Published: 26 May 2026
Summary
We talk with Lynda Petherick about what it really takes to become a data-driven retail business, from board-level understanding to frontline enablement. We dig into how New Look thinks about ROI for AI, modernising core systems, and building a customer model that drives smarter decisions without losing sight of privacy.
- Defining data-driven retail as enterprise-wide transformation
- The five pillars that underpin business change
- Why executive teams cannot delegate data and AI understanding
- Prioritising AI and data investments with clear ROI
- Managing risk from AI agents, hallucinations, and human checking costs
- Applying AI beyond merchandising into cybersecurity and operations
- Improving inventory visibility to reduce unnecessary ordering
- Enabling store colleagues to create moments of delight
- Modernising core platforms such as POS and commerce systems
- Building an advanced customer model with granular segmentation and practical cohorts
- Balancing personalisation with customers who prefer less engagement
Transcript
Host: 00:02
So welcome back to What’s in the Box. I’m here with Lynda Petherick, who is COO and CIO. That’s a brief and a half. COO and CIO at New Look. Lynda, thank you so much. Okay, welcome.
Host: 00:21
Lynda’s just come straight off stage here at the Retail Technology Show. And it was a fascinating talk, basically all about what it means to be a data-driven retail business. So maybe that’s a good place to start. What does it mean? And I know you know we’re going to talk for about 15-20 minutes. You could talk for probably 15 hours on that topic, but what does that mean to be data driven in today’s world?
Lynda Petherick:00:48
I think you know, ultimately to me, we need to be data-driven because that is the way the world is shifting. I mean, we’ve been disrupted as a sector, as retail for some time now. And for me, if you are truly embracing what I see as a revolution around data and how it can transform both your internal business process, but also the opportunity it presents to understand your customers fundamentally differently to move the top line, then the reality is that you’re going to be your competitors will.
Host: 01:29
Yes. Now I noticed that when you came off stage, you had quite a long line of people wanting to talk to you. There were a few, weren’t there? Was there retailers in that? Because I’d imagine that retailers would be very interested in learning from you knowing what you guys at New Look at are doing at the moment.
Lynda Petherick:01:50
Well, I think I’d say a couple of things. I’d say, yes, there were some retailers in the queue to do some knowledge sharing in terms of follow-ups. But I think also, there was obviously a slide in my presentation that talks about the five pillars that we’ve been going through that really underpins what our business transformation is. And I actually would think of it it’s more accurate to say it’s a business transformation, almost it’s data driven, as opposed to just a data-driven transformation because it’s kind of truly enterprise-wide. But certainly those five steps and those five pillars. I saw the phones move into the air to take a photo of that particular slide.
Host: 02:27
Yeah, that was a very, I wouldn’t say Insta slide, but yes, the phones went up at that point.
Lynda Petherick:02:32
I thought I might as well cue it because I thought that was what was about to happen.
Host: 02:35
So that slide is going to be reproduced in a lot of offices.
Lynda Petherick:02:39
I think so.
Host: 02:42
Now, one thing that you said during that, which again, this is all part of the culture of being data-driven.
Lynda Petherick: 02:51
Yes.
Host: 02:52
And you said, looking at his notes, delegation will lead to oblivion.
Host: 02:59
Tell us what you meant by that. I think it was in the context of the challenges and also the executive team.
Lynda Petherick:03:06
Yes, I think I mean what’s been really important to us is that everybody, at whatever level you are in the organization, and that includes our board, has to be on this journey with us. Even to the fact that we again so that everyone really understands and it’s seen it, we even show the scripts running around our customer model to our board to give them that real sense, the tangible sense, of what we’re actually talking about in terms of what we’ve been building. Because otherwise, what you find is that people see the output and the outcome, but they don’t necessarily understand all the dimension of what’s been involved in creating it. And for me, we all know as execs that there are whole bunches of topics that one might delegate to folks that work for you, but I think you know that the sense of revolution that is implicit around this data, and how data is changing, all organizations in all segments, not just fashion within retail, means that all executives have to be part of that journey. Because if you don’t understand it and you don’t understand its relevance to how the world is evolving and how you, even within your job, need to think about how that is then different. The reality is then you know, I think as I said earlier, in terms of oblivion, the problem is that you’re you stop being relevant.
Host: 04:27
Yes, exactly. And also, I wonder what your view is on this because I have so many retail conversations with people and talking about Gen Z or even now Alpha and the way that they shop your brand. And I say, well, you could you could sit any CEO down and you could explain this to them the fact that you know let’s say it’s TikTok or now from ChatGPT and moving to a GenSIC and so forth. They would understand it intellectually, but would they get it emotionally? And I say, look, if you’ve got teenage or even younger children, just observe the way that they use multiple screens and so on and so forth. Is that because you’re New Look demographic, your typical customer.
Lynda Petherick:05:17
We have a 9 to 15 range, and then we also have the 18 to 44 women’s wear market of our two main segments. We do have others, the first two of them are the main segments.
Host: 05:26
So you’re going to be not just digital native, but internet native and now AI native almost. But you also talked about getting value from data. Explain you know what that means for you and for New Look.
Lynda Petherick:05:47
I mean, no retailer has endless amounts of capex. It’d be nice if we did, but we don’t, let’s be honest. And with all those pressures externally, which just seem to be mounting. It’s critical that the way you prioritise what you’re changing within the business is based on ROI. I think the same has to apply to data, AI and agents as it does for any other investment. I think the only difference is it can be more difficult to necessarily demonstrate in kind of classic financial terms what that ROI is around a specific component. I do think you have to think about how your order and business processes can consider and evaluate what is value and what is return. I do think that is different, and that’s something that we’ve had to change as a result of that in terms of that thinking. But it still ends up being it has to be, there has to be an ROI. I mean, the problem is that if you just do it for the sake of having deployed AI, there’s a very real risk, but you require so many humans to actually check whether the agent in itself has come up with the right answer.
Host: 07:01
Yes, or hallucinating.
Lynda Petherick:07:02
Yes, exactly. You can end up spending more money than before you started. I think you have to ask yourself as well, how accurate something needs to be. I mean, there’s many things where you don’t need a definitive answer, and then that’s okay. You can try other things because you don’t need a definitive answer. But where you do need a definitive answer, then you just need to be certain that you’ve selected the right sets of products, innovations, and new technologies that allow you to have certainty that is the answer, without you having to invest loads of time and energy in terms of checking, yes, checking whether it is indeed true.
Host: 07:41
For me, it’s kind of you talk about ROI, and I think it’s answering the “so what” question. So, okay, right, AI. Well, why are we doing this? What’s the outcome? Okay, the outcome is we’re going to sell more products, yes, more efficiently at a better margin. Yes, and then we’re also going to be able to personalise that, those seem to be the messages. So, on the one hand, it’s back office demand and inventory management, and then driving efficiency, and then kind of like the front end, the customer personalisation. Or is that maybe too simplistic?
Lynda Petherick:08:21
I think it’s a bit simplistic because I think actually there’s an opportunity in just about every part of a business. I mean, you could think about the application of AI and cyber in terms of preventative and defensive measures, and candidly, you know, given what’s all over the news in terms of Mythos, if I’m pronouncing that correctly, will mean that your internal cyber approach and practices is going to have to embrace AI, if you haven’t already. Because you’re facing into AI engines that are potentially introducing a focus on your enterprise technology that didn’t exist before. So you’ve got to be able to operate machine speed. So I think it does apply to all areas without question. So if you think around the opportunity to move the profitability of a business, moving the customer, moving the top line around the customer piece and being able to very evidently target your most loyal customers, your most engaged customers in the right way in the right moment, is highly valuable. How you then customise experiences and how you look at personalisation, and evolve personalisation, is absolutely critical as part of that. I was speaking to the fact that you know our customer model is actually allowing us to think about how customer movement actually drives our five-year forecasting around what we expect PL growth to be over that period of time, and indeed PL performance, that’s quite a radically different way to how most finance functions are thinking about their evolution of a five-year plan. So that’s one dimension, yes. But I think the other dimension, as you spoke to in terms of cost serve ratios and looking at the opportunity around supply chain, again, for example, historically, you know, boats might have been in the water with product coming in. If you didn’t have a perfect understanding of where that product is, in the old world, the buying teams might have gone, oh my goodness, we are going to need more products in our stores because we’ve had a run on make it up knitwear, and they might have ordered more. Through the application, being data driven, that doesn’t happen to us anymore because actually you get the same information in supply chain and in the BMD function where you have near real-time visibility of where all stock is. So you don’t have to make that, you don’t have to order that stock.
Host: 11:10
Yes.
Lynda Petherick:11:10
I do think that whole cost serve ratio, there is lots to go at, but I think what you have to do, which I know was something I spoke to, is that you also have to be realistic about if you think about fashion and all, it doesn’t matter if it’s fashion or indeed another business, but all the components and capabilities within a business, how mature genuinely is AI and agentic AI, against each of those subprocesses and functions. Because what you probably want to be doing is thinking about that maturity and the maturity of that AI product or data product or technology product, as you always did. In some ways, I mean, let’s just be realistic with ourselves. That’s what we always do in technology, it’s no different in this. You need to be realistic about when the solutions really do make sense and how mature they are, because otherwise you’re using your precious capex to learn alongside someone who’s evolving their products. I’d rather be certain the product’s going to work. So, I mean, getting that piece right, I think it is really important.
Host: 12:26
And I’m talking to you here, and then when you’re on stage, it feels to me as if it’s also a question of bringing the whole organisation on this journey, which perhaps we don’t know where that journey is going to take us, but I don’t know. I’d imagine that also needs to include the frontline staff, the staff in in your shops, because they’re the ones who interact directly with the customer. Yes, we know the stats for online for the industry, and yourselves, but it’s the stores. I’d imagine that frontline staff have a big part to play in all of this.
Lynda Petherick:13:13
I think they do. Because ultimately, as I talk to the Omni, our Omni customers are most loyal, most engaged customers, and therefore, obviously, any opportunity for a colleague, whether in-store or otherwise, to engage with the customer is an opportunity to find that moment to provide that moment of delight.
Host: 13:33
Yes.
Lynda Petherick:13:34
So I do think having store colleagues and what you can enable them with in-store is an absolutely critical dimension to that. And I think alongside, and I think that’s it’s important when we think about the need to have modernised your core in terms of the core sets of technologies. We invested in creating new Point of Sale systems, which we’ve rolled out, leveraging an active technology, and that went in less than 12 months ago. We also invested in moving legacy hybrid web platform and moved that to Commerce Cloud with SAP. So there’s a whole set of other technologies which interplay within a store, and wider business context, that you also have to have either modernised or understand how they play a role as part of whether it’s enabling store colleagues with a customer in front of them or indeed making decisions around what you are doing through your e-commerce channel. I mean, it’s the entire enterprise architecture of which the data architecture is a part, yes, that is critical to it all.
Host: 14:52
Yes, yes. Now we’ve got a few minutes left. I just wanted to ask you something that you covered on stage, and that was the advanced customer model.
Lynda Petherick:15:04
Yes.
Host: 15:04
So tell us about that.
Lynda Petherick:15:05
I love a customer model. I think I referred to it, I guess, already in terms of understanding how you truly pull together all the different pieces of customer-related insight, product-related insights, to create a true understanding about what you think is going to move in terms of your top line and also around profitability. I mean, we’ve invested a huge amount of time and energy in terms of building that for ourselves as we feel it’s differentiating. I think the other aspect is every business is unique. So it didn’t seem practical or logical to go and ask someone externally to create that because ultimately these are our customers, and the way that we engage them in the way that we specifically want to be able to provide that personalisation and that experience. All of those things are within the customer model. It’s slightly mind-blowing, but when you break it down, we actually have segmented the customer base into 220,000 different segments. That’s what’s realistic within the model. So each segment only has about 50 customers in it. So just to give you an idea of the science that sits behind that, it’s that level. But if you roll it up to a simpler way of thinking about it, it’s basically around circa 14 cohorts, and whether those people are omni customers, whether they’re loyalty customers, whether they’re store-only customers, or if they’re indeed customers that we don’t actually know how they engage with us. It’s the mechanics of that and all the varying combination and deviation of those dimensions, which then was expanded out and extrapolated to create those essentially mind-boggling number segments. And then what you do is think about, and you apply into that, what your store profiles look like, understanding the halo effect associated with stores and the impact on customers, understanding which products those customers buy. What our loyalty schemes call loyalty customers, who are omni-customers, buy knitwear, for example, and then you extrapolate out because then what that gives you is an understanding back to your trading, back to your product, back to what you’re buying, being able to make better decisions.
Host: 17:40
Yes, it’s fascinating. So I did a Fireside chat earlier today with Mark Blundell of Harrods, and he was talking about, in a very similar way, in terms of what they do with their customers, probably quite different in many ways to a typical New Look customer, but it doesn’t matter. He was describing the sort of detail in terms of what they feel they need to know about their customer. Going back to AI and what it can do and the potential, I just think it’s so exciting because we talked about personalisation in retail. And I always use the example I’m a proud card carrying Amazon Prime member, I don’t mind admitting it. When they confirm an order, they’ve got all my information, my address, credit card data, everything. And yet they still say “Hi, A”. And it’s like, no, my name’s Andrew. Why can’t you just, now I know that doesn’t really matter, but it does kind of matter though. But I think that’s exciting because the journey that we’re on seems to be taking us all in the right direction.
Lynda Petherick:18:54
I think so. I mean, you have to always get a balance because there’ll be some people who actually don’t like necessarily companies knowing that much about them, and you need to know that too. I mean, you and I may be in the group that actually wants to be engaged with in the most perfect way possible, but there’s a group of people that don’t really want to be engaged with at all. So, understanding how to make it right at this single customer level; it’s the holy grail.
Host: 19:24
Yes, yes. Lynda, that’s fantastic. I think that’s a great place to end on. Thank you so much for your time, Lynda Petherick.
Lynda Petherick:19:32
You’re very welcome, thank you very much.