How Do You Use AI in Marketing to Save Time?

Alex: Cool. Okay. Nikki, do you wanna start off, just tell us a little bit about yourself and your role.

Nikki: Sure. So my name's Nikki Laker. I'm the marketing Director at Opus Technology. We are a managed service provider based in Reigate, and we offer five different solutions. So I have a very varied role, but also a very challenging.

Alex: And the reason you're on today is because you are leading the way very much on AI within your team. Do you wanna just tell us a little bit around kind of how you, how you got into AI and how you started using it?

Nikki: Yeah, sure. So I've been an early adopter of AI, specifically ChatGPT since October, 2022. And the way that I got into it was almost by mistakes. It came out. People know a little bit about it. We didn't really know what it did. There was a lot of chatter. But I'm very passionate about SEO and having a website that offers five different solutions is very difficult when it comes to SEO because I need authority across each.

So we're a little bit of a, jack of all trades when it comes to Google's perception. So one of the real starting points for me with AI was. Do I make a glossary for each of these solutions very quickly. So to do that without AI would've taken me probably two or three months. It would've also been incredibly boring. So that's where I sort of dabbled in ChatGPT in terms of what can this actually generate for me? On these topics, and before you know it, got it. We just clicked. And it's helping me generate landing pages for and I mean thousands of glossary terms. And I'm thinking, do you know what, this is actually really accurate. Like I've only got to do a slight enhancement to this. The more and more I used it the more I became obsessed the more curious I am. And yeah, it's part of my everyday life now.

Alex: So particularly as an early adopter, how did you get past that point of just having a high volume of. Dross to actually making it good, credible content.

Nikki: Yeah, so that's a really valid point. And I think this is why people fall short and certainly in my early days it is, oh my God, look how much I can do with the, LLMs. Whereas now it's like, how much can the LLMs do for me in terms of quality? So 2022 whacking out, God knows what for this glossary, and I'm sitting there and I'm still editing. But it's still a win because they didn't have to generate it in the first place. Whereas now the custom gpt that we have at Opus are outputting 95%, and I'm very passionate about this. So it's not now do more to a rubbish standard is do less internally. But output it to a really high standard.

So you are actually not doing half of the editing that you ever used to do because you've taught it in a way that it's replicating either you as a person or you as a brand.

Alex: And so you started off, with SEO. Tell us a bit more about kind of how you're using it now across the team.

Nikki: So, mean, how long have you got? Using it in lots and lots of different ways. My mantra for this year and last year was doing more with less. So we have an exceptionally large sales team at the business that I work for. We also have a pretty small marketing team the reason that it's pretty small. It's 'cause it's optimised and it's efficient. I'd say in terms of doing more with less, we the obvious one is a unicorn. Our unicorn that does 95% for us. And maybe I can talk a little bit more about that. And when we get onto custom gpt the other use cases for AI within our business, marketing and sales are very well aligned. We've actually done is enabled sales to sell more using AI. So there's a couple of different things we've done from a sales enablement strategist, for example, that will help train you and guide you, and answer any queries you might have on your objections or anything else in the sales process to a salesperson or assistant that would do a lot of the heavy lifting that you as a salesperson. Might not realise it's actually heavy lifting 'cause you've always done it that way. So it'll go out, find all the information about the prospect, tell you things you need to know, tell you what they're interested in, tell you what tech stack they're using, match their problems to our solutions as a provider. It'll go another step further and go, here's, here's core scripts for you. Here's email templates in the space of like 30 seconds. So I think like as a marketer, yeah, obviously I want my team to be powered by AI. Don't we all But the wider business too. And the way that I think we've had huge success with that is sales and marketing are big. But we're also leading by example and we are showing the rest of our business what is possible. I think that's really important. Than just saying, guys, everyone now needs to use ChatGPT. Great. How? Oh, I dunno. You just need to use it. Which a lot of businesses are there, whereas it should be, here's what it can do for you.

Here's what it can deliver. Here's the time saved. Here's the enhanced output that you as a person or team or manager can achieve by using this software. Go, you, we'll support you off, you go crack on and you know, win.

Alex: Why do you think marketing is the best place to do that?

Nikki: So if I'm honest, marketers are generally creative. However, I think the marketers that succeed with AI are creative and analytical. Curious. Um. We see use cases every day. So my team have to use AI, not because I force 'em to, because they see the real value in what it can actually deliver. Use cases only come about with people that sit there and use it and are curious and learn and listen to podcasts and, establish a problem or a process that needs optimising. So it's only yesterday actually. There was another use case that our team has found, is using AI to help conversion rate optimisation. I bet you 99% of people listening have probably not thought about that, but it can do a lot more. Than what perhaps you think it can do just by looking at it slightly differently.

So analysing your site based on psychology, based on personas. How does it come across, how does it read? How's the user flow? And that's one yeah, we found out yesterday and I'm just so curious as to how you get it to do this. But I'm absolutely gonna try everything I can to, you know, succeed in that space.

Alex: Yeah. So it's that curiosity as you've said of, exploring it and allowing yourself the time to explore it. And often the challenge is these things take longer the first few times. That you, that you do them and you're trying to figure out the solution and you're trying to train the models, et cetera.

How much has that been a blocker for people in the organisation?

Nikki: So I guess the mentality is. Certainly with us as a business, you don't know what you don't know, and unless you try, you're never gonna really learn. Failing fast is something we encourage and there's definite roadblocks. You know, oh, I want to make a custom GPT that does y, z, and this and that, and this, and everything in between.

But I've never made one before. Well. Don't start there, then start with step one and see how you go. And not, I think not having the answers. This would be a really good tip that I'd like to share. Get stuck with AI. Everyone does. We're all beginners. I don't care what you say. We are definitely beginners.

Some more advanced, but none of us are experts because AI moves so fast. When we get stuck with it, that can help us. LLM and we had another win today on this, in terms of a suggestion that I was researching over Christmas, something called a proof page, which everyone can go and go and research what that might look like, but it's something that l LLMs really value. So we put one together. It took 15 minutes to do this. I thought, oh God, this is hot air, whatever. Maybe it's elucidated. Is this backed by anything, but do you know what? I'm gonna have a go. So we've put that together over Christmas and then we've just had a really big inbound lead been generated from this page. Mad. Like it's almost when I, when I see things like that and we track everything. I'm like, okay, well it does know what it needs because it told me what it needed. I've delivered what it's needed, and guess what, it's now delivered me what I need. And not being afraid to do things like that. You know, certainly from my perspective, sometimes we do some crazy things and I'm like, I'd rather somebody said to me, that's not right.

You need to do it like this than we are just the same as everyone else.

Alex: Hmm.

Nikki: Because if someone else knows better than me, then great. If AI knows better than me, amazing. If we're all the same, then how can we win? I don't understand.

Alex: Yeah, no, that's a good way of looking at it. And I think it's, again, it comes back to that curiosity and that ambition. And, and the key thing, you know, I talk about all the time is you're not outsourcing the thinking. There you are, you're using it to help improve what you've already got and what you're already, what you're already doing there.

You mentioned the unicorn, so. Tell us, tell us about your unicorn.

Nikki: Okay the unicorn has had quite a lot of time spent on it in terms of optimisation. So it's a custom GPT. Most marketing teams will have one of these, but they might not be quite unicorns. And basically what it does, it's, it's particularly clever, it will analyse keyword gaps based on our content. So we've got five different solutions. So it'll say in this solution here, you are missing a trick on these 5, 10, 15 keywords, whatever it might be. It will then generate us high quality blog content around that keywords. It'll go out, work out what the most relevant title is, based on demand, our content for us. And before you know it, you've got a high quality blog or landing page. Written in Opus style, it's factually correct 'cause there's a lot of guardrails, implemented. But the best bit for me, which speaking to my peers, no one's quite managed to get, it maps the blogs that it outputs to our site map. So it pulls all the internal links in for me, which is amazing because internal linking, if anyone knows, takes some time. So generates the blog the title the meta. Then it goes, what's this blog? What's this landing page? What's this guide about? Here's the Midjourney prompt for you. So then all we need to do is bang the prompted midjourney and we've got our images. The other thing that this unicorn does for us is it spits out all of the information that I've just told you in an order that we need it to optimise the upload. And it also works out how much time's been saved based on what's been generated. And that's something that my team track, if you as a user don't understand the why and how this is helping you, how do you know you're doing the right thing? But more importantly for me, how much time has my unicorn saved our team this week, this month? And what's that allowed us to do? So that's our little baby gets refined quite a lot. It's backed up to the heavens. But it has been, it's been transformational

Alex: And,

Nikki: In terms of output.

Alex: And just to unpick a couple of things in there the timing, and the fact that you're measuring the timing is really interesting. I found AI is not great at measuring or estimating time. So have you preloaded that and said, this is how many hours we would typically spend on X activity.

Nikki: Yeah, so that's in the instructions. On average, we will spend x amount of time generating a guide a landing page a blog. Calculate how much time you will have saved us by generating this. The other thing which I actually inputted is the upload time that some people will forget. Whilst it spits out 95, this is why it can only do 95. The five the other 5% is the manual upload. And the day that comes when there's an integrative WordPress plugin that I can spray this into and just watch it upload and there you go. You know, that would be when I'm on a hundred. But it's, it's fascinating to see the amount of time saved in the first month alone of using the unicorn. We saved five full days worth of time for my team, is a hell of a lot of time, which allowed us one to do a lot more strategic activity, but two, catch up on all one of the other bits that we constantly have to push to one side just because of workload demand, all the different activities that are going on.

Like I said, we're a small team, so five days was absolutely, you know, magnificent for us.

Alex: It's interesting, I remember seeing a couple of years ago a piece of software that effectively did what what you've just described, as sort of in the early days of AI. And that was doing that, of looking at the keywords and then generating content for it. And it wasn't really very good.

But I do remember at the time as well thinking that the direction of travel, those things have such a short shelf life. 'cause people can just build them. Themselves And, and you are the case in point. So, and tell me about the linking, because that's not something that I've, I've, I've seen before.

Nikki: Yeah, so like I said, internal linking, unless you're using something like Link whispers, which is okay, but there's certain, certainly, when you are creating content clusters, I'm not sure I rely on it very heavily. The internal linking that the GPT that we have works by is knowledge, section area of the GPT.

We've uploaded the site map and then for each page of specified my keywords relating to the page. And when it's generating the blog, it will pick out all of the keywords and it will link them back. But there are set rules. I don't want keyword stuffing. I don't want a load of old toff when it comes to the internal links, that it has to follow to be able to do the output that we require. That's the part of the GPT that took the most time to achieve, but it came further on because all of a sudden you've got good quality content being outputted. But then I'm still thinking, we're still not saving time on this part. How do we solve it? How do we solve it? And that was a few hours work of. What am I missing in my instructions? How do I get you to do this? How can I help you? What do you need? A crazy dialogue that went on, to achieve the output that we have. And a lot of people say, oh, like, how have you done it? And, and that is the way, but it's a lot of, it took a lot of tinkering. Again, that's just being curious. You know, there we were probably 91%. How you get that up? Um. Ask AI, it will tell you what it needs. I promise you not read 50 manuals as to how to internal link A. A custom GPT waste of time. Don't bother.

Alex: But I think it really shows the importance of still having that domain knowledge. Yeah. I don't think that's something that you can just, press a button and, and have something like that. You need to a, have the education and the time to be able to spend on, on building it from an AI perspective, but you need to have that domain knowledge.

Nikki: Yeah, you do. And I would say perhaps, that might be where I'm slightly different because of the SSEO knowledge that I have. Also technical. Whereas perhaps others don't have the extent of that. But that said, it doesn't matter about your knowledge. There'll be other people around you that have that. If you're working a small team, you use an agency, you outsource whatever it is, you ask them for those types of things that they'll be able to, you know, provide the site map for you. Ask your SEO consultant if you have one to sort out your keywords. There's lots of. Options there, it might just not sit with you.

Alex: And you've spoken about a 90 day AI plan. Tell, tell us about that.

Nikki: Yeah. Okay. So my team, as most do work very well in sprints, and it was just before Christmas that we tried to launch this have successfully launched it since then. I wanted dedicated times 20% of the week on a Friday. Dedicated as a team to AI because there were two reasons. I wanted to give it the maximum effort to see what we could achieve, but I also really wanted to increase my team's adoption and learning and curiosity.

So every single Friday as a team, we sprint, which is great. Now there were two objectives of the 90 day AI plan. The main one for me was Forex Marketing Output 90 days. And that looked different in different ways, but the two main things that I wanted to achieve is doing more with less. Using our unicorn. How else can we develop as a team, custom gpt in other areas, which we have done since there's 11 of them. And also. How do we as a brand be in more places? So it goes back to the SEO thing, but also GEO, which is something I'm really passionate about, they go hand in hand, it's like, right, massive opportunity here to be an early adopter. Let's go. We've got all the knowledge that we need as a team. Let's collaborate and see what we can achieve in terms of being cited. How are we featured in prompts? How much traffic are we getting to our website? Many, MQ LS are generated from AI platforms? How many of those get quoted and how many sell 90 days?

We have had one that's come through. So the quality, it one, in terms of conversion, by the way, sorry. The quality of the, leads that we are getting from LLMs. Is very good. The timelines also much shorter. So that was a 12 week timeline, by the way, from initial inquiry to sale, which for us as B2B is not very, long But also month on month. How can we make the incremental gains to know that all the work that we are doing right now to future proof us in terms of AI strategy is gonna pay off? And I don't believe we're gonna see it tomorrow. I don't believe we'll see it next month, I do know that we will definitely start to see that ramp because we've invested the time and the strategy to focus in this area.

Alex: And that's specifically around GEO. How are you, how are you measuring the impact of that and how do you, how do you think that's gonna change as you go forward?

Nikki: Sure. So it is very difficult as a lot of marketers will know at the moment to measure the impact of AI on your efforts. So what I did in the 90 day plan is. I just literally outlined key metrics that we're gonna measure. They might not be the right ones, but we're gonna measure them to make sure that we are growing. So the first one for me is brand search volume. Generically, it's not just an AI in search, in absolutely everything is the brand search volume of us as a business going up. It should be brand mentions in LLMs can be tracked. AI visibility as a score. So as you'll know, certain platforms will give you different readings on this. That's fine. Stick with one and make sure it's going up. How many cited pages our website has in LLMs, again, as a chosen platform that we use for this, and we stick with that. How much AI traffic was generated to our website and how long did it stay there for? Did it go? And then like I said, how many leads do we get generated from AI? How many of those leads turn into M qls? How many are quoted and how many are one? And that's the key part. But at the moment, that's the bit that's not gonna be significant. 'cause you've gotta put all the effort into the top to grow that.

Alex: Yeah, because I guess when you're looking at it now and you're comparing it to some of the things you have been. Doing over the past years, it's, it's probably not comparable, but you are, you're kind of early on that stage of the direction of travel. Is that fair to say?

Nikki: Yeah, definitely. I mean, increasing AI visibility for me is the number one KPI right now. Obviously the dream would be increase visibility, increase the amount of inquiries, increase the number of creative deals, and guess what? Yes. Oh, we've got this amazing pipeline coming through. It's all selling. Well, that's not real world, right? Start where I did many years ago, Google, when SEO became a thing. Everyone was jumping on it and there was millions of different ways to achieve rankings. I remember sitting there in search console indexing pages, and within the hour I could see how it ranked amazing brand new ones. I just literally wrote them within the hour.

They're on page one. What a win. You can't do that anymore, unfortunately. And I think that's the same when it comes to, LLMs. It's understanding what works by measuring it. And just sticking with what you're measuring and seeing progress. And like I say, Alex, if I speak to you in a month's time, I might say, so, you know what?

Do you know what? I've changed the KPIs. Okay. And I'll change them because I don't think they're telling me the full story, or I'm curious at what else could be measured. But, right. Yeah. Right now, because it's so early, there doesn't seem to appear to be a single source of truth with LLM measurement for marketers. Which is why my recommendation is pick a platform, whatever it is, you can measure in that platform, just stay there rather than, let's take an average of three platforms. 'cause they all tell you something slightly different.

Alex: And so your view there is if one's going up, they're all going up, or is it that you want to pick one horse to back?

Nikki: As a me, so as a metric, as long as it's going up and not going down, I would stick with it. If you, if you are measuring, say for example, you might have a marketing team now that is measuring total number of leads generated by LLMs. Oh, God is going down, but they're forgetting to measure the amount of visibility that they have on LLMs right now and how many prompts that they are actually being featured in. We are trying to, I think as a world, how behavior is changing. And certainly from an e-commerce perse perspective, if you look at how LLMs are being used, it's, it's a radical shift. In the B2B space, it depends on your user. I speak to people in our ideal customer profile a lot. Some of them are really familiar with it, so they're using, they're using AI for deep research, amazing on vendors, providers, resellers, whatever it is they're using it, whereas others are using, ChatGPT as a search engine. 95%. I listened to another podcast the other day on Miss. Apparently 95% of ChatGPT users only use it for search. So we're trying to establish how people are changing their ways in terms of buying. A lot of people are using LLMs right now to review providers, so that goes back to that citation, importance. Where, where are you? It's all about who's the best, who's the most reliable, who's the most trusted. And it's working out where it pulls those sources in which it shows you, by the way, and how you get on them quick before anyone else does or to quickly catch up. 'cause your competition is way ahead of you.

Alex: Yeah. And it's, it's only gonna increase, how people are using it and the depth to which they're, they're using it. And I had a guest on here a few episodes ago, who work works for an SEO agency. And, and it's that importance of, and. How people are talking about your brand, not just what you are putting out yourself on your page,

Nikki: Yeah,

Alex: But what's, what's, what's the conversation about you out there as well?

Nikki: Definitely. The other. Thing that I wanted to kind of share, which again is curiosity, is the power of Reddit. If you're passionate about AI, you will know that Reddit is a real trusted source of LLMs, and it genuinely is. But how can you harness what's in Reddit to accelerate your marketing efforts? So we have a custom GPT that scrapes Reddit for us, suggests topics that we don't rank for example, but also can you be part of a conversation that perhaps you haven't ever thought of using Reddit. And I'm definitely not the stereotypical Reddit user, by the way. I find it incredibly hard to use. But what we're actually doing as a team is. AI. We are learning how to use Reddit and it's not easy, but again, things we found out from the platform that feeds lms, we would never have found out if somebody hadn't written to me on LinkedIn and told me all how wonderful it is and show me all the data in terms of Reddit authority. I've just been like, well, you know a platform. It's not for me. It's a techie person's platform. But inside there, learnings. And there's other equivalents of Reddit that are pulled into LLMs, we're now delving into those too. So yeah, you go down rabbit holes, but some of the rabbit holes that you go down are actually hugely beneficial, not just for your team, but also across the business. For example, we have a service desk. It's a managed service provider. I'm like, guys, where's your Reddit account? There's so much to be learned in there. Why are we not there? What are you telling me about Reddit? Let's do it.

Alex: So, so just, talk that through a little bit. So you're researching. Reddit around some of those topics, and then how are you using that? Are you then contributing to Reddit or are you just using that to inform your content strategy, or is it a mixture of both?

Nikki: We do both. So on Reddit, I'm a, I am a member of Reddit. I have a profile, and I'm also in the right subreddit. Funny enough, ChatGPT suggested which ones I need to be part of. So I just observe what's being said in there. And then the GPT that we do have that scrapes, Reddit find relevant topics that we don't have any content for on our website should have. The other thing is I am part of the conversation, even if I'm a beginner in some of these topics, 'cause they are very technical. But what I try to do is something called parasite, SEO, again, it's another whole topic that I don't have time to go into now you are seeding content from a human on Reddit, with useful resources. They're linking back to my site. But it's, it's been really beneficial so far to really understand how that com, how the community works. Are going to a world of communities. They're so powerful. We create them ourselves. Just see. Yeah. Seeing underneath the bonnet of your ideal customer profiles right now, how are they really, like, what do they care about?

What are their problems and how do you meet them and how do you be part of the conversation? I think is the really important bit, and I've gotta be honest, we haven't, we haven't cracked that yet on Reddit. We're ultimate beginners, I would say, although I do have, I think I've been on there for two months and I've only been banned once. But I would definitely say like, be curious and just delve into that, delve into that little world. And if you do nothing else, you'll learn something that you didn't know about either a current trending problem that you won't find anywhere else. You will find that in Reddit. It's a weird place, cool.

Alex: And let's, let's go back to, your custom gpt. You, you spoke about a few of them. Just briefly talk us through some of the main tasks that you've got other ones set up to do.

Nikki: Sure. So we got the sales personal system, is great, that overcomes a lot of time. That is drained on our sales team looking for relevant prospect information to be able to formulate stronger conversations, proposals, et cetera. So the personal assistant is programmed to go out there and find out all the things you need to know about that person. At the top of it will tell you. Five things you need to know. So about me, for example, what her interests, what does she talk about on LinkedIn, where she works, blah, blah, blah. It will then say, how long? How long has she worked there for, and who else is in her team that could be a decision maker to us. Really important. Then it will go, what tech stack? What tech stack does Nikki use? How does our solution meet the needs of her based on the tech stack that she's using? And also, how can I start a conversation with her that will be relevant? It'll do call openers and it'll do emails in terms of drafts based on what it knows. And like I said the thing the thing that's really good about that is it literally takes 30 seconds to output the results. And it's taking the information from a multitude of sources online. Tech stack, for example, not letting you all into secrets, but that's pulled from job ads relating to the company that we're looking at. It saves sales time, it helps them. The cool scripts are there as a guide. They're based on relevance. It's hard enough these days to get conversations with people or even replies to LinkedIn messages. If you're relevant you've done your research, you've got a much higher chance of doing so. So that's just one part of the sales process that is optimised using AI. And it's been particularly effective so far. It's also helped sales with their adoption.

Alex: And that is set up in a way that presumably they just put in who they're working. With or which account, so. So has the adoption for that been quite high, or are you still finding some resistance?

Nikki: No, at the end of the day, I said to them, first of all the way that we rolled this out was I want five volunteers to trial personal system with a caveat. Is minimum viable product. Like this is the bare bones personal assistant. Can you try it for me? Value your feedback. I want to understand if it's useful. So they did that and then before you know it, they've told two more about it. They've also told us a lot of feedback. For example, oh, it'd be really useful if you, if it could give me their LinkedIn, URL. Great. Just change the instructions. Well, it's that quick. Yes. It doesn't take days to do these types of things. And now every single salesperson pretty much give or take, probably two or three is using it. And if anyone says they don't want to use it, I think that would be quite a hard conversation as to why, because of the time that it saves. Why would you not want to save time when you could be selling

Alex: Hmm.

Nikki: Or having a more meaningful conversation like that doesn't make any sense. Only thing that I am having with it is it's a custom GPT. Some people are more Copilot. Well, that's absolutely fine. 'cause guess what? There's an exact replica in Copilot. It uses the same instructions, but it's in the Copilot environment. And I think giving people choice as well helps with the adoption. So I, it is only on those two. I don't plan for it to go anywhere else. But getting a little. Test group of people to be your early adopters, to give you the feedback, to feel that they're part of the vision is the way that you roll that out. 'cause Yeah. You know, word gets around the benefits are there that gets constantly updated.

Little tweaks along the way. I've even said to some of them. You all work very differently. If 95% of this is okay for you and the other five percent's absolutes off, you're not interested in it in the slightest. Well that's fine. I'll make you a custom one. Like by person if you wanna change the order of the output, absolutely fine.

I don't mind. We'll do it. We can do it together. I could show you how to do it. That's even better.

Alex: Interestingly there, you talked around having it in, Copilot and ChatGPT.

Nikki: Yeah.

Alex: Do you see a difference in the output? Yeah.

Nikki: Yeah, I do.

Alex: You don't need to be diplomatic here.

Nikki: I've just done extensive data analysis using both platforms, same dataset, same prompt. It's fascinating the outputs, so they think very differently as platforms, which you don't really understand unless you do the side by side comparison. Um. A Copilot in my experience, in terms of the output is a lot more literal. It's great on the numbers. It doesn't really gimme much else, whereas GPT will naturally give me numbers plus the y plus the solution. Which is really useful, and I would recommend when it comes to data analysis that you do use both platforms. The reason being, when you're doing extensive data analysis, and I'm talking like probably would've taken six weeks for manual human analysis of the data set that we looked at to actually be undertaken. You don't. You, you rely on the accuracy of it. 'cause you think, well, it can't get that wrong. But because I'm using both side by side, I'm like, oh great the numbers are the same or I will challenge it. I'll put the output of ChatGPT into Copilot and go, is this accurate? And it'll be like, yeah, it's totally accurate. So it was a really interesting task. Like I say to of my peers and certainly the employees within Opus, you are what you are when it comes to LLMs. We can't make you use a certain platform. Everyone has a preference. They are all different. I'm very much GPT orientated, although I do need to step out into Claude 'cause everyone keeps telling me I wonder it is.

So I need to move away maybe. But I think, yeah, it's user preference. At the end of the day, if you're a data analyst, crack on Copilot's probably the best thing ever. Gives you the literal thing. Whereas as a marketer and certainly as sales, we wanna know why, is that happening? When does that happen?

What can we do about that? How do we overcome? And it honestly, it's, it was actually incredible. What, what came out of that and the depth and understanding of that data via an LLM. Yeah, I don't think you could have done that without it.

Alex: You predicted my next question then actually, because, oh no, we started this conversation last time we spoke, around Claude. So I'm very much like you have been using GPT for a long time. I feel very, established in it. I've got lots of custom GPTs that I, that I work with and I am in a similar place to you.

Claude is the darling of the moment. Um. Do I hold my nerve and hope that ChatGPT leapfrogs it? Or do you start to make that move over? And that, and that's really just for me personally. For you, obviously it's for the, for the business more, it's, it's a reasonably big change. What's your thinking there at the.

Nikki: So I did this. I dunno if anyone's used GPT at there, which is their search engine. I did it just before Christmas. I dedicated a whole day. I was like, right, I'm gonna, I gonna made the switch. By the way, I don't know anyone else that's made the switch to that. I was like, I've gotta do it. Curious. It's brand new.

Let's, let's get on. And within like two days I was like, I can't do it. It's not for me. It's just not with Claude because of all of the stuff that I am hearing. I am also very active on LinkedIn. I feel like if I don't go curious in that area, I think I'm gonna be left behind. I don't, I don't think I'm doing myself. The best service by resisting it. And, and I'm certainly, not resisting it personally. It's just a time thing. You know how you get comfortable and you're like, there's my trusted platform. Why else would I explore? So what I am gonna do next week is actually dedicate one day I'm just gonna sit there and watch a load YouTube videos on Claude. There's a couple of influencers on LinkedIn that bang on about it all the time, and I'm just gonna see what I can actually do. And see if it's for me. And if it's not for me, then fine. But there will be a place for that. I think, you know, just going back to that Copilot being more literal data-driven, from my perspective ChatGPT being a bit more strategic, if Claude is something else and you are using a combination of all, then I think that would be incredible.

Like I'm intrigued.

Alex: Yeah, I think, I use Claude, but I don't feel it's embedded into. My way of working in the same way that ChatGPT is, and I think a lot of people are doing that. They're, they're using Claude for most things and using ChatGPT stuff for some things. And particularly if you're in house somewhere, you might be using Copilot.

But it becomes a kind of a balance, doesn't it? Of how deep do you go into any one platform and what information do you put into which one or I think. The interesting thing is the direction of travel that certainly Claude and and Copilot have is the integration with your systems and your existing platforms and your data, and I think that's the.

Nikki: I think, that's where I think we're missing a trick because ChatGPT at the moment, it can automate most basic tasks. I haven't explored, integrations in terms of GPT integrations yet. Whereas Claude, from what I've read, do that. So all of a sudden you've then got the GPT turning into the agent the genuine agent that acts whilst not prompted. And I feel that's the next step for me, but also for us as a, potentially as a business. But I, yeah, I'm like, I'm so new to it. I feel like a total beginner on Claude. If I delved in, I could probably pick it up, but it's just, yeah, it's a prioritisation thing, isn't it? It's speed. Doing more with less prior.

Like which things are you prioritising? Should you learn all five platforms? Absolutely not. Jack of all trades, should you have one that has two contributing, three contributing, depending on the need and the use case? Probably yes, I would agree.

Alex: So apart from Claude, what is next for you and for Opus?

Nikki: That's a tough question, you know, because I like, I mean the art of possible the thing that I always in the back of my mind think is. If we could on our list, there's a company in the states that's managed to do this. They had something like person marketing team and they cut it down to 10 and they've doubled their output using agents. It was very bold move. Amazing. And they're obviously incredibly early to be able to do that right now, but I just think like what could, what? Yeah. What is the art of possible you get this right. What could you achieve? Mm-hmm. And I can tell you now as a marketing team, we could achieve a hell of a lot more than what we're doing now.

But the thing that excites me the most the data and the learnings and how, and understanding how people do buy. And what is the role of the website in two years time, five years time? Is it still there?

Alex: Hm.

Nikki: Who knows? I'll predict No. But then how will eCommerce work without the website? Do you just build an integration to LLMs?

I'm, I'm just intrigued. I'm fascinated by the speed as well as to what things can be achieved in such a short period of time. When it comes to jobs, I wanted to just touch on this 'cause this is sometimes where you find resistance with AI. I would say if you're in a job, and certainly for me, like, you know the minute AI came out, I was like, if I don't adopt this a marketer, I'm pretty much on my way out.

Like this is early retirement for me. If I did adopt this. What could be achieved. So that's the sort of mentality that I would encourage people. If you're in a, if you're in a role at the moment that's process driven, feeling threatened by AI, that's your massive red flag to make it work and move into another part of the business or a higher level of the role that you are in right now. Think, yeah the opportunity for doing such thing is, you know, endless. But it takes a certain curiosity to do it, and I would just encourage people to take that step. Right now,

Alex: Good advice.

Nikki: Idea what my kids are gonna do, by the way, when they're

Alex: No, no. We often have that conversation.

Nikki: I mean, one is desperate to be a pilot and we thought, I dunno, actually, I'm sure there will be a need for a human on a plane, but I'm not sure you're gonna be driving it. And yeah, and the other one who knows. But again, where's AI being taught in schools? It's not. So I think there's gonna be that this mass problem in a few years time where they're all coming out of university and college it minimal knowledge all of a sudden the industry is like way advanced in it.

But the gap is huge. So, yeah, I mean, trying to get my kids into AI. Is something that I try to do, I want to do, I want to set for YouTube channel all the things you can do to enhance your homework with AI. That's actually really beneficial. But have I got time to sit and do that every day right now? Absolutely not. But if I was clever, I could probably build a little agent that would do it

Alex: That's what I was gonna say. Yeah. Yeah. That, that's a, that's a whole nother, podcast episode there, I think of how you're educating children with AI.

Nikki: Actually know one person that has started to do this, and he was the one that inspired me. But, yeah, I mean, look the opportunity is endless with it. There are resistors as there are adopters. Everyone is who they are. I just think it's all about identifying use cases. Failing fast is the way forward,

Alex: Exactly. I'm gonna move to the quickfire questions now.

Nikki: John,

Alex: What would you be doing for a job if it wasn't this?

Nikki: So this might surprise you. I would be a florist or a professional showjumper. There's quite a contrast between the two of those things. But yeah, it's always been the dream. It's just. It's just never worked

Alex: Both fairly safe from from AI as well.

Nikki: Actually. Yeah, maybe. I knew that all along. See if it all goes wrong, I'll be making bouquets.

Alex: And what's one thing you do regularly to protect your mental health?

Nikki: So I have a horse. I ride three or four times a week. And I've really prioritised that. I call it something re, it's in my calendar, it's called Recharge Time because it can get really intense. At the start of my AI journey, I felt incredibly overwhelmed, didn't know where to start. Everything was moving so fast. Felt like I needed to be at the edge of the curve, but I was right at the start. So, yeah, I ride regularly and take breaks. It's really important for the old brain.

Alex: Nice. And in the grand scheme of things, does AI excite you or terrify you beyond your job and everything when you look at society?

Nikki: It excites me in general.

Alex: Good. Thank you, Nikki. That's been a really insightful episode, so I really appreciate you coming on.

Nikki: That's okay.

​