How to Drive AI Adoption - A Step-by-Step Case Study

Welcome back to the Digital Marketing Podcast, brought to you by target internet.com. My name's Daniel Rolls, and in this episode we have a step-by-step case study on how to drive AI adoption and embed AI into your organization. Okay, so AI generally hasn't actually decreased people's workloads.

In reality, we're doing what we've always done and now we're also constantly testing and experimenting with the latest AI tools and prompting techniques. So what this episode is really about is how do we go beyond this kind of pilot purgatory, uh, and really embed things into our every day. And in this episode, I'm joined by Emma Troon, who is Deputy Director of Marketing at Aston University, and looking at how she's taken a really structured approach to embedding AI into a complex organization with lots of different stakeholders and her step-by-step process and the lessons she's learned along the way.

What was interesting is I met em at the beginning of this adoption journey, and it kind of fascinated me how this process is essentially a really classic case of a really well structured digital transformation with a big focus on innovation, stakeholder management, and skills. So let's go over to the interview and learn from this kind of step by step process. Daniel Rowles: Okay, I am with Emma. And Emma, before we get into it, just give us a bit of context.

'cause your job, your background is maybe not necessarily typical of the person that's in charge of implementing AI in the organization. So how did you get into it and what's the kinda background of that? Emma: So, with the risk of going too deep, too quick, I actually had a bit of a setback at work. Just over a year ago.

I didn't quite get the job that I really hoped that I was gonna get, and I thought, how can I make myself indispensable. I. Emma: What can I do to my skillset? What is gonna be desirable, both for future employers and my current employer too?

Offer me a new better job really. And that's where I thought instead of doing dry Jan, I'll do ai, Jan. And I really just threw myself into everything I could do with learning ai attending webinars, experimenting and just kind of putting myself at the forefront of AI learning. Daniel Rowles: So let's go back to like March, 2025.

And that's when you initially pitched this AI task force. Is that right? And then what did that look like? Emma: I did.

So I went to I director and I said, right, I've got a pitch Mission AI Impossible . And I knew that there was a massive gap for upskilling our department, for implementing ai, for having some kind of guidance. I mean. This is, you know, a year ago now.

I think a lot of companies have come on since then. But really it was like the wild West. Everybody's running around using AI as different search engines and they just felt like there wasn't any kind of policy, any type of guidance and any kind of upskilling. So I took that opportunity.

I went to the director at Aston director of marketing and said, look. I want to align this to our institution strategy. I think that's really key. Make sure that you are showing that you are in tune with what your institution, what your company is looking to do over the course of the next however many years.

So we have the Aston 2030 strategy. I made sure that it aligned with one of the core pillars of that I asked for a reasonable amount of money. I think, you know, it's easy to kind of go in and hardball a really high price. But actually I asked for something that was totally reasonable under five grand, but the biggest ask was that I could get time and commitment from a small group of people to join the AI task force and to focus some of their work time on upskilling themselves in ai.

Which I then supported with as well. But that was really important. You know, I think a lot of people will ask for money when they're asking for it, putting a pitch forward, but actually the resource and time is key, particularly in this area. I demonstrated the benefits and the reasons why an AI task force would help the business, what it would do for us and for the workforce.

And I also just clearly mapped out how I was gonna execute the plan and what the tangible objectives were. So it was a real clear and easy yes from my director to, to support me in this initiative. Daniel Rowles: A couple of things there is that. First of all, this wild West, everyone doing different things.

I don't think that's changing a lot of organizations. I think it is still pretty chaotic because it's moving so quickly. We're in this kind of purgatory of forever trying stuff out, but never really embedding it in. So I think that structured approach is really interesting.

And then this piece about, actually, it's not AI for the sake of ai, it's, there is a business strategy and we're aligning how we're using AI to help us deliver on that as well. Now, it might be that the. The strategy changes because of ai, but actually those two things should be baking into each other. And I'd really love that you had that focus on skills and learning at the beginning, because unless you understand this stuff, how are you gonna apply it?

so what was the, the key purpose of the task force, what were you gonna try and do with it? Emma: So I had four objectives in this task force and actually reviewing these just last week and. Being able to celebrate that we have achieved all four of these was a massive win for me. So the first was to build AI expertise to equip the select staff with AI knowledge tools and practical applications.

Notice that is select staff and with this approach, I did start small. I wanted to make sure that I could give the right amount of support and guidance and time and then do a bit of a hub and spoke approach. So that was the first objective. The second was to pilot real use cases, so test applications in marketing, communications, student support and operations.

Specific to the role, as you say, it's so easy to spend a lot of time messing around having a play, and there's definitely space floor experimenting, but my ethos is. Always try and relate it to the work that you're doing. Otherwise, you'll end up doing double the amount of work. So the quicker that you can practice and pilot real cases rather than just having a play, the better in, in my opinion.

The third was to establish governance. So build internal confidence and profit governance and guidance around AI use in the department. Again, that, that's a huge one. I think.

Whilst AI's been established, and whilst a lot of organizations are deciding which tool to go with, which company to, to buy a package with, it's been very difficult for anyone to give any kind of kind of proper policies or guidance and be able to back themselves while it's such fast changing. So I can speak in a bit more detail about that, but I lay out some. Obvious examples of how to use and how not to use AI for the wider department. And then the final one was supporting digital transformation and contributing to the university's wider digital transformation strategy.

So that's part of the Aston 2030 strategy. Daniel Rowles: So you then went off, I guess in, this was like April time to recruit the team of people that you are, you're kind of dealing with. How did you go about that? Did you do anything kind of differently that you would've done normally?

What was the approach? Emma: So first of all, I would say you need to really build the interest and show the benefits to the individual. Don't talk about how AI is gonna help the company and what it's gonna do for the business. Let's be real.

Most people are working for themselves or to benefit their career to make money that you know. They want to know and they're going to be driven by how AI can support themselves on their own learning. So I started with a lot of stats and lucky for us the marketing area is one of the most heavily affected by ai. So, you know, it's quite easy to evidence how this is changing.

I dunno whether you feel this way, Daniel, but there is a lot of talk about skills of the future and upskilling students and students getting the right skills. But I'm sat here thinking, well, what about workforces that are currently in place? You know, we've still got 30, 40 years left of our career and we still need this kind of upskilling as well, so. I really went in on the benefits and kind of sold them the dream.

I also communicated in a different way. I didn't just send out another email. I made an avatar, which was a good use of showcasing ai, but also it's something different for people to see within a business and to get them excited than just another scheme that's opened with a Microsoft form. So there was also a Microsoft form, but it started with an avatar.

Daniel Rowles: So what were you really looking for in the people that you were recruiting? What is it that you you were particularly trying to focus on as part Emma: on part of task I wanted a range of different skills. I wanted a range of people from across the department not just people that I knew well and that I knew that I got on well with, but people that would be able to reach colleagues that I didn't know. So, at Aston University, I work in the department for marketing communications and also admissions.

Now admissions is particularly my area or my strength. I know a few colleagues there, but I really did want somebody who knew most of the admissions team. I didn't want a whole task force of yes people. I wanted people that were gonna challenge our thoughts that were gonna check for bias.

That, you know, in some ways at times were anti ai. And I think, you know, really having that spread of different viewpoints makes such a huge difference to your outputs. Daniel Rowles: What's fascinating to me is that all the stuff that you're talking here with stakeholder management is all the learning that we found when we were building our digital transformation strategy program at Imperial in the, that stakeholder management's key, and there's this quote from Alistair Wellum, and Alistair has contributed to the program over the years.

And these thing is like, you need to bring the naysayers on board because they're not negative. 'cause they want it to fail, they're negative 'cause they want it to succeed and they might actually have some genuine concerns. And if you don't bring those voices in, you just end up alienating those people. So I think it's really good that you kind of manage to bring those into the kind of task force as well.

So from there you built this task force, you've got buy-in from kind of leadership level. What did you then introduce, you did these kind of areas of focus, right? And what did that look like? Emma: So this is where I think my AI task force differs from most that I've seen.

There's a lot of different AI champion models out there for implementation where. I bought in areas of focus. So these are specific to marketing or kind of marketing admissions in our world. I split them into six categories.

Those categories are data and analytics, general work, support, copywriting, visual and creative, SEO, and web and social and community. So. By splitting those six, I had six Task force members. It meant I could align each task force member with one specific area.

And that just meant that they were more focused, they had a bit more of a direction. It also made it a lot easier for myself to triage when questions came in or requests, I knew exactly the right person to look at that because it fit in their category. So I've had a lot of feedback that actually that's really helped their focus and helped them make it more bite-size and manageable Daniel Rowles: Right. Okay.

So, when you're going through and bringing those different taskforce members then into that, how do you kind of align that and then what support did you give them through that process as well? Emma: so. Wherever possible. And with my first cohort, there's a spoiler for what's coming.

With my first cohort, it was really easy to actually fit them into the specific areas of focus. They had those human skills to back up what they were doing with ai. And for me, that's absolutely key. We've all been there when AI's hallucinating, when it's telling you what it thinks you wanna hear and it's completely false.

So actually, if you have humans who can understand immediately what good looks like and what isn't correct, you've solved half the battle there. But also what I found was if say I put a task force member into Seo O and Web who knew nothing about SEO and Web, that's almost double the amount of learning for them. 'cause they've gotta learn all about what SEO and web looks like, what's correct there, as well as the AI side of it. It also helps their confidence, you know, that they think, oh, well, I don't know ai, but I do know copywriting.

I, you know, I'm a published author and I do a lot of copywriting, my role, so actually it's kind of halfway there in to, into what they need to know, Daniel Rowles: Okay. And then what about the support? Emma: So for each task Force member, I provided clear guidance. So first of all, what does that task force member cover?

I've got an example here for the copywriting. So to support creative writing, ensure it still sounds human with the correct tone of voice was kind of the overarching brief of that task force member. Typical cases would be supporting, writing a blog, looking at email copy, body subject lines, course page writing. Who is it relevant for?

Aligning it to different stakeholders, doing some analysis on tone. And then I also gave them a list of AI tools to explore. So Jasper ai, Claude custom gpt, and kind of wrote out a whole list. So.

It was that guidance and those, the steer of what they needed to do, but for them to go off and do that on their own. Daniel Rowles: And then how do you kind of keep them motivated through that cycle then? So you've kind of got 'em in place, you've given them tasks, you've given them that kind of support. What we've seen a lot of these initiatives is it's, we go back to the day by day and it's quite easy to kind of fall out and forget this kind of additional task force you're working on.

So how did you get 'em motivated? Emma: So the first thing that I ask them to do is book in lab time in the diaries. Like if you don't have that time in your diary, like you say it, it's so easy to just go back to the business as usual and not continue with your, you are learning, you're practicing your upskilling. I didn't specify what that lab time needed to look like, whether that was webinars, whether that was experimentation you know, reading news, whatever it looked like for them, just.

They actually did dedicate some time to it, and that was kind of written way from the beginning when they were asked invited to come on board, so it wasn't a new thing that they needed to add into their diaries, and they knew that was a commitment that they needed to make. But it was also very loose and there was no expectation just that they spent the time on it. Daniel Rowles: And when you call it lab time, did they have a real clarity, like lab time was back, experimentation, or you know, how was that defined for them?

Emma: Yeah I explained that it could be experimentation, it could be meeting up with another task force member. It could be finding a webinar and doing a catch up. Anything kind of within the AI space that's helping support their development for me would class as lab time. The other thing that we would do is we'd have a working group every other week and actually meeting at the working group and just doing a quick round robin of what they've been up to what's worked for them, what hasn't worked for them, and sharing.

Sometimes you're actually doing, you're doing AI without realizing you're doing ai. And when you're forced to think of, oh, what was the latest thing that I have worked on that helps you realize how far you've come. But it also really helps teach others you know, what works and what doesn't work. Daniel Rowles: I mean, it's interesting for me, what's really coming out from this is the amount of structure that you've put around this as well, so that there is there's the freedom to experiment, but there's rails around this and you're kind of giving a very kind of clear structure, so.

You kind of keep them motivated. Then when you'd got to, so this took you through to about September time. What did you then do? You had like a fuller launch, a wider launch later after that.

Emma: Yeah, so we've gone all the way from March when I launched Mission AI Impossible all the way to October now, and I wanted to really take this time with the task force to allow them to develop their skills. I made it very clear when launching the task force that you didn't have to have any experience of AI to join. So that meant that we did have a range of skills of people who came on board, and I was very conscious of. Launch into the wider department when everybody was ready.

Because if you had people coming to ask for advice when they didn't feel confident and they weren't, you know, perhaps ready for that next step, then that wouldn't help. That wouldn't help anyone. So. I had meetings with everyone in September, made sure that they felt like they were really then in October I launched to the wider department, a live q and a with guidance examples as well as introducing to them their task force members and what areas they looked after.

So before then, the rest of the department had absolutely no idea of all this work that was going on behind the scenes. But I think actually building that framework was really important for us to. Feel confident for the next step. Daniel Rowles: It was the idea then that the task force would then just go in and support people on the particular topic area that they'd been kind of assigned to.

And that was how you'd embed it throughout the organization. Emma: Yeah, so I mean, October, it's pretty new when you're factor in the Christmas break. We've not got too far into this area, and I would say this is one of the parts that, that we will be working on and focusing on for this next year. But the idea is that, queries will come in.

I think a common situation is people know that you can use AI for a certain job, but they just don't know how. Or they might know that a job is too long and monotonous and they actually have no idea how they could use AI with it. But they see that there is potentially an opportunity so they can come to anyone in the task force. Usually people will come to me and then I will triage them to the right task force member, or I'll speak to the task force member and maybe pair them up with another.

Another common thing that I saw was task force members themselves coming to me with an idea and saying, okay, I've identified this. And that's where they really shone and where I was really seeing the kind of fruits of the labor. Then I'd match them up with somebody else in the task force that I knew was working on something similar that would be able to help support them with that. And kind of, I guess it's a bit like a jigsaw fitting together all the puzzles and seeing how each other can support each other, and that is one of the key roles of the task force leader, I'd say.

Daniel Rowles: So we've, we are very nearly then I guess that timeline a year into things. So what's worked and what hasn't is probably a good learning as well. Emma: Definitely. So what has worked is the expertise of the task force members.

They have come on so far from the beginning, and really, like I said, you know. Upskilled themselves, identifying ways, being innovative and sharing what they're learning. They're doing brunch and learns. We have the teams channel and all of these initiatives are really kind of showing where their expertise have grown.

I say also kind of seeking out opportunities to collaborate. I've linked up with the University of South Hampton and we have a bit of a buddy scheme going now between their AI task force and a our AI task force. And allowing that space for experimentation with no defined expectations is, yeah, that all of that has really helped support their growth. Daniel Rowles: and what hasn't worked.

Emma: Well, I'm sure a lot of people can relate with this, but system integrations has been a real challenge. I think there's still such a mistrust and hesitation, and as I alluded to earlier, our. Digital services department hadn't quite committed to who we were gonna partner with. We have now, but it's not been announced yet, so.

We've had this last year of working out, okay, well which systems shall we use for now? For, for what things? And actually we have done a lot of activity, but it's more, been more outward facing. So it's been things that are super low risk, you know, competitor analysis where you can set up an agent to go and do that for you.

But actually it's all information that's widely available on the internet, so there's no risk there. But it does feel like we're kind of scraping the. Possible. We're just at the tip because there's so much more we know we can do.

If we can get those system integrations in place, Daniel Rowles: Yeah, I, this is a huge issue ever at the moment. 'cause all the legacy systems are in place. There's huge risk adversity in a lot of tech teams that are really worried about this stuff. And that's a natural tension I think between, we want to embrace everything, but we also wanna do it in a risk.

Mitigated kind of way as well. But the thing is you don't get the true benefit until you've really embedded that stuff in. But I always kind of go back to, you know, your average digital transformation has taken two or three years, whereas the problem is this technology is moving so quickly that you're kind of seeing we're missing opportunities unless we're embedding this in more quickly as well. So I think some patience with that sometimes is hard to develop, but it's an important part of it.

So, so based on that, what's next? Where's the task force kind of going in 2026? Emma: Well, I say one of the other challenges was keeping the wider department developed and engaged and coming to us you know. I purposefully kept it a small group so that I could manage them.

You know, this is on top of my usual role of being deputy director, marketing and student recruitment. So it's not like I can dedicate my whole working hours to this. But I do see that there is a need and a desire for the wider department to be on this journey with us. So I launched cohort two.

We are doubling up on the areas and also this year. Because we're now implementing systems as of next month, we'll roll them out and we'll have, be able to have much more control over, lots of different areas and lots of different systems, but. I'm also introducing three kind of work streams because I think that these are the three that I've identified as the biggest opportunities to really turn that dial. So the first one's personalization.

I think if we can nail personalization at scale that'll be a huge win for us. The second one is data. Again, you know, we've done bits with data, but because of needing that closed circuit AI platform. We've been hesitant, and I know that there's so much more we can do there.

And then the final one is custom builds. We've got a real whiz on our team this year. I'll do a shout out to Kian, who is just incredible. I go to him and say how could we do this?

And he's got the technical capabilities to do the backhouse of the automations to build and add on to a website. So. He's like a bit of a secret weapon at the moment, but I will say all of the task force are fantastic and they've all brought so much to the table. Daniel Rowles: I think that's fascinating because what we found is that all this talk that, oh yeah, it's gonna AI's gonna replace developers, we don't need developers anymore.

It's nonsense at the moment because there's still so much technical difficulty in actually doing something really effectively. But the secret power is if you've got AI and a team that are embracing that, and you've got some technical people. That can do the development, things like that as well. The combination of the two is huge because it's allowed us to deploy AI driven content to our website, which would've been quite tricky to do otherwise.

So I think that's really important as well. So it's quite interesting to see that, how that allows you to experiment a little bit more as well. Emma: I'm not in the camp of ai. Gonna take all of our jobs.

I. I do think that people need to embrace it and learn how they can make themselves better at their own job with the use of ai. That's the way that we need to go and, you know, look at social media. I dunno how many years ago now, but everyone thought, oh gosh, that's nothing.

And we won't need to learn that. And actually now it's just a core part of marketing and our everyday, everyone's everyday is life not just marketing professionals. And that's why I'm really fascinated on how companies are implementing and embedding AI into their workforces because there, there's no other time that we are gonna be experiencing like we are now. I.

Daniel Rowles: Yeah I think absolutely right. And we just, before we came onto this recording, we were just talking about the fact there needs to be two camps at the moment. There's the. Everyone's job's gonna go.

It's gonna be a, you know, an absolute apocalypse and AI's gonna create this absolute carnage in the jobs market. And then there's like, no, it can't do that much. And it's not that great. And yeah it's not quite as impressive.

And I do think there's somewhere in between which is saying it's gonna radically change what we're doing. Doesn't mean the jobs are gonna go, but we do need to really embrace it. And it will get rid of some jobs. There's a bit in between.

So I think it's really how we embed that in now that becomes really important. And it's that idea of just baking innovation into everything. It's the whole piece that actually we need to constantly innovate and if we don't, that's a huge risk and learning is part of that. Emma: Absolutely.

And if you are listening and you are in a business where you are thinking, God, we're not doing any of this, or, no, nobody's, you know, teaching us anything, or upskilling, take the ball by the horns, and you be that change maker, you become that leader in that AI space. Like you can absolutely do it and it's there for the taking. Daniel Rowles: Well, it's interesting to me because when we spoke about this just over a year ago, this was really early stages for you and you know, your confidence is building this stuff and you're being asked to speak at conferences all over the place now to talk about your kind of journey with this as well.

So it does go shape, if you embrace it forces you to learn. You don't have much choice. 'cause if you're gonna go on stage and talk about it, you better know what you're talking about. We've had conversation with you saying that I.

I've got my core job, you know, there's a lot of distraction potentially, if I'm not careful from that core job by speaking at conferences. Is it a good idea? And I've always found that if you have to go and teach somebody something or stand on stage and speak about it, it really forces your hand to learn about it as well. So I'd be interested, what's your learning from all this as well?

Oh. Emma: Yeah, I mean, you did give me really good advice. It certainly builds your confidence and does force you to learn. I think it's really easy in the world of AI to get imposter syndrome and I know we've spoken about that as well.

It's so new and it's ever evolving and you could have a month off and already become out of date. It's a wild ride, but I think the more that you realize and understand that we are all learning because of that, because of the rate of change, we are literally all constantly learning. , That's what. Be at the forefront of this, but it also means that there's opportunities for everyone to become an expert and to become a change maker, and I think that's really exciting.

Daniel Rowles: I think it is. I think that's brilliant. Well, Emma Johnson, thank you so much for joining us for those fantastic insights. If people wanna get in contact, what's the best way of doing that?

Emma: On LinkedIn, so Forward slash emmason. I'm also starting with Sally from Southampton, a AI in he marketing. Collab group. I think at the moment the working title is collab lab.

There is no kind of monetary agenda from it other than to just share our experiences, find out best practices, see what other people are doing in other institutions, what works, what doesn't work. So if there is anyone out there, particularly from outside of the UK that works in education marketing, please do get in touch on LinkedIn because what up? Daniel Rowles: Brilliant. We'll put those links into the show notes, so Target internet slash podcast Emmason.

Thank you so much for joining us.