From Hype to Habit: How Accent Group Operationalised AI

BELINDA LLOYD: All right. Welcome everybody. It is an absolute pleasure and privilege to be here at the Amperity Amplify 2026 event. This session today is "From Hype to Habit" and how Accent Group has operationalized AI. I will be speaking in my sultry tones today as I've lost my voice. My name is Belinda Lloyd and I'm the customer success lead for the Australian and New Zealand region. I've had a front row seat to some incredible transformations from our enterprise customers down that way, including the awesome Accent Group. Real magic happened recently when we rolled out the Amperity AI tools for the teams at Accent, and it is my pleasure to introduce the team that is making it happen. I want to make a note that Accent Group is currently the biggest user of Amperity AI tools by far across our global customer base — so absolutely the right people to be hearing from today. Amanda, why don't you kick us off and tell us about your role in waving the magic AI wand at Accent Group.

AMANDA DOAN: Hi everyone. I am Amanda, group product manager of customer data and digital at Accent. I was there at the beginning of Amperity being implemented within the business, and now we are also using the AI part of Amperity. It's been an amazing journey.

BELINDA LLOYD: Amanda is being humble — she's an absolute champion and advocate for Amperity and does wonders with customer data at Accent. And Julia, you are the one making it happen for the brands every day. Can you tell us a little bit about your role at Accent Group?

JULIA WOOKEY: Yeah, I'm Julia. I am a CRM and loyalty manager at Accent Group on one of our brands, Platypus Shoes — which if you haven't heard of, I wouldn't be surprised, because we don't operate in the US. But it's one of our multi-brand retailers, stocking the likes of Nike, New Balance, et cetera. And I'm in the tool every day, so yeah.

BELINDA LLOYD: Julia truly is an internal hero at Accent Group for making this data come to life via AI. In fact, the story of why Julia is here on stage is an absolute cracker and we'll get to that a little bit later. Now, we are here today because the transition from AI experimentation through to true operational impact is what we are seeing as a missing link in many modern enterprises. While everyone is talking about the hype of AI — and we had our beautiful keynote on where AI is headed and what we can do with it — it's about how we are making impact now and bringing real, measurable business value. It's also about the habits we're building into workflows and developing internal heroes. But before we look under the hood and talk to Accent Group, I'd like to share a video from Claire Best, who is the group head of customer data at Accent Group, on what she thinks of the Amperity AI tools. Claire could not unfortunately join us on stage today, so this was the next best thing. Let's see how we go getting this video started.

CLAIRE BEST (video): I was speaking with a member of our incredible CRM team yesterday, and she was saying that people on her team who are fairly new to this space are able to hop into the tool and type something into the AI and find just what they need when they need a number, just like that. I think it's always going back to customer first and how we can utilize efficiencies to bring that customer-first mindset into everything we do. All of those AI efficiencies are going to allow us to do that, because we are just customer-obsessed across so many business units. It's taken us beyond silos and I think that's really critical for us. It's transformed how we look at and think about data — we've almost gone from a siloed channel view to a customer-first view. I really want to start exploring how we can activate insights in the moment, and I think that's going to be really critical for us — whether that's looking at things like the real-time tables functionality and really leaning into that when we can.

BELINDA LLOYD: Julia alluded to Platypus not trading in the US, so let's take a look at who Accent Group is to the Australian and New Zealand market. They have over 900 retail stores throughout Australia and New Zealand, 20-plus retail banners, which equates to over 32 e-commerce platforms where their customers are buying either online or in-store. That's billions of dollars of revenue spent every year on apparel and footwear. In fact, one in three Australians and New Zealanders will engage with Accent Group in their lifetime. That is the scale of the business and what these two operate in every day. So we are now going to talk about how we take AI hype and move that to habits within workflows, and develop internal heroes — all to get to business value.

So firstly, when we look at hype to habit — we know that everyone is talking about AI right now, we're all talking about it today. But very few organizations have actually embedded it into how they operate and into their everyday workflows. Amanda, when did Accent Group know that this wasn't just experimentation with Amperity AI and that you wanted to operationalize it across the business?

AMANDA DOAN: Yeah, we realized it was something we should really focus on once we started seeing speed to value. We were trialing it with just a few select team members — the business has 30-something brands, so we did have to select a few people. Julia was one of them, and it had an immediate impact. That's when we realized: let's dig into this and focus on it for the business. We then rolled it out to the rest of the brands and everybody loved it. The other thing is that it was actually really easy for the teams to adopt because it had this really easy ChatGPT-like interface. And aside from that, at Accent Group we had a lot of support from the leadership team to implement AI into our day-to-day workflows, which was very important to making this work.

BELINDA LLOYD: That was something I noticed about Accent Group immediately in the partnership — how supportive the executive team was. The Accent Group board had spoken about AI and wanting to operationalize it, so it really was top-down support for the initiatives we were running. So many boards are talking about this now, and I'm sure many of your executive leadership teams are as well. There is a big gap right now between companies experimenting with AI and those who are actually gaining a competitive advantage from it. Where does Accent Group sit in that spectrum today, and what have you done differently to get there, Amanda?

AMANDA DOAN: I would say we're in the early stages of gaining that competitive advantage. We made sure it wasn't going to be a siloed initiative — with an umbrella as big as Accent Group with 30-something-plus brands, we needed to make sure all brands were on board. We trialed it with a select few and then rolled it out to all of the brands. We focused on embedding it into workflows for the marketing teams early on — using the AI to gather insights, analyze the data, develop a strategy, and then execute that strategy, all within the one platform: Amperity CDP, the AI, and then execution and activation. And again, lowering the barrier to entry — that ChatGPT-like user interface makes it really easy for teams to use.

AMANDA DOAN: And they're able to use natural language, which the teams really took on board and were really excited to see how easy it was. They thought they needed to have more knowledge to be able to use the AI, but it was so easy to pick up — and there were questions they didn't think it would be able to answer, and it could. As someone mentioned earlier today, the more you prompt the AI, the more it learns your business as well, so it's getting smarter as we use it. And with that easy adoption, you didn't need to have highly technical skills either. Very easy to adopt.

BELINDA LLOYD: You talked about the more you prompt AI, the better it gets. We've got some more information on prompts coming up and we're going to show you one shortly. But first, let's start at the beginning. Julia, what was the core problem you were facing as a CRM manager in your day-to-day that meant you were looking for these tools?

JULIA WOOKEY: Yeah, for us it was definitely the complexity and fragmentation of our dataset. Belinda mentioned we have over 900 retail stores — Platypus alone has almost 200 across Australia and New Zealand. So we're dealing with huge amounts of offline transaction data as well as online purchase data and engagement, all sitting in separate siloed systems. Stitching that together had historically been really difficult and created a lot of reporting bottlenecks. But the AI has been able to unify all of that information so that we can really take that customer-first mindset to how we're generating insights. It's also created a culture of data-led curiosity, because it is so user-friendly — anyone can get in there and ask it anything, which makes getting into the data feel like something that isn't difficult to do.

BELINDA LLOYD: I love that — a culture of curiosity. And you talked about data bottlenecks. Why was business as usual just not cutting it anymore?

JULIA WOOKEY: We operate with a pretty lean team. There are probably four or five equivalents of me across the different brands at Accent Group, all asking questions to one centralized data analyst. When everyone has questions coming from leadership as well as their own questions about customer, stores, and marketing performance, it really does reach a bottleneck. We needed a way for marketers to self-service those insights, which is what the AI really gave us. And on top of that, a lot of the data we used to look at historically was transaction-based: what products are selling, where are they selling, how is that trending over time, seasonality. Which is fantastic — but now we can look at it from a customer perspective: how a person is engaging with us over time, across our ecosystem, across multiple brands. That has been a huge unlock for us as well.

BELINDA LLOYD: A hundred percent. Those bottlenecks meant you had to wait in the queue every time a request went in — which is why business as usual wasn't cutting it. Okay. Many organizations are trying to layer ad hoc AI tools across fragmented data, and we often hear that AI outputs are only as good as the inputs. We heard Laurie from Alaska Airlines say "junk in, junk out" and it still very much applies to AI tools. So how critical was it that Amperity had that unified data foundation before you even typed your first AI prompt?

JULIA WOOKEY: Very, very critical. Even the most perfect prompt isn't going to compensate for gaps in the underlying data. Our customers interact with us not in silos — they expect the same brand experience and see us as the same brand regardless of the store or channel. The underlying data really needed to reflect that reality. If we were prompting a system to give us information on our customers as a whole, based on a fragment of the data, it wouldn't be giving us reliable outputs to inform decision-making and execution. Having Amperity was really why the AI worked for us — because we knew that every insight and every metric we were getting was based on our own validated dataset that we already trusted. So yeah, super important.

BELINDA LLOYD: It's like Gregor mentioned with the CDP — your customer dataset needs to be governed and configured correctly, and you know it's the source of truth. Fantastic.

So we've moved from hype to habit in a few steps. It was really important to identify core challenges and use cases where value could be seen for the teams. Leadership support was another key element, as was the unified data foundation, and of course a low barrier to entry. Julia and Amanda both touched on prompts, and we have one to demonstrate here today. Let's load up the AI instance where we'll be using the customer data assistant across the Accent Group database. We're in demo mode. Julia, do you want to narrate and talk us through what we're seeing here? This is one of your incredible prompts.

JULIA WOOKEY: Yes, so this is a real use case. It started with a really broad question I had for a project I was working on — I didn't really know where to start. We knew we had the data, but we asked on a really broad level for the AI to explain our three most important personas. It's thinking, looking up all the data and referencing it, and straight away it comes back with really comprehensive analysis — demographic, psychographic, transactional, everything. The depth of insight straight off the bat is amazing. But what's also amazing is how it outlines what it did and then also summarizes it. From there, we dig a little bit deeper — we can see how those customer groups are changing over time and also get it to come back with financial-based KPIs, sizing up the opportunity we have in those segments. For anyone in conversations with leadership about where to invest time, energy, and marketing budget, it's always good to have the numbers and financials to back that up.

JULIA WOOKEY: And it ranks them in terms of revenue impact so we can help assign budget. All of these things are obviously things we would sense-check ourselves once the AI comes back to us, but it's a really good way to go back and forth and understand: where can I start, what's a good jumping-off point? They were talking this morning about segment fatigue, and this is a really good way to get out of that and get out of your own head. It's kind of just like chatting to a team member — it always helps to bounce something off someone. So that works here too. In a few prompts we've basically gotten an entire customer analysis that would have historically taken a paid agency or an analyst hours to come back with, and it's done within a few minutes. The real shift for us though was what happens next — it can actually create the segments and then activate them downstream to different channels. You've got it coming up with who to target, telling you where to focus, then building the segments and pushing them to your activation channels.

BELINDA LLOYD: We had so much fun creating the base prompt, didn't we Amanda? Amanda was actually part of that. Do you want to tell us a little more while it's creating the segments?

AMANDA DOAN: Yeah. We're also teaching the AI Accent Group context. In the back end of the AI, we've added in prompts — its voice, essentially — and also how we use terminology in our business: our brand structure, our data structure, acronyms we use. We're constantly adding to this as well. And at the same time we've made it fun with its own tone of voice.

BELINDA LLOYD: We did.

AMANDA DOAN: We call it our data bestie, and it's got a little personality. I love what I saw — it's like hanging out with another colleague.

JULIA WOOKEY: Yeah, it's fantastic. From an execution perspective, you've seen how it helps with strategy and guidance, down to creating segments, down to executing those segments. And because it's all housed within Amperity, you can close out the loop by measuring the impact, seeing how those segments behaved over time, and optimizing. So instead of just a static report or a set of ideas, you've got a full end-to-end idea-to-execution-to-measurement solution.

BELINDA LLOYD: We also prompted in for each brand what destination needs to be sent to. If the user says "activate to paid media destinations" — which is what we've just said — the journey will be built and Amperity will understand what that means. You'll notice they said "send it to Platypus Facebook paid media" and it will know which destinations within the platform that refers to. Also: send it to TikTok and Google, which you've all got set up with caps and everything — fantastic.

JULIA WOOKEY: Yeah.

BELINDA LLOYD: And we close out with a bit of tone of voice from the prompt as well, right?

BELINDA LLOYD: So we said "say hello to the crowd at Amplify." And you see fire emojis, cool faces, sneaker emojis — "Accent Group is seriously cool, bringing the heat with iconic brands, sharp customer strategy, and a proper strong sneaker game culture." Love that. So what we are seeing here is that prompting isn't just a hack at Accent Group — it's a strategic capability layer that's being unlocked and embedded into the teams, truly upskilling them for the future and for agentic AI that we're seeing coming. Now we're going to look at some specific use cases being unlocked through advanced prompts. Julia, let's walk through these. First up is loyalty — tell us about how you're using the AI to redefine your loyalty program.

JULIA WOOKEY: At the moment, loyalty programs really do center around reward-based programs. We'll see if AI can make them smarter and more one-to-one in the future. But most loyalty programs now, as you all know, are reward-based, and a lot of the analysis that comes with that involves complex calculations: commercial liability, redemption rate, how people are returning over time. Because all of that data was housed in different places — some of it in spreadsheets, some in different systems — it was historically very difficult to pull any reporting on loyalty. That was always something we would have to brief in to our analyst. With the AI, I can ask what rewards are in circulation, I can have it forecast how many rewards will be in circulation in three months from now, I can have it calculate liability for me — all of those different things — so we can really quickly and effectively track whether consumer behavior is changing in the direction we want it to over time as a result of the mechanisms we have in place.

JULIA WOOKEY: And based on that understanding: are our minimum spends effective, are our expiry dates effective? All of those things. Being able to prompt the AI — which is something we couldn't do just through regular Amperity and segments, because it would have required query logic — we can now just ask it and it will tell us straight away. That actually helped us to remap how our loyalty program was performing and actually redesign parts of the commercial proposition going forward.

BELINDA LLOYD: Yeah, and there was real value attached to that too. So that was one use case we unlocked. We saw the "know your customer" or persona prompt in action — but tell us a little more about that and how other teams have benefited cross-functionally from having their personas evaluated.

JULIA WOOKEY: Obviously the prompt you just saw had been recreated for this use case — it was a real thing I had done, but we wanted to make it a bit clearer for everyone here. What you didn't see was some work done separately beforehand, where we used the AI to manipulate large raw datasets. It can be simple things like converting a birthday into a usable attribute such as generation, or determining whether someone's postcode means they're metro or regional. All of that kind of data manipulation to turn raw data into actionable segment attributes had been done, which was really useful for us.

On top of that, I had been given a series of personas by our buying and planning teams — the personas they typically look at from a merchandise perspective — and I recreated them as actual segments grounded in our customer data based on the brief I'd received. What we found through that exercise was that the buying personas, when we fed them in, were quite general and had a lot of overlap between the three segments. What we were able to learn from that was that we needed to get more specific and more data-driven so we could have higher differentiation between our segments, and then understand the right way to talk to each of those segments.

BELINDA LLOYD: And even create new segments — like those micro-segments that Laurie was alluding to with AI. Fantastic. And the final use case was reporting. Talk to us about how your reporting has changed now that you have AI.

JULIA WOOKEY: We used to work out of so many spreadsheets — lots of static spreadsheets, or reports that would get exported from random places on a Monday morning for all of our BAU reporting. I started with a lot of those really common questions or reporting aspects that we would do every week or every month and just tested and iterated to get the perfect prompt — so that in one prompt I could get the data I needed for those scenarios, for the most common questions we would get or the recurring metrics I knew I needed to report on. That saved me a ton of time.

But over and above that, I'm obviously not with the team right now — I'm here. And they've been able to take the prompts I've created and generate the exact same outputs as I would have, which has given them a lot more confidence in what they're doing without the person who typically does it there on the ground. It's made everything a lot more consistent and user-friendly for anyone on my team to pull that data. And as a business where senior leaders look after multiple different teams, consistency in how we're getting our outputs is really important. So it's helped streamline a lot there.

BELINDA LLOYD: Absolutely. And it's not just your team while you're away in New York who can do it — senior leaders at Accent can jump in at any time and get the same figures, consistently. Fantastic. And most importantly, all three use cases are helping to guide strategy and drive revenue. Now we're going to look at how we take those habits we've embedded into workflows and create internal heroes that can bring this to life within your organization. Amanda, driving adoption across 30-plus brands wasn't a walk in the park. How do you ensure AI is fostering internal heroes and developing measurable business value?

AMANDA DOAN: Firstly, we needed to make sure we were working with a CDP where we had a really good relationship. We worked with Belinda and the team to implement the platform, and from there we had weekly and fortnightly sessions for different aspects — we were pretty much working together all the time. Belinda's expertise on the Amperity platform, what it can do and what's on the product roadmap, combined with the fact that she understands our business because we talk about it constantly — that partnership was really, really important. The other piece we leaned on was the culture at Accent Group. With 30-something-plus brands, it can be quite competitive — some teams want to do better than others. Platypus wants to do better than Hype. So we leaned into that and leveraged that competitive nature, and we held hackathons — of which Julia won.

BELINDA LLOYD: Oh my gosh, this is the cracking story we wanted to tell earlier. So we had two hackathons. The first one, Julia and Amanda were in the winning team. We upped the ante for the second one and said: show us what you can do with Amperity AI and we will fly you to New York to be here today. Julia was the winner of that. She was only meant to fly over and take in the sights, but we started talking to her about how she'd done it and what she was doing with AI — and the gold that came out of that made it a no-brainer to have her join us on stage. Let's all give Julia a round of applause for being the internal hero.

AMANDA DOAN: Yeah, I just felt FOMO so I was like, I'm coming too. So we leaned into the "find your champions" model, and that's really worked out well. We'll be holding more hackathons at Accent Group — the teams find it really fun, and they're able to talk to each other and share strategies even though they're competing. That was really effective. And the office hours we held — I touched on that earlier — we bring all of the team members together, get feedback from them on what they want out of the platform, and Belinda comes to these office hours and teaches the team how to use the platform. There's that constant feedback loop: what do you want to see, and what have we got in the platform? That's been really effective. And what we do is make AI an opportunity to have fun and not a chore to learn.

BELINDA LLOYD: I love that — an opportunity to have fun and not a chore to learn. What a smart way to get more people involved with AI tools within your enterprise. The office hours really do feel like that. It's playful — we're not trying to bat anyone with a stick, we're saying: tell us what you want to do, play, experiment, have fun with this. And we always get fantastic results and cool use cases from across the business.

So we've spoken about Hype to Habit, and now Habit to Heroes. Here are the four steps we've covered on how to take something that's embedded in workflows and develop internal heroes. You want an enablement strategy — think of it in a fun way, what can you do within the enterprise to bring people on board? You want to offer a feedback loop. We offer that monthly in office hours with the team at Accent. We have them do live demos on the call and feed back on what they're doing. Develop heroes such as Julia — she is such an amazingly skilled person within Accent Group, and that's spreading to her team as well. And then we want to measure value. We haven't covered that off just yet. Amanda, what's the bottom line — is it just saving time, or is it changing the game for Accent Group?

AMANDA DOAN: I would say it's definitely doing both. We did the maths — Belinda and I sat in a room, looked at our campaigns and all the hours we might have saved, and we can see that we've saved thousands of hours with the teams on the manual creation of segments and on reporting and getting insights out of the platform. The AI has definitely saved a lot of time for our teams, which means they can spend less time doing manual work and more time focusing on strategy. They're less operational and more strategic now, which is what we were after. Faster decision making as well. Time saving in all aspects, and stronger campaign performance. We are always looking at ROI — I'm really excited about that ROCD thing we saw earlier today, I want to go to that session. Always focusing on ROI, running control groups, looking at incremental revenue. It's definitely been impactful and a game changer for Accent Group.

BELINDA LLOYD: That's powerful. And these hours are only from the AI initiatives. We've done further maths with the ROCD team — there are thousands of hours being saved from this one initiative, and nearly 8,000 hours a year being saved across the base from the CDP alone. Looking ahead, where do you want to see the AI go and what's the biggest untapped opportunity for your business with Amperity AI over the next 12 to 24 months?

AMANDA DOAN: I find that people usually think of a CDP as a marketing platform, but I definitely see the opportunity to expand out from marketing to more business use cases as well. We want to bring different teams into the product and using the platform. We're going to reach out to the product team, the buying team, and the loss prevention team, because we really think there's an untapped space there. Just having more of the business using the tool and the AI and getting them used to using AI in the business. Yeah.

BELINDA LLOYD: Brilliant. In closing out today, tell us exactly what everyone in this room needs to do to boost or start their AI momentum.

AMANDA DOAN: Reach out to your Amperity rep — I hope they're just as amazing as Belinda. And just get stuck into the tools. It's really fun, engaging, and the teams love it. So give it a go.

JULIA WOOKEY: I think just take your team along on the journey — ideally halfway across the world if you can. But if things are built and rolled out in a silo in MarTech, it can sometimes miss the goals of the end user. Having that open feedback loop, helping your teams see how it's going to save them time and make things they're already doing easier, and making it not a chore to learn — if you can do that, adoption will come really naturally. That would be my advice.

BELINDA LLOYD: Thank you. So — we've shown you under the hood of how Accent Group is operationalizing AI, how they're saving nearly 3,000 hours a year and making it a daily habit rather than a side project, and really making strategic gains and competitive advantage every day. Now it's over to you. If you're sitting here thinking this is a little bit of all right — this is your sign. Don't just let your data sit there. We saw earlier that only 26% of customer data is being used. Activate Amperity AI across your base today and start harnessing the hype of AI into powerful, repeatable habits to develop internal heroes that drive real, measurable value for your business. Thank you so much for having us, New York City. If you've enjoyed this session today, please scan the QR code and provide feedback. And please give Amanda and Julia a round of applause for flying halfway around the world to be here today. If you're here for the next session, please stay where you are — we will have Matthew, our VP for EMEA, on stage next. Thank you so much.