The Future of Shopping with Miya Knights

Podcast cover art for In Clear Focus episode The Future of Shopping

IN CLEAR FOCUS: Our guest, Miya Knights, author of "The Future of Shopping: Using Digital, Data and AI to Win with Customers," explores the evolving retail landscape. As AI assistants begin making shortlists, Miya explains why product data must be clean and machine-readable for brands to remain selectable. Learn how to optimize retail media networks, when to intentionally add friction to the customer journey, and why consumers trade personal data for recognition rather than surveillance.

Episode Transcript

Adrian Tennant: Coming up in this episode of IN CLEAR FOCUS:

Miya Knights: We were very obsessed about pretty PDPs and 360-degree spin and true color resolution when we first embraced e-com. But homepage merchandising and banners can lose influence, and the battleground becomes more about pricing, transparency, promotional effectiveness, and most importantly, machine-readable product attributes.

Adrian Tennant: You're listening to IN CLEAR FOCUS, fresh perspectives on marketing and advertising, produced weekly by Bigeye, a strategy-led, full-service creative agency growing brands for clients globally. Hello, I'm your host, Adrian Tennant, Bigeye's Chief Strategy Officer. Thank you for joining us. The reasons people shop haven't changed much in decades. How they shop seems to change every holiday season. This year, many people will ask an artificial intelligence assistant for gift ideas before they open a retailer's app. In stores, digital screens and shelf-edge cameras can now connect what shoppers see with what they buy. For brand marketers, that raises a practical question. When more of the choosing happens through algorithms and retail media, how does a brand make the shortlist? My guest today has spent more than 25 years watching retailers answer that question with technology. Miya Knights is a retail technology analyst, editor, author, and consultant. She owns and publishes Retail Technology magazine and is a member of the KPMG Retail Think Tank. Earlier in her career, she was a senior research analyst at IDC Retail Insights and global technology research director at Planet Retail. Miya judges industry awards, including The Grocer Gold Awards, and in 2021, the Black British Business Awards recognized her as their Arts & Media Senior Leader. Miya's new book, published by Kogan Page, is "The Future of Shopping: Using Digital, Data and AI to Win with Customers." The Bigeye Book Club selection for November, it's her third book for Kogan Page, following "Amazon," co-written with Natalie Berg, and "Omnichannel Retail," co-written with Tim Mason. To discuss what shopping missions and AI agents mean for the brands competing to be chosen, I'm delighted that Miya is joining us today from London in the UK. Miya, welcome to IN CLEAR FOCUS.

Miya Knights: Thank you, Adrian. It's lovely to be here. Thank you so much for having me.

Adrian Tennant: Well, first of all, congratulations on the publication of "The Future of Shopping." Now, this is your third book for Kogan Page, but it's your first as a solo author. So, Miya, what was different about writing this one on your own?

Miya Knights: Oh, there are a number of things. It's certainly a lot more daunting, a bit more challenging, and twice the amount of work. I think co-authoring gives you a built-in sounding board. I wrote "Amazon," the first edition, with Natalie Berg in 2019. Then I wrote "Omnichannel Retail: How to Build Winning Stores in a Digital World" with Tesco's architect of Clubcard, Tim Mason. A second lens really does pressure-test every claim as you go. As I said, going solo means the argument is wholly yours. You know, the burden of truth sits wholly on your shoulders. It was good to be able to just not have to negotiate that shared structure. I remember Tim very eloquently once said to me, I think very appositely said to me after the first chapter review that, "It's like you've been in my head. My head is my house and you've moved the furniture around. I'm walking in and I can still see everything, but you've moved the furniture around." So this is very much my head, no furniture being moved. And I think I had to replace the missing co-author bit with everything that I could bring, which was 25 years of practice. I did over 30 practitioner interviews, and I definitely refer to my work with the KPMG Retail Think Tank in the UK because we're the only one of our kind forecasting what's going to happen in retail as well. So, 25 years of experience trying to bring to bear just on my shoulders.

Adrian Tennant: Well, your book draws a distinction between needs, missions, and journeys. Now, shoppers' underlying needs haven't changed much, but every search and contactless payment now leaves a digital trail. Miya, what can that trail tell a brand about the mission behind a purchase?

Miya Knights: So, start with a bit of context. I think Amazon established that having a digital connection to a customer certainly tells you a little bit more about them than a one-time cash-paying anonymous customer. You get certain demographic information, maybe where they live, what kind of credit card they have, etc., etc. And that digital summation gives you the data that drives the inference around their needs. But needs are enduring. We're always going to need to eat. We're always going to need to find shelter, etc. To your point, missions are temporary configurations for fulfilling that need with context and in terms of the constraints faced in the moment. Missions are defined by three variables: urgency, effort tolerance or friction, and perceived value. So I might need something, but in the moment, how much do I need it? How close am I to getting it? How much am I willing to pay? These are the kinds of things that guide our journey to fulfill the need, which is to buy and use the product or the service that we wanted. I think retailers are drowning in data and they can really start to understand the needs, particularly if you have frequency. You get that wealth of information that Tesco's in the UK showed the rest of the world that you could access through Clubcard. It was the world's first mass grocery loyalty scheme and now it has a UK penetration of about 84% of all households. If they don't know what kind of toilet paper and washing-up liquid and what we like to buy to treat ourselves on Friday, who does? But in terms of why we bought what we bought, we can see the who, the what, the how, the when, and the where, but that still only allows us to infer the why. What I argue is that missions, the motivations behind the need, that sends us on the journey, whether it's fast or slow, informed or high touch with human information, human reassurance, human reviews. That all depends on the mission as well as the need.

Adrian Tennant: Well, of course, much of the digital investment of the past decade has gone into removing steps between a shopper and a purchase. Your book usefully sets out a matrix that runs from low-value, high-volume purchases such as groceries to high-value, low-volume purchases such as luxury goods. Miya, when should a brand add friction on purpose?

Miya Knights: Yes, I'm glad we get to talk about this. This is at the very heart of calibrating how much of a digital connection you need to your customer and how much you can rely on the data that it drives to guide insight and strategic technology decisions. If you think about an XY graph, on the lower left-hand side you've got low value. But low value products tend to be sold in high volume. And the apex example, sector example of that is grocery. But in the absolute total opposite right-hand top quadrant, you've got high value, something like a luxury good that's actually sold at low volume. And I think the principle of friction optimization is the important thing here, not friction elimination. If you accept that mission value matrix maps against economic value, the human touch that mission really needs tends to focus on the high value, low volume products. You need more of a consultative sale. I always make a joke about, would a gentleman buy an engagement ring through an agent? No, they wouldn't. They'd want to go and see it and touch it and feel it and talk to experts and possibly get some specialist customizations in terms of fitting. This is where friction in terms of adding the human touch, maybe adding scarcity to drive value up can actually create value. Those high-value, high-touch missions like luxury furnishings made to measure, consultation actually reduces return risk, scarcity preserves price integrity, and friction becomes a quality signal. My anecdote is always from COVID when I interviewed a furnishings retailer's digital director. At the time in this country, in the UK, the UK government was throttling footfall in a store, so you could only have a certain number of customers in the store, which led to queuing. When they were allowed to reopen and allow customers in, they introduced a booking system, an appointment booking system online, and they actually found that they arrived better informed, converted more often, and spent more. Luxury waiting lists and private previews are scarcity economics. They're not operational failure, but only if they're calibrated against urgency and perceived value. So, strip out friction in low value, low touch, high-volume sectors like staples and commodities, but you have throughput and you have frequency, things like one-click, reorder, self-scan, automation. But the question isn't whether to remove friction, it's where. A £3 basket and a £3,000 basket, buying one of those should never feel the same as the other.

Adrian Tennant: Digital screens and smart carts are turning the store into a media channel that can be measured in ways that used to be possible only online. For a consumer packaged goods brand buying in-store retail media, what results should it expect a retailer's media network to report?

Miya Knights: Right, diving straight into that sort of area where retailers have, as you said, become very used to being able to track our every move online. I think it really has shone a light on the store. In "Omnichannel Retail," Tim coined the phrase "stores are a digital black hole." I think actually he may have even borrowed it from Stephan Schambach's book, "Mobile Makeover." I've been talking about stores being digital black holes for a long time. Since COVID, we kind of do the work for them as well when it comes to groceries, scanning and picking and packing. So when it comes to retail media networks and the digitalization of the store, every digital surface becomes an opportunity to personalize the journey, to react to customers at a much more granular level. But I think the difference is that CPG brands that are now being asked to maybe divert funds from brand-funded advertising to retail media, shouldn't be looking for impression volume. A store run as media, as one of my interviewees says in the book, Ruth Harrison, is where impression volume is the only metric trap, and it becomes noise if that happens. It erodes trust, and physical retail depends upon, I see a price, it's the price that you're going to expect to pay. I think the more powerful thing that CPG brands should look for is closed-loop attribution. It will allow them exposure, understanding of behavior in a much more granular way, and through to transaction in one system. It's not reach. I think it's possibly the holy grail for brands, being able to match an activation directly through to a transaction. Expect things like margin per visit and coupon redemption by surface type. Margin per visit means product margin plus supplier funded media. These are the kinds of new metrics, I think, that we need to move on beyond. To bring it to light, the proof point that I use in the book, and I think it's possibly one of the very few that are out there at the moment, is China Resources Vanguard, which is the national-run grocery in China. They worked with Suntory, the brand, on an NFC electronic shelf-edge label pilot. They used tap to engage the shelf-edge for coupons and they found 24% redemption versus 1% to 2% for self-operated in-store coupons. And five to nine off-site and they still got the same results, 15% for beverage and 35% uplift for liquor shelf share as well. So, I think it's in that moment, that very basic simple example in that moment engaged with the product and offer a closed-loop resolution to the brand but their advertising actually resulted in dollars. It matters because it's a new revenue stream with incredibly high margins. You know, in Europe, EBITDA is about 2.8%. Walmart's a bit healthier at about 3.5%. Retail media operating margins run at 50 to 70%. Amazon's retail media margins are just vast and revenues are vast. But I think the interesting thing that Amazon can't play into, maybe a little bit with Whole Foods, is that they don't have a store network. We've seen 7-Eleven playing with in-store radio. We've seen Walmart do the same thing. Finally, I think it allows adjacent brands to also get into the retail media game. With retailers as publishers, that wouldn't necessarily work together before. A working example would be Ocado, the online grocer in the UK. They work with Halfords, an automotive retailer, to go and get your tires checked and your windscreen wipers and things like that. You can't buy a Halfords product on Ocado's website, but they want to advertise to Ocado's audience around a bank holiday, seasonal event, knowing that they were as high as they wanted to target. So all in all, it's a win-win for retailers, but you really have to close that digital loop, remove the digital black hole, and offer brands some tangible extra closed loop value over their existing funding.

Adrian Tennant: Let's take a short break. We'll be right back after this message.

Adrian Tennant: Welcome back. I'm talking with Miya Knights, author of the Bigeye Book Club's November selection, "The Future of Shopping: Using Digital, Data and AI to Win with Customers." Miya, let's talk about shopping by delegation. A recent study from Accenture forecasts that more than a third of all shoppers will use AI assistants for holiday shopping this year. They also reported that more than a quarter of shoppers trust a personal AI agent more than themselves to research products and build a shortlist for them. Miya, what makes a brand's product selectable when an agent is doing the first round of choosing?

Miya Knights: That's interesting. I've seen so much research saying that consumers don't trust AI, but that research points to the very same reason that technology and digital have become such profound enabling tools in terms of their impact on our industry. They're able to parse vast amounts of data beyond what we could do in terms of our cognitive load. You could argue that digital has increased our cognitive load, and so answer engines and AI have really kind of appealed to that essential human nature of help me find what I'm looking for, make sure it's right for me, make sure it's the right price for me. But what makes a brand's product selectable, I think, is what's really worrying brands at the moment when an agent does the first round of choosing. That delegation shifts power from interface design to data architecture. You know, we were very, very obsessed about pretty PDPs and 360-degree spin and true color resolution when we first embraced e-com. But homepage merchandising and banners can lose influence, and the battleground becomes more about pricing, transparency, promotional effectiveness, inventory accuracy, and most importantly, machine-readable product attributes. I think what we're seeing with agents and answer engines, Gemini, Claude, ChatGPT, and the rest, these agents are weighing history, budget preference, delivery constraints, brand exclusions, loyalty memberships. I often say nowadays that to be chosen, your product data must be clean, structured, and legible to a machine. Structural defenses, not embellishments. Things like loyalty benefits, subscription pricing, exclusive assortments, and bundled value really have to be machine readable. I think the most exposed are mid-market brands with neither a cost advantage nor a strong symbolic differentiation. I think every brand can do something to increase their exposure and their chance of being chosen. Those that are going to win that race, as it were, win the answer engine's results, are going to be those ones that are really good at what they do, use their digital flagship that if you land on it, really wants to buy from them, but equally are really good at what they do in terms of talking to those answer engines as well. Oh, finally, I'd say when the algorithm is making shortlists, your product data kind of becomes your storefront. If the machine can't reach you, you won't exist.

Adrian Tennant: Right. Well, of course, the holiday shopping season kicks off this month with Black Friday and Cyber Monday. Which shopping missions are consumers most likely to hand over to AI agents first, do you think?

Miya Knights: So it's back to that framework we talked about. It's chores before choices. Shopping is still fun for some people, particularly in-person shopping. But when you're looking at those more digitally driven touch points, I think high urgency, low value, routine replenishment missions will be the ones to compress first. We regularly see every year now, Christmas delivery slots for online groceries, online shopping, go really quick. People don't want to schlep around the supermarket buying all of their products and, God forbid, there isn't something in stock. So, at least if you order it three weeks in advance, you know it's there. So, for those low-risk routine buys, I think funnel compression, with this year being the first proper agentic Christmas, will be near total. Low-touch, low-value categories. Again, agents already add obvious value. You can monitor consumption, you can monitor prices, optimize delivery windows, auto-substitute. By contrast, I would suggest high-consideration, identity-laden missions. Because when you think about need, we used to buy based on our basic needs, but now we also buy based on wanting to identify, wanting to be part of a community as well. Those kinds of missions are going to resist agentic disintermediation for longer. They definitely still need reassurance, ceremony, and human judgment. I think over the holidays, expect known-item gifting and replenishment to go to agents before anything symbolic. As I said, we'll delegate the chores before the choices. Nobody outsources buying that engagement ring. Plenty will outsource buying toilet paper.

Adrian Tennant: Yeah. Well, a moment ago you mentioned the use of digital shelf labels. In the US, there's definitely more awareness now among consumers of dynamic pricing. Most of us are used to the concept with airlines and hospitality. Miya, I'm curious, do you think dynamic pricing is inevitable in some respects across a broader range of categories in the future?

Miya Knights: Yes, I do. Yes, I do. But I think we have to be really specific about what we mean by dynamic pricing right now. I think there's a lot of confusion out there about whether or not I'm going to get the same price as you or not based on certain data that the vendor has on me. That's definitely not dynamic pricing. I think we are a long way off personalized pricing, but we already see dynamic pricing and I think are familiar with it in terms of surge pricing. Who would argue with a business who sees something that's in greater demand at a certain time and raises their prices? I think it's... we don't necessarily like it, but we understand it. I remember saying years ago that Amazon could change hundreds of millions of prices a day and compare those prices with hundreds of millions of other prices from retailers. Every time a retailer wants to change prices, a retailer like Walmart wants to change prices in its stores, it has to send an army of people through the shelves, through the aisles with belt printers. That's a more realistic, I think, near-term use of dynamic pricing to make sure that the price at the shelf edge is what you pay at the till, and also to maybe change prices more frequently to be more attuned to the pricing fluctuations in really thin margin sectors like grocery, where you can make or break margin on a product by raising it by one or two pence, one or two cents.

Adrian Tennant: Well, as regular listeners know, we love case studies on IN CLEAR FOCUS. Miya, could you walk us through a retailer or a brand that's connecting what happens in its stores with what it knows about its customers?

Miya Knights: Yeah, absolutely. There's a store I always talk about. I actually call it the store of the future because I think the store of the future is here in Lisbon, in Portugal. You might have heard of MC Sonae. They are Portugal's largest grocery food retailer, and they have opened a store under the banner Continente. It's a Continente Bom Dia store in Leiria. Essentially, they've worked with a technology supplier who I admire for their sticking to their guns over all this time of development of autonomous stores. It is the world's largest autonomous grocery store. It's got over 1,700 cameras and over 2,000 shelf sensors. The reason it works and is different to Amazon's Just Walk Out technology and various other retailer trials is that they have kept the checkout. What they've provided is a scan-free solution. So instead of having to identify yourself by picking up a gun, pulling out a smart cart, everything in the store is as you would expect it as a normal store. There's a deli counter, there's a coffee machine, there's croissants in the pastry, in the bakery section. But the very difference is that you just walk up to the checkout, it pre-populates your basket, you check it, you pay, and you leave. The problem with the Just Walk Out technology was, one, it was very expensive, and two, it barred customers that didn't have the app from coming in. I don't know what retailer can afford to do that. And then a lot of customers were always saying, are you sure I can leave? Please can I leave? And there was always that fear, that lack of trust that I'd get double charged for something. And even though Amazon's a behemoth enough to say, "Just keep it," if the technology didn't work, there was that doubt that they didn't trust it. By keeping the checkout, Continente Bom Dia has managed to keep that level of reassurance that yes, you can walk out of the store, you have just authorized the payments, the transactions. And there's two stores on either side of this. One has the technology and the other one doesn't, and they are finding people are actually driving further across town to use the new stores because they prefer it. The queues are much shorter, throughput is much higher, but I think the final payoff from a return on investment perspective is that when I interviewed MC Sonae's head of research and development and innovation on the stage in Athens last week, Marlos da Silva, who said to me, "We used to have a black hole. We now have a glass box. We can deal with loss prevention, we can deal with shrink, we can deal with out-of-stocks, we can optimize planograms, we can optimize retail media and brand-funded advertising. Dwell time, throughput, traffic, hot heat maps: we can do it all through these sensors and these cameras." And so we can actually not just give people more of what they want, but we can use the technology to serve them better and run our business better as well.

Adrian Tennant: Fascinating. Yeah, that's really interesting from the retailer's perspective. Of course, shoppers are sharing far more personal data with retailers now than they did, say, a decade ago. What do you think we expect in return as consumers when we walk into a store?

Miya Knights: So I did some research on this. We expect to be treated better than a one-time cash paying anonymous customer. Recognition is what it boils down to. The research that you do as an analyst in this industry when you talk to consumers is what is it that influences you to choose a retailer or a brand? And every marketer and every retailer knows it's always been the three Ps: price, promotion, and proximity. When I did this research back in 2020 for "Omnichannel Retail," it had changed to price, promotions, and recognize and reward me for my continued custom. So the payoff for data is being remembered in a way that shortens decisions, reduces repetition, lowers perceived risk. You know me. I trust you more. I give you my data. So trust, consent, and restraint aren't constraints on sensing. That's what makes it work. Systems that respect uncertainty and preserve choice will generate better signals over time than those that overreach. There is a very clear give to get dynamic. To boil it really right down to the essentials, it's what are you asking for the data? What is it that you need to know about your customers in order for you to know them better so you can find more like them, serve them better, and run your business better as a result. Probably best to focus on who your best customers are and the characteristics of them. And then the expectation, the payback is relevance. It's not surveillance. So personalized help that feels earned. A food tolerance example, you know. It's cool, not creepy. Customers will trade data for recognition and reward, not surveillance. The deal is, remember me. So, remember me and reward me.

Adrian Tennant: Brand and agency teams can now use generative AI to produce product copy and campaign content at a volume that would once have needed a much bigger team. When you think about the future of shopping, where does human judgment make the biggest difference?

Miya Knights: Wow, that's a big question. I think human judgment still makes the biggest difference when it comes to creativity and innovation. Again, the rule applies that it doesn't create innovation or make you creative, it just enables it. So, from the very inception of ideas and new product development, I would hope that AI tools can be used to speed time to market, vary by sector in terms of regulation and compliance to all those boring areas. You talked about generative AI. I think you've got to be really careful and sensitive about using generative AI in marketing. There was an equestrian clothier's firm, I'm sure he won't mind me mentioning the name of the company, it's LeMieux.com, and they've banned generative AI in their advertising because their customers absolutely hated it. I think when it comes to packaging design, making sure that your printed package on the shelf renders in as powerful way on the marketplace digital shelf, absolutely there's room for flexibility. But if you're generating images of people or scenarios, videos, that kind of thing, I think we're still seeing retailers and brands stick to what's real and what's new. I believe there was a Louis Vuitton Christmas campaign just launched that used actual fine artists and painters to create it as an antidote. I think we're entering a phase where people crave authenticity as well as relevance. As long as the AI can be relied on to manage its scale, manage by exception, and harness creativity and innovation, in much more powerful ways, we'll be okay.

Adrian Tennant: Great conversation. Miya, if listeners would like to learn more about you and your work, or of course your latest book, "The Future of Shopping," what's the best way of doing so?

Miya Knights: Thanks, Adrian. "The Future of Shopping: Using Digital, Data and AI to Win with Customers" is out on Kogan Page; that's KoganPage.com. I'd like to offer your listeners a 25% discount using the code BIGEYE25. And they can also find the book on Amazon, Barnes & Noble, and Bookshop.org. You can find me on LinkedIn for the running commentary. You can look at the trends that I think are shaping the industry through the coverage we have of retail technology implementations and operations and trends in Retail Technology magazine, and you can see what I'm up to with KPMG Retail Think Tank every quarter as well.

Adrian Tennant: Perfect. Miya, thank you very much for being our guest this week on IN CLEAR FOCUS.

Miya Knights: Thank you, Adrian, for having me. It's an absolute pleasure.

Adrian Tennant: Thanks again to my guest this week, Miya Knights, author of "The Future of Shopping: Using Digital, Data and AI to Win with Customers." As always, you'll find a complete transcript of our conversation with timestamps and links to the resources we discussed on the IN CLEAR FOCUS page at bigeyeagency.com. Thank you for listening to IN CLEAR FOCUS, produced by Bigeye. I've been your host, Adrian Tennant. Until next week, goodbye.

Timestamps

00:00: Introduction to E-commerce Trends

02:30: Miya Knights: Author and Retail Technology Analyst

03:45: The Challenges of Solo Authorship

05:30: Understanding Shoppers' Needs and Missions

07:15: The Role of Friction in Shopping Experiences

10:44: Retail Media Networks and Measurement

15:35: Miya's Book Promotion

16:38: Shopping by Delegation and AI Assistants

20:11: Holiday Missions Shoppers Will Delegate to AI

22:21: The Future of Dynamic Pricing

24:09: Case Study: Continente's Autonomous Grocery Store

27:38: Consumer Expectations for Data Sharing

29:47: The Importance of Human Judgment in Creativity

31:38: Conclusion and Resources

Other Episode

View All