Looking at Retail Challenges from a Data Perspective - with Nick Masca of Marks and Spencer

Welcome, everyone, to the AI and Business Podcast. I'm Matthew DeMello, senior editor here at Emerge Technology Research. Today's guest is Nick Masca, head of data science for growth and personalization at Marks & Spencer. Marks & Spencer is a prominent British multinational retailer headquartered in London, known for offering a wide range of clothing, beauty items, home goods, and food products. Nick joins us on the program to surmise his views on the data-driven challenges currently facing the retail and e-commerce sectors, With a focus on change management rather than traditional digital transformation, Nick outlines the key obstacles retail leaders encounter when leveraging data tools to optimize processes like price setting, supply chain efficiency, and customer experience. He shares insights on the friction that arises when introducing automation, particularly in areas like content development, and how data teams can work closely with stakeholders to ensure seamless implementation. But first, are you driving AI transformation at your organization? Or maybe you're guiding critical decisions on AI investments, strategy, or deployment? If so, the AI in Business podcast wants to hear from you. Each year, Emerge AI Research features hundreds of executive thought leaders, everyone from the CIO of Goldman Sachs to the head of AI at Raytheon and AI pioneers like Yoshua Bengio. With nearly a million annual listeners, AI in Business is the go-to destination for enterprise leaders navigating real-world AI adoption. You don't need to be an engineer or a technical expert to be on the show. If you're involved in AI implementation, decision-making, or strategy within your company, this is your opportunity to share your insights with a global audience of your peers. If you believe you can help other leaders move the needle on AI ROI, visit Emerge.com and fill out our Thought Leader submission form. That's Emerge.com and click on Be an Expert. You can also click the link in the description of today's show on your preferred podcast platform. That's Emerge.com slash ExpertOne. Again, that's Emerge.com slash ExpertOne. Without further ado, here's our conversation. Nick, thanks so much for being with us on the program this week. Great to be here. Yeah, thanks for the invitation to join you. Absolutely. It's not every day we get to talk to somebody who's really on the data side of things in retail. We did a really great podcast series a little earlier this year with leaders from Etsy and Instacart and in the gap talking about, you know, different data related problems from like the operation side, the loss prevention side. But we all know and long term listeners of the show also know that looking at things from the from a data perspective really, really changes things. So just from your vantage point, what do you see as the biggest challenges currently facing retail and e-commerce sectors right now from a data perspective? So I think the retail industry is a fascinating place for data people. There are huge opportunities to improve operations within retail organizations and to directly impact and improve that customer experience. So my own role, I lead a data science function really focused on personalization of our experiences for customers. There's a huge opportunity there, whether that's personalizing, say, marketing communications or personalized experiences on our website or on our app, for example. That can drive automation as well, talking about those operational aspects within the business as well. More broadly, we've also had teams using data to improve things like the way we run supply chain in our organization and how we run other business critical operations like markdown events, things like that. We've got really popular with our customers loyalty scheme. We relaunched a few years ago and again, have tailor-made offer campaigns for our customers, which again, are driven by data and machine learning. So yeah, quite an exciting time to work in data in the retail industry. Indeed. And just kind of pulling apart a couple of the notes that you mentioned there, there's the loyalty program, which really drives a lot from the data perspective. There's also price optimization. There also elements of the supply chain Where do we find and I know you were speaking a little bit in the last answer just about the loyalty program especially and I know we were talking even in the lead up to these conversations about how a lot of retail organizations go about what typically is for other organizations, what gets called like a digital transformation. They do a little bit more change management. And we'll get into what that means in a little bit. But just in terms of like, across the business where data science is being pulled in to help solve the problems outside of the loyalty program. Tell us a little bit maybe about what's involved in change management and how that's touching different areas of the business that comprise where data science is needed the most. Sure. So, yeah, firstly, data is a central component of our business transformation at M&S. So, firstly, yeah, to improve business operations and customer experiences, but secondly, more broadly, to change that culture, to use data to inform, data and other evidence to inform decision-making. We've been championing a culture where we experiment, where we can create freedom for teams to try things, measure the results, and then make evidence-informed decisions, which, you know, for traditional organizations, you know, that's quite a big change there. In terms of how we organize ourselves, So, yeah, we have, you know, customer focused teams whose purpose it is to improve that customer experience, drive customer value as well as business value off the back of that. And we have what we call enterprise focused teams working on business facing processes, you know, like the supply chain, for example, improving efficiencies in the supply chain, working on things like markdown and pricing. And just in terms of having a seat at those tables, especially for an organization where they know things are becoming more data-focused, but they're going through it in the change management model rather than the digital transformation model, which to me sounds like you know that data is going to be at the heart of this, but at the end of the day, your business is brick and mortar. It is fundamentally just not online. And I'm wondering, how does that change the conversations with stakeholders, especially in terms of emotional management, working with folks to change their processes based on data, but knowing fundamentally, you know, we're not slapping them with like, you know, a co-pilot tomorrow. Their job is going to change. Their workflow is going to change based on the data they're seeing. But, you know, it doesn't mean that there's more technology necessarily involved in their new process. So, yeah, I'd say like Eminence is a traditional organization with around 15,000 stores globally. But we've had a big focus over recent years, especially on driving like omni-channel growth, driving growth of our website. I think we launched our website in 2007, launched our app in 2010. So we've got some years of experience doing that now. The transformation probably until recent years, I think COVID was a bit of an accelerator of our digital transformation to drive more shopping online. Sure, sure. So just in terms of the emotional management and the stakeholders in the conversations that you're handing, I'm just trying to draw a starker contrast to maybe a change management process rather than typical digital transformations as they occur in other sectors. You know, when you do a digital transformation in logistics or FinServe, you know, the underlying assumption is that data and the data tools are really going to drive, you know, the core values of the business. To me, it seems, and correct me if I'm wrong here, change management deals a little bit more in data complementing the process. Maybe not necessarily be, you want processes to be data informed, but the end goal isn't necessarily making it so that everybody's attached to a new system. You know, I guess it's a little bit more of a minimalist view of how far you need to take the technology. A little bit more driven by what are the SMEs saying? where do they need the help? Sure. So, yeah, I guess for an established organization with many existing processes already in place, like I think we are in a digital transformation M&S, but for us to replace or perhaps just adapt an existing process, there is change management needed in order to ensure that performance is not slipping, that you can, say, take steps out of a highly manual process of lots of people contributing, perhaps using data or algorithms to automate parts of that or to optimize parts of that. That change, to do that smoothly, requires careful management. You know there are people issues involved You need buy from the people who already kind of manage those processes at the moment I think when you driving that digital transformation digitizing operations that kind of hand-holding of the business is really crucial to gain buy-in and get people on your side contributing rather than playing against you. And I think where people can be disrupted by technology. Yeah. If you don't invest in that appropriate change management, you can really find progress very slow, very hard. And initiatives can ultimately fail without gaining that buy-in, investing that effort. Yeah. I'm really curious as to the preparation in the process, especially for dealing in that emotional management. What do you want to bring to the table from a data data governance perspective in terms of winning that buy-in. And it seems more that you want the buy-in or it's more crucial to get the buy-in from the subject matter expert necessarily than the executive. Or tell me if I'm wrong there. Yeah, I think, yeah, you use different tactics for those two types of people. For the people on the ground, you've got to highlight, firstly, the reason why you're doing things and perhaps the advantages to what you're doing, hearing about their pain points as well, listening to their experiences, taking into account their pain points and finding common ground is really important. There are other processes that can help in terms of aligning interests between perhaps a technology team or a data team and a traditional team. We've been using OKRs, Objectives and Key Results, for a few years at M&S. OKRs are a goal-setting framework, but a neat thing about OKRs is these are openly available objectives for the business. So it helps you can share goals between different areas. If two different teams have aligned goals and can see that, that can make progress much smoother as well. And just in terms of SMEs especially, I know we were talking in our lead up to this recording about how adapting machine learning answers can change who owns the process. And that adds more friction to the process in terms of implementing automation. Tell us a little bit about those frictions and maybe some strategies that you've seen really succeed in at least addressing how that gets straightened out. Sure, yeah. It's a good one. There can be friction where you do have these changes in accountabilities and roles and responsibilities when it's just new technology. The classic one we've seen in a couple of big initiatives at M&S is traditionally we've had commercial teams or marketing teams, for example, who are responsible for creating business rules And, you know, typically the technology team, whether that's in-house or outsourced, would just be responsible for implementing the rules that they're given. So, you know, acting as a service, you know, that operating model has to change if you're going to start to be using technology to optimize those processes or to, you know, automate some of that, you know, take some of that manual effort out of that system. Yeah, it can be difficult for people to go through that change, to lose the abilities, to give their rules for a team to implement versus relying on a different team to propose an approach for them. I think if you can measure your results, you know, I mentioned before, like an experimentation culture, you know, having measurement as a key part of your work as well is crucial. Like if you can show the performance is not slipping or even you're improving performance through introducing, you know, new automation or say a new machine learning system, like that can be quite convincing. like we you know we all have targets to hit and if we if you're helping people meet their meet their targets you know that's typically a good thing i think how you how you work with these people as well like sometimes people's smes roles can may have to adapt a bit when you've got automation like rather than just you know handing you rules to implement they can end up being a bit more having an analytical role where actually then they're seeing like areas where algorithms are not performing well. So surfacing areas for opportunity that you can improve the automation rather than having to handcraft all the rules in the first place. So it's kind of creating new opportunities and slightly different responsibilities for those people. Absolutely. And it almost sounds like to really win that subject matter expert buy-in, the SME buy it about making them feel part of the data science team or that data science is inaccessible to them and they should be with you as something of an equal in the process of looking at the changes, looking at the results, and do those results match our hypothesis for what's going on? Does this give us a stronger indication of what are the core business processes and the core business values that we need to have a better grip on in order to streamline processes. I know even in terms of manual processes, content development is really big right now. That still has a lot of manual processes about it. All at the same time, I know that especially in this moment of generative AI, everybody's kind of holding their breath knowing that this is going to change. We've talked about it elsewhere in other episodes on the show about the gap between development and design. I'm just wondering if we can kind of use that maybe as a not a full blown use case, but like an impending case study for right now. How do you imagine that that might suss out into the future in terms of those manual processes becoming more automated in that gap closing with new generative AI or even the older, more traditional capabilities of artificial intelligence? so yeah if you speak to the people who are responsible for things like writing product descriptions or tagging attributes on on products that you can use say in filters on a website like usually that's quite a laborious job it's not that exciting for people to do it's repetitive it can be error prone people are often not not unhappy to to give up some of that repetitive work And people usually have broader roles than just doing that work. So a lot of the people we speak with whose roles may adapt a bit, can be brought in to using technology to get rid of some of that manual work. But yeah, in time, who knows? If I had a crystal ball and could predict the extent of what Gen.AI will be able to do in coming years. Yeah, we're still in a little bit of the fog of war with generative AI. I think the comparison I like to make, listeners will feel a little bit like I'm a broken record, but I think it's apt. We seem to be in this moment with generative AI that it's a lot like the dawn of the internet in 1998, where just about a lot more people had it than they did in 1995. and you can get your internet at your local library if you don't have it in your house yet. And everybody's astounded that it takes, that they can get a website in about 20 seconds. And by the time you get to like 2002, four years later, that's unacceptable. That's so long. 20 seconds for a website or 40 seconds for an image to load, that's absurd. And by 2002, maybe everybody in your neighborhood has internet. It's probably pretty high speed, I'll be attached to an Ethernet cable. And yeah, I think that is, especially while we're in this moment of hallucinations and still trying to really nail down that misinformation problem with a lot of LLMs and generative AI, that yeah, there's a little bit of the same fog of war of like, we're impressed by this technology now, but maybe in a couple of years, will we think this is absurd and moot? But Nick, really appreciate you being on the program this week and spelling out how these problems look a little bit more from a change management perspective than the traditional digital transformation that we see across industries. Pleasure. Thanks for having me. For more in retail, tune in to our June 1st, 2023 episode of the podcast titled Retail Fraud and Loss Prevention in Data, Brick and Mortar and Beyond with Chris Nelson, Senior Vice President of Asset Protection at Gap Incorporated. Chris joined us on the show to share his experience in loss prevention, going from the military to on-site security and onto the world of data. He describes the challenges in each realm, where they're interrelated, and what the difference looks like in data signals. If you enjoyed or benefited from any of the insights of today's episode, consider leaving us a review on Apple Podcasts. Let us know what you learned, found helpful, or liked most about the show. Also, don't forget to follow us on X, formerly known as Twitter at Emerge, that's spelled E-M-E-R-J, as well as our LinkedIn page. On behalf of Daniel Fagella, our CEO and head of research, as well as the rest of the team here at Emerge Technology Research, thanks so much for joining us today, and we'll catch you next time on the AI in Business podcast. Outro Music