typing into Chachapi, TGEM, and I, cetera. So you can do, for example, for one topic, maybe 10 prompts. How do we come up with a prompts? Honestly, it's by understanding your target audience, putting yourself in their shoes. What do think they'd be typing? So it is a guess. Yes. Because I mean, there's no way that you can, there's like millions of variations. You're never going to get it exactly right. So you put together kind of maybe prompts. And then after that you.
create short articles, maybe 500 to 800 words, answering each of those prompts. So the way I think of it, like a bit of a hub and spoke model. At the center, you've got your SEO piece that's very in depth, that goes into detail about X topic. Let's just say open banking APIs, for example. And then you create your GPT articles. These are short articles, and they answer different questions about open banking APIs. So for example,
tell me an Open Mac API that does CVRP or what's an Open Mac API that works across Europe and not just the UK. So all these different variations of that topic and then you answer them and make sure they're bottom of funnel. That's the main thing, right? So I think the number one way to ensure that your piece of content is LLM optimized is to first go after the right prompts. I think that's 80 % of the work. If you go after the wrong ones, then there's no point because say you go after a topic like
What is open banking? ChachiBT will just give you a straight up answer, but it's not going to mention any brands. So there's no point in really targeting it. It's going to use your content. It probably won't even source you. So start with the bottom of the funnel. And then after that, honestly, we've been experimenting and testing. But I think the main thing is make sure you answer the question in the article. It likes a few different things like tables, headers, all stuff that is good SEO.
practice as well. It likes long URLs rather than short ones. Maybe that's a bit of a differentiator. We used to think that, you you had to answer the question in one sentence immediately at the beginning of the article. You had to always add a table, but we've been playing around with it and we've realized actually, no, it can just be a normal piece of content that's really good. yeah, make sure that it's, you know, good quality for a person to read as well, not just for edit lems. So.
So the way to answer your question is it really starts with picking the right props. Everything else after that is kind of similar to SEO in that way.
Tom Rudnai (19:56)
Yeah, and I like that my thing is always like if you algorithms or whether you say if you say that search and then AI now is a kind of incremental improvement on search and it's very incremental. But if you want to kind of draw that line, they're just getting smarter and smarter at reflecting what humans want from content. So start with that is a pretty good rule. As they get smarter, you have to think less about optimization and more about quality.
Araminta Robertson (20:11)
Mm.
Tom Rudnai (20:19)
The two things I want to come back to in what you said. One is, so your hub and spoke model, which I think makes a lot of sense. You're kind of splitting off splinter pieces that answer very specific questions. One thing a lot of people talk about is FAQs. Have you found that it works dramatically, like significantly better to split those off as separate pieces of content rather than FAQs at the bottom of an article? And do you have a hypothesis on why?
Araminta Robertson (20:31)
Mm.
Mm.
Yes, we have found that answering, having multiple pieces does a better job than adding FAQs. So each piece will have FAQs as well. We like adding them because ultimately there's like millions of ways that you can ask a question. So might as well just add them. I think it's because, and this is me not being like an expert on this, know, but my theory is that just LLMs, when you have exactly the question, when you're answering exactly within a piece and providing some context, adding that table,
it just, it's easier for it to know, okay, you're answering the question here. When it's a separate page, it's like a lot easier for it to, crawl, find the URL, which is exactly what that question is, find the answer. It's just directly there. Whereas in FAQ, right, there's a lot more information. The URL is maybe shorter, different, all that kind of stuff. So we have found that, yeah, creating those shorter GPT articles makes a difference. And then the important part that I maybe didn't mention is you then want to track those prompts that you're.
targeting, we use peak, but profound does this. There's a lot of other tools that do this as well. And you can see like it's often within days, maybe even hours. Once you publish your GPT articles, you'll see visibility shoot up. And then the next question is, okay, visibility, great. Does it bring in conversions so far? Yes, we have, I have an article in this that I can share with you. And we have seen that when you start publishing more GPT articles and going after the right topics, you'll start to see an increase in inbounds as well.
which proves the theory that if you target one prompt, you will automatically start appearing for related prompts. You know what I mean? Like even if it's not exactly worded the same way. So that's why I don't worry too much about how you word your prompt. If you're going after the right sentiment, you'll start appearing for related questions as well. And we have found that client or companies that are quite early stage, maybe even too early for SEO, they're the ones who benefit the most from this because
LLMs are so much faster than Google at showing brands and all that. So yeah, it's a really exciting space.
Tom Rudnai (22:50)
Yeah, that does make sense. And I've got to put it in there. Demand Genius also does that as well as Peak and Profound, which do a fantastic job, but it's fine. No offence taken. No offence taken. The other things that I wanted to come back to, you mentioned starting at the bottom of the funnel, which I think makes all the sense of the world. And my one thing that we've known from our research where we were looking at top middle bottom of funnel and how AI behaves at each stage of this is that
Araminta Robertson (22:56)
Okay, perfect.
Tom Rudnai (23:16)
Retrieval only gets invoked at the bottom of the funnel. Really, a little bit at the middle of funnel, but mainly at the bottom. Like, have you had customers where you started at the bottom and got to the point of now we're working up and what have you found works then and how do you measure it? And because if retrieval isn't getting invoked, I think it's fair to say none of this is going to work at that point, right?
Araminta Robertson (23:34)
Yeah, I don't think so. I think people always underestimate. There's always that statistic out there. Oh, only 3 % of your customers are at the bottom of the funnel. I think first of all, OK, 3 % sounds very small, but it's so much bigger than we think it is. And then there's also a few things you can do in terms of targeting people at the middle of the funnel, but who are closer to the bottom. So for example, I call this damning demand.
Tom Rudnai (23:47)
Mm.
Yeah, they've crossed over.
Araminta Robertson (23:59)
So for example, say your product is like a credit card for a business credit card. Someone who types in AMEX fees, sorry, AMEX business, AMEX card, no fees. That is kind of, where are they in the funnel? I don't know, they want an AMEX card, but they want an alternative that's without fees, right? What they're writing is a pain point, really. And so there are ways to be creative.
and figure out the topics and the keywords that people are using to indicate they're looking for a solution like yours. And so I think people underestimate that section of their target market. It's a lot bigger than they think it is. So with GPT articles, have definitely, we're still doing lots of the bottom and like more middle, but aware. We haven't run out of that. For SEO, because the keywords are shorter, there's not like, it's a more straightforward process.
We have reached a point where you have to go more middle of the funnel and all of that. That definitely has happened. But for GPT articles, think because there's so many variations, you just never... And what's really cool with prompts as well is that you can almost do it the other way around from STO. But because we're able to do so many more and there's so many shorter articles, often we have like a hunch. Like, I feel like people looking up for this topic might like... But we're not sure. We'll go after it with GPT articles.
And then we'll see that it starts to work and people start to come in. Maybe even some of the GPT articles are ranking. And then it's like, okay, let's do an SEO article now targeting. So sometimes you get more data just by doing these shorter articles because it quickly tells you maybe which topics people are actually looking for. And you'd be surprised at how many, like search volume zero or NA on SEMrush, know, like barely any. And actually it's people are searching for it. I think a big reason is also potentially
people are starting to use Google more like Gemini already. So they're already starting, the long tail's already getting longer and longer as people start to personalize their searches. yeah, anyway, there's a lot to talk about here, as you can see. It's an interesting space.
Tom Rudnai (26:07)
That's very interesting
though, because one of the things that I get quite interested in is how SEO and AEO or GEO overlap, because I do think that, I love the opinion that they're quite different. The kind of SEO side is one component of GEO, but there are a lot of differences between the two. And I do think at times those things actually conflict. But that's a really interesting use case or example where the two go hand in hand, because I one of the big...
kind of edges that you can achieve in SEO is if you're better at finding the long tail keywords that are getting relevant traffic or high intent traffic for your product. That's huge. So if you can use GEO as a testing ground for that, that's cool.
Araminta Robertson (26:45)
Yeah, exactly.
Tom Rudnai (26:47)
The other thing that have to come back to that you said in yours, sounds ridiculous to me is long URLs Why on earth does AI prefer that?
Araminta Robertson (26:56)
I think it's just because it's more specific. there, again, I'm not the right person to break this down. There's lots of articles that do it better than I will right now. But anyway.
Tom Rudnai (27:05)
It
signals the content better.
Araminta Robertson (27:08)
Well, when it is rag, right? When it like researches on online, let's say chat GPT, what it does is that, and you can see it when it does the fan out queries, right? Is that it's going to take your prompt, it's going to break it down into keywords, right? And then it's going to search and see which URLs match that. So if your URL has all those words in it, in that prompt, it's going to go immediately there because it's like, okay, boom, you've, literally got all the words that,
Tom Rudnai (27:12)
Yeah.
Araminta Robertson (27:35)
that this person put into the prompt. So I'm going to go there first. And then it might go to some other ones as well. But because you're giving it exactly what it wants, and also in the meta description, always make sure to answer the question exactly as well, then it's going to, that's why it prefers that. It will prefer, I mean, it's an algorithm.
Tom Rudnai (27:55)
Yeah, we like to we like to personify the algorithm for people who spend a lot of time doing this you get to know it Okay, awesome, that's super interesting. Let's look at the other side of the coin Are there I always think it's a difficult space to wade through at the moment and there's a lot of kind myths and misinformation out there Are there anything that you see very commonly? Kind of recommended or talked about that when you've tried it. You simply haven't been able to see that that has an
Araminta Robertson (27:56)
Yeah. Yeah, exactly. Yeah.
Yeah.
Yeah.
I think the biggest thing is that there's just so much vague advice out there. People will be like, yeah, do chunking or make your headers like this or make it more conversational. And I'm like, people are already doing this. This is not actionable. I like a framework, like something repeatable that I can scale over time, which is why I love our GPT framework. So I think that's the biggest issue. It's not that it's wrong. It's just like unhelpful and people who are calling themselves GEO experts and then they just, you know.
give out this kind of generic advice, it's not very helpful. I don't know if there's anything that's specifically wrong. I mean, I'm always going to be against mass AI content. So doing like hundreds of pages a month, all AI generated, that's a very good way to get penalized by Google. So definitely don't do that. And any of the, what do they call it? Ranked to tank or something like that? Something like that. Oh yeah, ranked to tank. basically it's like these...
SEO experts or not experts actually, just people who are mass producing pages to rank very fast and then it tanks because they're doing too many. So you have to be very careful with that. I think there's a lot of stuff out there about Reddit, which I find it's just so hard to manufacture. Like I think either it's authentic or you're going to struggle. I've been testing a bit like LinkedIn articles and Medium articles haven't seen a big impact. We did do a test of digital PR.
versus GPT articles. And I think digital PR is good for the long term, just in general, because you're building authority. But I still haven't found a repeatable framework that works as well as GPT articles, honestly. Like this is one that we have seen work best. Maybe we haven't been as rigorous as, we're not like a research organization, so we haven't done like rigorous testing with others. And maybe there are others out there. But the answer is simple.
create more content that answers very specific questions. And I find it funny because like when LLMs came around everyone was like, my God, content's gonna die. What's the point of becoming content marketer? And I'm like, it's become even more important now content because that's the only way to appear in basically prompts. So yeah, anyway. I...
Tom Rudnai (30:34)
Yeah,
no, I say often we're in the kind of black cat a phase and some of it will work in the in the short term. think there'll be a cost. I'd add is I'm just so much echo to anyone listening. Don't go down the route of just spamming out AI content. You will create yourself and you'll get short term benefit and create a massive headache pretty soon because the thing that you have to do now is you have to have a plan for how you maintain all of the content because as soon as you start getting kind of clarity, consistency issues that causes a big, big problem.
Araminta Robertson (30:39)
Mm-hmm.
Yeah.
Yeah, mm-mm.
Yeah.
Tom Rudnai (31:04)
Good luck in picking that. The thing that you just kind of came on to, which was the kind of importance of content, I also completely agree with, I think the value of human input into content as we come out of the black cat AEO phase will be actually greater than it probably was before. But I guess with that in mind, like, and when you're hiring at Mint, how does the skills that you look for from a content marketer
change in today's world.
Araminta Robertson (31:37)
Yeah, it's interesting because, so first of all, we hire writers mainly and then eventually they become strategists. it always, everyone starts out as a writer because having, you need really strong writing skills to edit an AI. So first of all, we never do fully AI. We always do AI assisted and I don't think that will ever change because if not, it's just very, it's still not reached that point where you can fully rely on it. And also I think you always want an editor just to make sure that it sounds good.
So when we hire, we're mainly looking for writers, we're looking for very strong writing skills. And the main thing that we're looking for is clarity of thought, a lot of empathy, putting yourself constantly in the position of the reader, and curiosity. These are the three things that we're looking for. Someone's got these three things, then usually they'll be a strong writer. And also all the other things like a communicator, but that's like, you know, of anyone in a white-collar job, that's important.
So for content marketers in general, think an interesting shift that I've seen is that it's no longer enough to be a great writer. And this is why we are, people hire us and why we're getting more work now is because they're like, okay, great writing is great, but like, how do we connect it to revenue? And I am seeing even freelance writers who used to be booked up constantly and always with a lot of work are now like coming to us and saying, can we work with you? Because
I'm good at writing, but I don't know how to do the strategy side of things. And so I think that's the key thing that's changed is that being a great writer isn't enough yet now because a lot of companies will use AI. And instead, what is still going to be difficult, and I think for a long time because it's a strategy, is, OK, you're great at writing, but how does this turn into customers? And that's a skill that's more important now than ever. And it's also what we love to do, so good timing for us.
Yeah.
Tom Rudnai (33:36)
Yeah, I think it's interesting because I speak to a lot of freelancers as well in a similar position, but there I think a lot of them are struggling to make the transition into strategy. They bump up against a challenge where in-house marketers don't want that from them. They want to do it. But I guess it's again, it's probably that that's a lot of selection bias there, right? Because when you want just content written, maybe that's when you go to a freelancer and when you're looking for strategy, you're looking more for an agency or a...
Araminta Robertson (33:46)
Hmm. Yeah.
really?
potentially.
But I mean, there's still like definitely freelance content. I definitely think if you are a very strong content marketer, you can be freelance for sure. You can if you're good at the growth side, there's still so much demand for that. So I think there's definitely it's just like they might not think of that first. They might think, yeah, agency. But I think a freelancer. The thing is that if you're going to be a freelance content trash, you'll need a writer to help you out. It's a lot of work for one person.
And that's why I set up my agency. was like, I want us to do this, but I can't do this by myself. I need a team because writing is a lot of work or editing or whatever it is. So yeah, I think you can be a strong freelance content marketer, but you'd have to be very good at marketing yourself as that because if not, yeah, you're right. They will not, an in-house marketer might not think of that immediately.
Tom Rudnai (34:51)
Yeah, no for sure. And what, so, because you keep kind of actually preempting where I wanna go, which is gonna make everyone think that we've planned this much more than we have, I promise, we're not that competent. In terms of in-house marketers, so one of my kind of views, I think one thing that lot of in-house marketers I speak to struggle with at the moment is,
AEO is this massive new thing and they don't like it's something that isn't just a project It's a muscle that you have to build up a new skill that you have to learn but it's very difficult for them to hire at the moment because the general pressure is in the other direction in terms of headcount so I think a lot of them are caught between that like How much of the AEO job do you think there is a benefit to having in-house or
versus using an agency. And I know obviously you have a pretty strong bias on that one, are there elements that you think are best kept internal?
Araminta Robertson (35:47)
Hmm. It depends so much on the agency. Like if it's us, I would say, well, we only take on clients where we do SEO content and AEO. We wouldn't do AEO only. And we have had people reach out and say, can you just do the AEO side? And we'll say no, because it's such a small thing that like, it just doesn't make sense to take on a full on agency just for AEO. I think it only makes sense if you hire them to do SEO and other stuff as well. So yeah, exactly. It just doesn't.
Tom Rudnai (36:11)
Yeah. It's so intrinsically linked to content. I don't think you can separate that
Araminta Robertson (36:17)
No, no, no,
Tom Rudnai (36:17)
because that's a, it's not an optimization job in quite the same way as I think people want it to be.
Araminta Robertson (36:21)
Yeah.
Yeah, yeah. It requires producing content. And if we're to produce content, might as well do SEO as well. It just doesn't make sense to separate the two. So it depends so much on the stage, on the agency they're going to hire. It depends so much on where they are at in their SEO and AEO journey. So what I would say is if the agency, because we developed our own framework, if the agency uses that framework, then I would say pick the agency because they know what they're talking about. They're doing the right thing.
If it's an agency that is telling you to create tons of AI spammy content, I would say don't pick an agency. So this is why it depends so much on that strategy that that agency is going to implement. Having said that, whenever someone comes to us and says, can we just do AO, I just send them to our article that explains our process. And I'm like, you can do this internally. It's not that complicated. And I think that if you, and we actually did have this with a client of ours where the work
had reached a natural point where results were really good, there wasn't much left for us to do. And we were like, honestly, you guys are doing really well. We're just going to pass on our framework to you and you can implement it internally. And it's literally happened. a few months ago, well, actually a month ago, we checked in with them and they were like, yeah, AI, LLMs bring in 20 to 30 % of our amount now and the best quality as well. So that's why it's, again, based on where they are in their journey, it's hard to answer that question. I think if they're very early stage again, maybe agency helps with strategy.
framework, but if you're quite, you have quite a mature contact program doing well with SEO, I think you can do AO internally for sure. And, and I can send the article that I talked about and implement, implement the GPT framework and you're good. You're going to do well. So yeah, I kind of try to answer your question, but as you can see, it really depends on the situation.
Tom Rudnai (38:07)
Yeah,
it's a difficult question. mean, maybe a better way to frame it is like, what do you need from a client alongside what you do in order for to do in order for what you do to be as effective as possible? So if I'm an in-house marketer thinking of bringing in an agency, what do I also need to think about in order to maximize that investment?
Araminta Robertson (38:28)
Yes, so I'll talk about specifically for us rather than any agency because I don't know every agency will be different. I think the first thing is, it sounds really like simple, but having a good personal working relationship. So they're willing to trust us. They're willing to listen to what we have to say and rely also on our experience. I think if they want to decide on all the topics, for example, this will make it hard. understand where we're coming from. And that's why I always say like, you've heard us to bring us customers. Let us do that.
That's what we're going to do. So I think inherent willingness to trust and listen, I think that's a big one. I think the other one would be access to subject matter experts. I always say this before we start anything. I'm like, if your experts are super busy, no chance to talk to them, then I don't think it makes sense for us to work together because we really need access to the experts. And in the first two or three months, we'll need a bit more. And then after that, it tails off and not so much, but they do need to set aside a little bit of time.
And I think maybe ideally an internal person who is going to be dedicated to either content or who can be our point of contact and has the capacity to work with us. So especially the first two months, there's more reviews necessary to make sure that we're getting the product correctly, we're describing it correctly, and to make sure that we're getting the tone of voice correctly, all of that. Someone who can spend a little bit of time the first two or three months just to educate us and train us as well. And then just the last one, this is just in general.
I think it depends on the maturity of the company as well, but strong positioning, strong USPs really helps us. So if it's definitely true that earlier stage companies, might be changing their positioning every few months, that can make things complicated because then we're not sure, like, what do we say in the article? What are your USPs? If those keep changing, that's not good. Also for positioning, like the keywords we're going to go after, that keeps changing. That's going to make things tricky. So think having a clear, like positioning, branding and all that, that's just, that's not just for us, for anything in general.
any agency or freelancer you work with that will make life easier.
Tom Rudnai (40:31)
was just
the job of content is to communicate and kind of amplify the message or positioning a product, right? If the product is crap, then there's only so much you can do. You can't put lipstick on it. Yeah.
Araminta Robertson (40:37)
Yeah.
Or if it keeps changing as
well, or if the priorities keep changing. But more mature companies, doesn't tend to happen. Like that's once per year, that's fine. Once every six months even, that's okay. every month it's like, okay, well, makes things tricky.
Tom Rudnai (40:55)
or you'd hate working with me. Final question I want to ask you, just because it's something that I've been thinking about lot recently and I'm really interested in, Chad GPT launching ads, do you have any kind of take or hypothesis on what impact that might have on the kind of organic, AEO side of what we both do?
Araminta Robertson (40:58)
Okay.
Yeah.
Well, I haven't really, to be honest, looked into this too much. I think because first of all, they have said that it shouldn't change responses. That's what they've said. In reality, we'll see what happens. But if what they say is correct and true, then truly it doesn't change anything. So we kind of to see what happens. See what happens there. And then also only free users will have ads, right? So paid.
Tom Rudnai (41:21)
Hmm.
Araminta Robertson (41:42)
head users won't. So that's also something to think about. think if you're targeting people who, because the reason we primarily focus on ChatGPT is because it still is the tool that has the highest number of users, but that is decreasing. I know Gemini is catching up, so that's also something to keep into account. yeah, so far I'll just say it hasn't changed anything because it seems that it shouldn't bias or change the answers. If it starts to change the answers, then that's when I'll, you
then we maybe need to change something or look into it. But I think it's too early to say. I haven't really got any thoughts on it. Have you got any?
Tom Rudnai (42:18)
I have thoughts but not conclusions. I think I'm very interested in how it potentially opens up another lever for an AEO strategy, right? So how can you find ways to use it? Yeah, if you have a budget that you can invest in ads, then maybe you can find a clever way of finding, I'm going to use the horrible word that I hate using, you'd find a synergy between your organic and kind of paid. The other thing I'm just curious to see,
Araminta Robertson (42:21)
Mm-hmm, yeah.
Yeah, that's true. With ads.
Mm-hmm. Yeah.
Tom Rudnai (42:48)
In order to run ads, ChurchyBT has to give you a way to explicitly communicate who you are, who you're for to it, right? Because it needs to make sure that it's nailing the ad content versus what you want. Now, most of what we both do day in and day out is help brands communicate, inexplicitly, implicitly, who they are and who they're for. So I'm curious whether that presents an opportunity to educate the model much more explicitly in a way that actually...
Araminta Robertson (42:55)
Mm.
Yeah, that's interesting.
Tom Rudnai (43:16)
carries over beyond what they're doing in Adverts, it's going to be interesting to see how they
Araminta Robertson (43:19)
That's a point.
Tom Rudnai (43:20)
implement that. But I don't know. Yeah, I think that's
Araminta Robertson (43:22)
Yeah, that's possible. I don't know, it's very
interesting. We'll see. think definitely once there's more, I think it's only live in the US at the moment, but once it's more worldwide and stuff, I think it'll be really interesting to experiment and test with that because yeah, you're right. It could have other benefits.
Tom Rudnai (43:40)
Yeah, for sure. that's the, I don't know about you, but that's what I love so much about working in this space. It's so much fun because it just like changes every day. And you just get to spend so much of your day figuring stuff out.
Araminta Robertson (43:51)
Yeah, definitely.
Tom Rudnai (43:52)
Cool, right, before I let you go, let's do a couple of quick fires that I like to do at the end. So the first one which I'm going to give you is an AI use case or an AI tool that you absolutely love.
Araminta Robertson (44:05)
What I've been getting into recently, anyone who follows me on LinkedIn will have seen is building little AI tools with lovable. I'm completely addicted. My Sunday afternoons are spent building these tools. And I want to say specifically not for public use, for internal use. So for me, my team, and our clients, the whole doing it public, you need so many other things in place, compliance, all these things.
start charging people money. No, I think that's too complicated, especially for us. We just use it for internal. So I built an internal linking tool that's actually useful that we actually use. And we've built a bunch of other tools for clients that are really exciting. that's, I've been addicted to that lately.
Tom Rudnai (44:47)
I'm exactly the same I've got two groups with Claude code and the stuff that I can build with it in like a weekend is I mean it's awful with my weekends now I just sit in like a dark room coding but I like yeah I keep making fun of our engineers like I'm so good at this they're building me little playgrounds where I can like do stuff and like right just stay over there but don't touch the actual product but I can kind of mess around in staging and stuff like that
Araminta Robertson (44:57)
I know, it's not good.
Mmm.
That's cool. Yeah, that's true.
Tom Rudnai (45:12)
Yeah, cool. I love that. And then for you personally, is there a skill or a trait that you have that you think has really helped you in your career?
Araminta Robertson (45:23)
I think this could be, it's a positive and a negative, which is being very stubborn. Which I think that just when things aren't going well, just being like, I don't care, I'm going to keep trying. Like even if everything is telling me to stop or everything is like, I think that's what's allowed me to keep going when things can get tough. But there's also downside to being stubborn and my husband will always say that. Which means that sometimes it can be...
You have to also know when to quit and when to stop. But yeah, I think it's benefited me a bit as well. bit of the two sides of it.
Tom Rudnai (45:59)
Yeah, I think you could probably find some people who would say that I have a similar trait. And if I, maybe, how we can rephrase this one for you, but I always find it quite fun. If I were to approve your plan A budget request and maybe, you can go at this in two ways. The one that your board would never be stupid enough to let you have, let's pretend, I don't know if you have a board, pretend that you have a board. Or the thing that just you could never justify, but think would be really fun to do.
Araminta Robertson (46:03)
Yeah.
Tom Rudnai (46:26)
or maybe that a client would never give you the budget to do if that's more fun for you, what would you do?
Araminta Robertson (46:33)
So if we're talking about Mint Studio specifically, I think I am so bullish or excited about events. If I could, I would hire a bunch of fantastic event marketers and just do lots of events. I love being one of the organizers and attending them. I don't like being the only organizer because they're very stressful. But hiring a bunch of event marketers and just doing lots of things like dinners, workshops, side events, everything, I just love it.
Tom Rudnai (46:46)
Hmm.
Araminta Robertson (47:00)
It's so much fun, especially when we spend so much time in front of a laptop, being able to do in-person stuff with pretty smart, interesting people. I, yeah, I love it.
Tom Rudnai (47:09)
Yeah, I've had the same thought actually quite recently because I think it's something that works really well at the moment, but I'm just not like an event organisery person. I've been really bad at organising stuff for my birthdays. I barely ever invite my friends over. I'm not the host. But I think it's a very effective thing to do at the moment. And then final one is always a bit more fun. I always think on podcasts people sound really smart and I think it's nice for people to realise that they're often actually not.
Araminta Robertson (47:17)
Yeah.
Tom Rudnai (47:39)
Like what was the biggest screw up that you have or that you remember in your career?
Araminta Robertson (47:45)
I'll just say the most embarrassing one because I haven't really, I'd have to think about it for a lot more longer to think about it, to really have an answer. I think two embarrassing moments stand out. One is on stage at one of our conferences, I think. I accidentally called someone by their company name rather than their actual first name. I was exhausted and it was the end of the day. Luckily, everyone laughed, but I felt mortified. And then I think at a different event as well, I...
Tom Rudnai (48:09)
You
Araminta Robertson (48:14)
completely forgot someone's name. I knew this person well, I knew her name, not a problem, but on the stage I actually said, I'm so sorry, can you remind me your That was very embarrassing as well. So it happens, yeah, when you do a lot of events, things are gonna go wrong and yeah, that's happened, unfortunately.
Tom Rudnai (48:30)
Yeah, it's that thing.
Do you have like the lights? And they're so disorienting. It's so easy to just have like a mental block in that moment. Can you share the name of the company? What did you think this person's name was? Okay, yeah, fair enough.
Araminta Robertson (48:36)
Mm-hmm.
No, I don't want to embarrass them even further or embarrass myself even further. On that stage,
that was enough. No, I don't need to do it anymore. But luckily they took it very well, really chill. And I'm still in touch with her and everything and very nice people. to me, I'm always like, especially someone's first name is so important. People always mess up my name. Like, I know how important it is to get that right. So it was it was a weak moment for me. But I try to not let that happen again.
Tom Rudnai (49:11)
Yeah, I mean, luckily that's not one that I empathize with as a Tom. don't have to I don't really have to come across one too much. That's not a battle I have to fight my son and that one I get. And then just before I you go one recommendation for everyone. So a book, a podcast, a thought leader, something you love that people should check out.
Araminta Robertson (49:16)
Yeah.
Yeah.
Yes, so I love April Dunford, huge fan of everything she creates, especially her book, Sales Pitch. I think if you're selling anything, whether you're a freelancer, agency, or if you're an in-house marketer trying to, maybe if you're product marketer especially, and you're trying to understand sales more. What a great book. If you're in the agency world, Blair Enns, David Baker, love them. Anything they do is fantastic. And then solely marketing, I'm getting into Emily Kramer. I think that's how you say her last name, Kramer, I'm not sure.
Tom Rudnai (49:54)
Yeah, market one.
Araminta Robertson (49:56)
Yeah, so good. Wow. It's so nice to have like a prolific creator who's also just so great and really actually genuinely useful stuff. I loved her sub stack on how to run dinner dinners, like executive dinners or founder dinners, called them so good. So I think, yeah, those three or four resources there are really great.
Tom Rudnai (50:19)
Great, and then I normally at this point say anything that you want to promote that you're doing, but I guess I can do that one for you probably and say go you go check out Mint Copywriting, but I'll give you the space to do that yourself. Sorry, anything that you'd like to promote that you're doing.
Araminta Robertson (50:28)
Studios, yes. Sorry, what's question?
promote?
Well, we just relaunched our podcast. It was under the Fintech Marketing Hub brand. Now it's under our own brand. So if you'd like to check that out, under, it's called Fintech Marketing Lab. And we've got some fantastic guests, including yourself, who are coming on. So I'd love if you could check that out.
Tom Rudnai (50:54)
feel free to skip the me episode. I will not be offended. I don't blame you. Cool, Armenta, I'm gonna let you move on with your day, but thank you very much for joining us. think one, like, I always love, I think anyone who's trying to like sort through who to listen to in the space that we're in, the AI search space, ask them questions. And if they don't ever say they're not sure, then just leave, don't trust them. And you said you're not sure like five times in this switch.
Araminta Robertson (51:18)
Yeah,
Tom Rudnai (51:19)
You could
Araminta Robertson (51:19)
definitely.
Tom Rudnai (51:19)
frame as a negative, but I think that's a fantastic thing. It's a reason for people to listen. So thank you for joining. been fun.
Araminta Robertson (51:29)
Thanks a ton for having me, it's been a pleasure.
Tom Rudnai (51:31)
Bye bye.
