AI in the Retail Industry: Re-defining Operations & Customer Experience

Paul Morrison

Hello and welcome to Retail and Consumer Pulse, brought to you by WNS. In this series, we explore the world of retail and consumer goods, spotlighting the latest strategies, trends and innovations. And to help us, each episode features an industry expert or leader sharing views on what it really takes to succeed in this hyper-competitive market.

My name is Paul Morrison, and I lead the WNS Retail and Consumer Practice in Europe. For today's session, Checking Out Retail in AI, I am delighted to be joined by a special guest, Richard Lim. Richard is CEO of Retail Economics, based in London.

Hi, Richard! Thanks for joining us today.

Richard Lim

Hi, Paul. Thanks for inviting me along.

Delighted to be here.

Paul Morrison

You are very welcome. The team here is a big fan of your commentary on global and UK retail markets. And I particularly like your crunching of income statements from Marks and Spencer, Sainsbury’s and others into very powerful visualizations, almost works of art. So, we are big fans of your analysis. And perhaps you could quickly kick us off and tell us about your journey in retail.

Richard Lim

Sure. Retail Economics is an independent economics consultancy. We've got a laser-like focus on consumer and retail.

As you said, I'm the CEO and founder of the company. So, we started the company about seven or eight years ago now. But essentially, we put economic analysis, behavioral economics, forecasting, analytics at the heart of the insight that we provide the industry.

I have been in the industry, though, for around about kind of over 15 years now, keeping a close eye on what's happening in the retail sector, let's put it like that, and consumer behavior.

Paul Morrison

Brilliant!

So, an ideal perspective for our conversation about retail AI, and we're going to dive into the details in a moment. Perhaps you could just give us your big-picture take on what's going on with retail AI.

Richard Lim

Yeah, I mean, we're going through a period of such enormous change and such rapid change as well. And if I just expand it out for a second, I think what we're seeing with Gen AI is going to have a profound impact on all corners of the economy and all businesses.

So retail, I think, is one of those industries that will see some of the biggest changes and some of the biggest challenges. And from my perspective, I think it's going to touch every single part of the retail value chain, right from kind of manufacturer and supply chains, right through to chatbots and customer service and everything in between.

Paul Morrison

I think that's absolutely right.

I guess my builds from a top level are firstly that this is absolutely happening. It's also building on a long period, a long history of movement and developments in AI. Looking way back, the very first business computer in 1951 for J. Lyons & Co. was a retailer and was focused on what it called time intelligent processing. And then you fast-forward through some of Walmart's innovations, Amazon's, Ocado and so on. There has been a multi-decade journey to get where we are.

But there is a massive explosion, as you say, at the moment. We hear of annual growth rates of 30 to 50 percent in terms of the uptake of AI in retail. And by the end of 2027, some estimates will be that 80 to 90 percent of all retailers will be extensively using AI across their operations.

So, absolutely backing up that picture there. I completely agree with that point about proliferation. And I think one key point at its top level is why is this happening now. From my perspective, because the technologies we'll talk about have reached maturity. It basically means AI is available at a lower price point.

That's really what's happening. So, it's not new, but it's more pervasive. It's more capable and available at a lower price point, which means that all sorts of organizations, perhaps much smaller than in the past, are able to tap into this technology.

Very exciting!

Richard Lim

Yeah, just building on that. I completely agree with that. And I know this conversation is focused on AI, but actually, there's a wider technology piece where we're seeing almost like a democratization of digital and retailers are able to leverage the benefits of past investments over decades of digital infrastructure.

AI is definitely the center point of that. And as you say, the barriers to entry to actually leverage the benefits have come down so significantly that it's being utilized by companies of all sizes, in all sectors, across all different channels. And so, this technology is there for the taking.

Paul Morrison

Exactly! Democratization, I think, is the word indeed. Well, let's put some more parameters around what are we talking about in this massive topic, and I'll just kick out a few. I wonder if you agree or would add or amend.

Firstly, what are the technologies that are shaping retail AI?

We could spend a whole podcast on this, and we're not going to. But from my point of view, I would summarize it as there is a wide and evolving set of technologies that make up AI and retail AI, such as computer vision, NLP and other language technologies. Obviously, Gen AI being the foremost of those now. And then, some of the other AI technologies of the last ten years or so around deep learning, neural networks and machine learning-driven analytics.

These form a group of technologies that have really unlocked the opportunity in the last 10 years, and particularly in the last couple. So, there's that wide mix. Sometimes, we see them deployed on a relatively standalone basis.

So, there's a computer vision solution for problem X, let's say. But very often, we see them in combination, and very often, combined with other changes around workflow or devices or digital signage or connectivity, whatever it might be. So, very often, it's baked into a wider solution.

I think a third parameter is, you know, taken together, all of these AI investments are trying to carry out activities that were formerly requiring high levels of human intelligence and judgment - prediction, holding a conversation, writing content, creating images, finding patterns, making recommendations.

Those judgment-based activities are the objective of these technologies. And I think the final sort of parameter or dimension I put on this is that we are absolutely talking here about e-commerce, but also bricks and mortar. And I think from my perspective, that makes up the big universe of retail AI, which I know from your perspective is something that you see or you see it slightly differently.

Richard Lim

I think you've kind of provided a very comprehensive helicopter view of what's happening. So, yes, I think, all of the things that you've mentioned have resonated with me and reminded me of various different conversations I've had with retailers and other kind of service providers. You can probably then start peeling back the layers and start thinking about specific use cases or how they're using chatbots.

And so, some of that service is built on internal data to drive conversational insights. So that's something that I think has been really interesting as a development within this space. And then also what I've seen is quite interesting is around retrieval-augmented generation.

So, RAG! That specific type of technology is being used to drive insights out of unstructured data. I think that's kind of a really interesting area.

Perhaps we might kind of cross that a bit later as well.

Paul Morrison

Absolutely. That sounds great.

Well, why don't we move the spotlight to this - the key question about the use cases and where retailer AI is making a difference today and in the future? And perhaps, Richard, you spoke when we talked last about this difference between AI that is being driven by the retailer versus AI that is being driven by the consumer. There are two different overlapping ways of looking at the space, but perhaps, could you give us a flavor of that retailer-driven AI? What stands out for you?

Richard Lim

So, I think that's right for me from a really simplistic view. If I think about how it posed with the question of how is AI, Gen AI affecting the retail sector, immediately my thoughts kind of fall into two camps. How are consumers embracing or potentially will embrace the future Gen AI that's going to affect customer journey?

And then there's, of course, how retailers are embracing Gen AI to drive operational efficiencies, deeper insights about their customers or perhaps using it to generate more intelligence around commercial operations and optimizing it in that kind of way.

If we start with the retailer side of the equation, then I think, there are probably four main use cases that I see retailers embracing. And I think that the journey the retailers are going on at the moment is looking at the technology and actually just mapping out all of the different business, the kind of the use cases across their business and then going through that process of thinking about what demonstrates the best return on investment, what's the low hanging fruit, where they can really focus on having the biggest kind of impact.

So, the four areas for me, I would put out around personalization – using customer data to provide more personalized marketing messages, thinking about the next best action, but in a really sophisticated way. Price optimization would be my second area, optimizing prices for better sell-through, but then also trying to use more dynamic pricing for the sell-through end of season and things like that. Better demand forecasting would be my third area that I see a lot of activity.

And then the fourth area is around supply chain optimization. But I think those are the kind of the four areas that I'm having conversations with our clients and other retailers in terms of where they're embracing Gen AI.

Paul Morrison

Absolutely.

And I think that definitely resonates with our client work and client conversations. Just to pick out one – personalization. My colleagues recently worked with a retailer in South Africa around groceries and generating a personalized mail campaign across its customer base using Gen AI-created personalized messages. Instead of taking around 15 days, which is what the usual lead time is when using a creative agency or sort of human solution, it took around 15 minutes and the effectiveness of that campaign was four times higher.

So, four times higher click-through and resulting in higher sales as a result of that. Gen AI is real today and delivers clear, very clear ROI. I think if you look at those other three examples, you'd see a similar pattern.

So, I think that's great. And from your perspective, Richard, these are live. These are current investments that your contacts and network are making.

Richard Lim

Yes, absolutely.

I mean, what I would say is that from an industry perspective, I do think that we're seeing a significant gulf between those retailers that are embracing this type of technology and those that aren't and for lots of different reasons as well.

While some larger retailers are embracing this, looking at it as a huge opportunity, a way to create competitive advantage, a way to really speed up their digital transformation plans, you've got some retailers that have the right culture of innovation, of change, of agility to want to try to embrace that.

But then you have other large retailers that are suffering from a bit of paralysis around what it means for the company, the best way to incorporate this type of technology, but also with a bit of fear around what it might mean and concerns around privacy, data privacy and things like that. So, there's a camp that is embracing it and definitely using it.

But then also, I think there is definitely some kind of paralysis in other companies just because of the fear factor and the unknown.

Paul Morrison

I think that's a good point.

If I think back on some of my clients or contacts in this space over the years with other waves of technology change, many and most of them have taken time or years to roll out.

And Gen AI has been with us for 18 months to two years. So to some degree, these technologies are fairly new. But I recognize that the range of leaders and market size is definitely true.

Before we move on to the consumer-led side of the retail AI equation, I just pull together a couple of other examples, two individual companies just to illustrate. I think the key point you made Richard earlier on is that AI is not one technology and it's not one use case. It is multi-tech and multi-use case in any company.

So, the two examples that may be different, end with different parts of the market. Take Amazon. So obviously, world world-leading e-commerce player arguably uses AI from the very beginning.

And if you look at its business, AI is key to things as diverse as the recommendation engine the Alexa voice agent. It's been developing. It's just using walk-out technology in its ‘Go’ stores.

It uses AI in venture management, in its fulfillment centers, for the robotics there, in route optimization, in detecting fraud and security issues and in optimizing energy. And then more laterally, it's been updating how reviews are created using product reviews, and product summaries are generated. It's been using machine learning for years, but it has added Gen AI to that.

Then, there are some new recommendation engines that it's using with AI. So, an organization is permeated by AI use cases, and that's to be expected from an organization like Amazon. If we take a different type of organization like UK fashion retailer, AllSaints, there’s lots of information.

It's about 500-million-pound revenue and a store-based or store-led operating model. But that too has embraced AI in its algorithms for its personalized shopping e-commerce platform. Again, it uses it in venture management.

It has the functionality for customers to upload images so they can compare and feed in images to the website and get recommendations based off that. Customer service, pricing, you mentioned that, marketing campaigns and security as well. So, a very different type of organization also richly using AI at the moment. I thought that was interesting.

Richard Lim

It's really interesting. And yeah, there were just so many different examples of retailers embracing it and doing it in some really clever ways.

While you were going through that list there, you reminded me of other companies providing services in the retail sector. And you reminded me of TikTok as one example. So TikTok! Obviously, social commerce is having a really disruptive impact on the industry and TikTok is doing some really interesting things and leveraging Gen AI in interesting ways.

One example would be how influencers are using text to video to be able to create AI-generated video content to be able to showcase their products. That's one example. And then, some of the latest things that they're doing is around being able to use video technology.

You can record yourself on a kind of green screen. They will turn that into an avatar of yourself, and you can then promote your products, do product demonstrations, do whatever you want. But it's text to videos.

So, you can do it in different languages. You can do it. So, you can create video content to promote a product or your own product across, let's just say 15 different countries in 15 different languages.

You could probably do that in about five minutes. So, it's just the level here of how it's going to change the tactics and some of the operating model for different retailers, even small businesses.

Paul Morrison

That's really good and really disruptive. And that's just part of it. I guess you were going to unveil or you were going to tell us about that sort of wider consumer-led AI sector in general. Your view is that this is going to be an area of massive growth. Can you explain what you mean by consumer-led or consumer-driven AI?

Richard Lim

Yeah, sure. So, I mean, a lot of the focus on Gen AI at the moment is on retailers and how they're implementing all of this different change.

But I think less attention really is placed on how Gen AI is likely to affect the customer journey over the coming year, two years, three years. And actually, my view is that it's going to have a significant impact on the way that consumers are discovering and researching brands and finding new brands. We're already seeing some of this, some interesting things through Google and Bing and through the generative experience that Google is offering.

Probably if I'd said this three months ago or even six months ago, people hadn't experienced it then. But now when you're searching on Google, you get this kind of at the top of the page generative experience results. And so, you might ask questions around, you know, what are the best trainers or what are the best boots to use on an outdoor holiday in the winter? And previously, obviously, you'd get a set of results, your traditional kind of blue underlying clicks.

But now Google is serving up a generative experience, which turns into a more conversational search that's more intent-based. From a consumer perspective, it's going to change the way that we are discovering products and the way that we are searching for products. I think that's really interesting.

Anyone who's used ChatGPT and used the voice on ChatGPT to have a conversation with ChatGPT realizes that we've come so far from the days when Alexa was launched and even what Alexa is like today. So, I think that we've seen this kind of huge acceleration in the capabilities of voice commerce. And that might be another disruptive area where you can have a conversation with a chatbot and they understand context, they understand meaning, they understand sentiment.

It's going to serve up some suggestions on products that are actually kind of relevant to what we're looking for. I think other interesting areas, just as an example, but IKEA has built something in ChatGPT where you can go on and you can have a conversation in GPT that's specific to IKEA and the products that IKEA sells. So, that conversation might be around, you know, I'm going to redo my kitchen. These are the kind of the dimensions of my kitchen. I'm looking for this type of style, this type of feel. And then it serves you up content that's based on the conversation.

So yeah, I think there are lots of really interesting areas. But I don't think we're too far from being able to walk down, be in the middle of central London, double tap on your AirPods and start having a conversation with AI that's integrated into hardware, where we can say things like, “I'm going to lunch in the next two hours. I need a table for three people with a top table rating of no less than four stars, 50 pounds per head. What are the available options?” You know, and we start having a conversation in that way. I honestly don't think we're too far from that.

Paul Morrison

I definitely agree. We are more or less there. And I think, in that list of ideas you went through there, there's so much to unpack. We have so little time.

So, we will have to come back and go into more detail. I think a few of the points that stood out to me, I think that the point around how search is done and discovery is done is really important. I've seen a lot of news and developments from different types of retailers.

I think Walmart and Instacart are two that jump out that have been investing a lot in generative AI-enabled search. You search for an idea or a need as opposed to a specific item – “I'm having a barbecue at the weekend, help me shop for that as opposed to the specific items.”

I think that's going to become very, very normal, very, very quickly. I think in this space, it seems that in many of these sorts of technologies, AI approaches are giving the consumer the sort of driving seat. It's interesting in many ways. It's very often fashion, retail, fashion and apparel that are leading the fray in terms of things like virtual try on at different retailers or cosmetics, players like Sephora and sites like Zalando in terms of generative design, creating new designs on the hoof, as it were. So, I think that's interesting.

We're going to have to move on, I'm afraid, to our sort of next part of the discussion and really sort of taking all of this opportunity that we've only really just scratched the surface on. The big question is, well, you know, how do you make moves to tap into that and make your investments in this space successful? I wonder if you could kick us off with two or three top tips as organizations move further into AI.

Richard Lim

I think the big one that is hard to measure and hard to quantify, but ultimately is kind of like the DNA that runs through the successful kind of digital transformation plans for me, is the culture and having the right leadership, having a clear strategic direction and goal, having the culture of innovation, change, agility, fail fast kind of mentality. I think that culture at the moment, just given the pace of change, is going to be a really significant competitive advantage for companies. So, some of the companies that are stuck in old practices and old kind of mentality will really struggle to keep up.

So, I think - culture - although it's difficult to quantify and really put your finger on, I think that culture is going to be a really important part of success over the next few years, or it has been, but even more important. I think partnerships are going to be absolutely critical. Most retailers do not have the in-house skills to build their own applications and really integrate the benefits of what Gen AI can offer.

And so, finding the right partners is going to be absolutely critical in order to kind of keep up with the pace of change. And then, the third thing for me is just data. Of course, data is the oil that makes this Gen AI engine run.

Lots of businesses, large and small, suffer with getting data in the right format, in the right place, in the right kind of central location to be able to really drive those insights. So, for a lot of companies, that is a 101 of getting the data into the right place, into the right format and moving away from working in different silos and not having that kind of single view. But I'd say those are the three big things for me.

Paul Morrison

I think they're three really great points. And another item you mentioned earlier on around the need to identify, filter, and prioritize the right opportunities and build a strategy from that is, I think, I'd add as a layer or a follow-on to those three. I'd also add a question that needs to be solved around understanding the build versus body equation. I think, maybe about 10 years ago, the emphasis was on this functionality didn't exist so it had to be built. But already there's a massive ecosystem of retail AI solutions as well as wider general solutions, so the buy options are getting stronger all the time. I think that needs to be understood.

This doesn't mean that, as you're saying there needs to be a culture of innovation, and the skills to innovate behind that around computer science and data science are absolutely essential. But it is also the case that the wider skills around process design, strategy, business analysis, change, communication, all of those wider transformational skills are equally powerful.

I'd also just flag this point, and it's not to deflate the AI bubble but it's just to make the comment that in many ways the AI of the past is just software and has become normal. And so many things we're talking about now will become embedded in different platforms, be that ERPs or core systems and it will just be there. Then AI will be the new functionality that hasn't previously been automated so we might just keep an eye on the fact that AI is constantly moving as a target. The final thought for me is around the importance of human intelligence along with AI.

We don't have time to dig into this but I think there is, in many discussions or strategies around AI, a need to think about the darker side or the potential negative consequences of some of the massive change that we're talking about.

Consumers don't care about AI in itself but they are interested in service and experience and impacts, and there are employment questions or authenticity questions or control questions about AI. I got a great quote from Erik Brynjolfsson from MIT who said, “AI is the ultimate personal shopper but it's also the ultimate manipulator.” So, that may be a sinister take on it but there needs to be a sort of care around all this change. I don't know if you have a view on that dimension, Richard.

Richard Lim

I think it's a really valid point. And yeah, with all of these kinds of technical solutions and the integration, there's a huge element of automation in terms of - Is this hands-free automation? Is it the kind of automation with a kind of human interaction as well?

There's still lots of issues around things like hallucinations and making sure the accuracy of AI is a level that is acceptable. So, there are certainly lots of risks around implementing the technology and it's putting kind of guardrails around that to ensure that you're not making any kind of significant mistakes. I think, you know, as all of this is going on as well, you've got consumers that are becoming or have become increasingly unforgiving and also intolerant of bad experiences. So, there's a balance that you know the industry needs to walk in terms of automation and AI versus human and the personalized side of things. So, there's certainly lots of things and there's a balance to strike to get that right.

Paul Morrison

Absolutely lots to dig into but we are out of time. That's how you burn up 30 minutes in a blink. So we have to draw the curtain down there. Many thanks, Richard for a really great discussion. Look forward to crossing paths soon in your next retail roundup.

Richard Lim

Brilliant. Thanks, Paul so much! I really enjoyed that. As you say, I do lots of content on LinkedIn so if anyone's out there please do follow me on LinkedIn. I share lots of content on there but really enjoyed it today. Thanks so much for inviting me along.

Brilliant and thanks to you our listeners for joining Retail and Consumer Pulse today. If you've enjoyed the show please do like and follow us and stay tuned for our next episode. So, thanks and goodbye. Thank you.