- 22 Jan 2020
- Managing the Future of Work
From opt-in to check-out: How digital platforms are transforming retail
Bill Kerr: New digital technologies have dramatically transformed how consumers buy things. While new businesses born in a digitally connected world have thrived, traditional retailers have struggled to keep pace. Welcome to the Managing the Future of Work podcast from Harvard Business School. I'm your host, Bill Kerr. Today I'm speaking with Dan O'Connor, CEO of FrontFour Ventures and a visiting executive with our Managing the Future of Work research initiative. Dan's career advising America's largest retailers has spanned two dramatic transitions. He's going to give us an inside look at the changes that are unfolding and share guidance for leaders working to navigate them. Welcome, Dan.
Dan O’Connor: Hi, Bill.
Kerr: Dan, let's start with a little bit of terminology. You use a phrase, G1, G2, G3 and G4. Define for us, what are the generations of retailing that we're talking about?
O’Connor: In every business it's useful to have a framework to map, view, model the transitions as industry moves from one business approach to another. Sometimes people use a term called S curves or Experience Curves to describe this.
Kerr: Where you go from 0 percent penetration up to 100 percent penetration.
O’Connor: Sure. And a great example and my favorite case here is the Netflix case where they talk about that industry being one that originated and grew out of the pick, pack, and ship of DVDs and grew into this industry that was built on streaming. And that inter-generational rotation between DVDs and streaming was obviously a significant change in leadership, asset allocation, talent and so on.
Kerr: Okay. And so in your terminology for retailers, you're having the three transitions from G1 to G4. What are each of the transitions?
O’Connor: The traditional retail trade—the fragmented trade or the mom and pops as it's often called—these are retailers that sort of operated one market at a time. They were traditionally supported by a wholesale distributor network and were often important in the distribution of large national as well as regional brands.
Kerr: Okay. So Main Street businesses that were on every street corner for a long time. Then G2 comes along. What’s G2?
O’Connor: G1 was very fragmented and G2 represented in a sense the Walmart era, the Carrefour era, the Tesco era, the era of the modern trade, where they brought a level of standardization As they aggregating concentrated volume into the merchandise that they chose to resell, they drove even new economies with the suppliers and those economies started at manufacturing and through transportation out through the store. It's important to understand that G2 retailers didn't create new demand. They took it from preceding generations of retail.
Kerr: Some lowering of the cost structure so that maybe we're buying more goods and we aggregate up some of that demand into specific locations.
O’Connor: Absolutely. And I think people talk about Flywheels in the modern vocabulary today. And that cycle is what really drove a 25-year cycle of what we call the modern trade. And that really spread out to many parts of the world, top 40 countries in the world.
Kerr: Okay. G3 comes along and I'm guessing that e-commerce have some variety. Tell us, what’s G3?
O’Connor: Marketplaces. Companies like eBay, companies like Amazon, Later, companies like Alibaba and jd.com over 250 of these sort of marketplaces grew up in the world. And they were just digitally connecting businesses with businesses, businesses with consumers and consumers with consumers. They were giving consumers an alternative in many instances, to buying a directly from a store. The second new entrant that we saw into this marketplace were what we call the intermediaries. And the intermediaries were companies that were digitally influencing the outcome, what people bought, where they bought it. So companies like Facebook and Google for example, were really in the business of aggregating really large audiences.
Kerr: Very symbiotic relationship between the intermediaries in the marketplaces.
O’Connor: And initially these were very different companies, that is, the marketplaces were not intermediaries. So search was being done outside the marketplaces. And the third level of innovation that we saw that was really different was the growth of digital influencers. These are individuals who aggregate at large audiences typically around a specific topic. And they were really influential in moving the consumer's point of view from learning and studying products at the shelf in a store—and in a sense they move that first moment of truth from the shelf to a destination, a digital destination.
Kerr: This being Instagram feeds or other ways that they're impacting what you're thinking about even before you use this search term inside Google.
O’Connor: That's true. And then even when you do search, you may find yourself in a chat room or in some kind of a YouTube library with an expert who has a point of view. And these are people who are out there that are building audiences of 25, 50 a hundred million viewers that are regular viewers, subscribers.
Kerr: So give us the last generation in this sequence.
O’Connor: G4 is the emergence of large-scale platforms and large-scale platforms that have are multi-market place, they incorporate eight-, 10-, 15 different digital marketplaces. They provide a space and a point of connection for influencers. And in many cases, they actually own and operate large intermediaries such as social firms, gaming sites and search. And so companies like Alibaba, which I think is really the gold standard in the world of platforms today, you pair that with this underlying architecture, which is really the elegance of these platforms. It's the ability of the platform not only to enable their digital commerce, their social platforms and so on, it's that they have data and marketing data and analytics, they have supply chain capability, they have delivery capability and so on. And they build those capabilities beyond their own need and they resell those. So examples of those are the Ali[baba] Cloud, are the Amazon Cloud [Amazon Drive] and these turned into very large revenue streams.
Kerr: Okay, so this is e-commerce on steroids and with lots and lots of aggregation up from, what were these individual actors before.
O’Connor: Yup. I believe that the world that we've just described is going to move from one where there was 250 large marketplaces and there was tens of thousands of intermediaries and influencers and I think that world aggregates down into about 25 large platforms. The way we create demand inside of these platforms will change substantially.
Kerr: So walk us through … when you've worked with retailers, how do they go from the G2 big box into the e-commerce?
O’Connor: If you really looked at how this evolved, why did they successfully make the transition? One was, is that they got to some new operating norms. So they reset the vocabulary, they reset the organization around what it means to have online and offline and to try to create a sort of synergistic relationship. Secondly, they established a vocabulary around that so people were all talking about this the same way. And thirdly, very common metrics, so that people understood that they were being rewarded and not penalized by the growth in both their store business and their digital business. And the really high level picture on this is that I think history will show that there's companies that successfully made this migration were companies that really got their capital in the right place, the right kind of store network, the right kind of digital architecture and they got the right talent in place. People who were already versed and well trained and have a good understanding of digital commerce and yet pairing that with this culture that comes out of generation two, a very strong store operations capability.
Kerr: The capital allocation is very important point, I want to surface that up a little bit more because you can think about advances as being things that increased your operating efficiency, you get more profit per asset base. And then there's the overall asset intensity and are your stores in the right locations? Is your manufacturing footprint the right footprint? Talk us through how much e-commerce has changed capital choices. And then that gets to where companies have to make their bigger bets, their bigger decisions as to how to move forward.
O’Connor: Every element of the supply chain has started a level of digitization of their own business. And the outcome of digitalization is data. Every player in the value chain can begin to share that and begin to synchronize. And in fact the long-term impact of all this is that the data itself will really orchestrate the supply chain.
Kerr: So this is between Walmart and Procter and Gamble & then Procter & Gamble and its suppliers all the way back up through in that. To what degree would you characterize as being an automated information flow or how much was that all in the same kind of platform?
O’Connor: We're moving into this world of the algorithmic economy, enterprise to enterprise connection or sharing of data. Your bot can go and access my data in a very safe way, leave the data there and go back with a better forecast for example, or go back with the pricing analytics or the cost analytics. The algorithmic economy is really going to change the way, for example, planning takes place all the way along the value chain.
Kerr: Okay, so from G2 to G3, and then G3 to G4, we're evermore allowing the data to freely flow up the supply chain, value chain, have people making decisions, products get shipped down and it becomes more algorithmically based versus me calling you up as the Walmart rep and saying, "This is how detergent's doing this week."
O’Connor: It's algorithmically managed where there are people carefully placed along that digital value chain that are reviewing and editing. But not having a hundreds of planners running very large Excel spreadsheets to try to figure out what's going to get shipped here and what production is going to get allocated to what demand points. So the very nature of the work has changed significantly from you 1990 and the year 2000 to where we sit today.
Kerr: You also see things like dollar shave club or Nave deodorant. These brands that come up and they don't even have a physical presence before they're online. How has that changed the dynamic for a retail commerce organization?
O’Connor: With the level of data and the ability to use that data to isolate specific opportunities, I think what we'll always see is the rapid introduction and scaling of what we'll call “microbrands.” And sometimes microbrands will be built with an attempt to scale them beyond a couple of hundred million dollars and try to create billion dollar brands. That as the world digitizes and as entrepreneurs as well as “intrepreneurs,” people working inside of large companies, learn how to work with these kinds of analytics and work with this data, they'll see more and more consumer opportunities, production and supply opportunities and they'll find a way to build brands around this.
Kerr: Okay, because we all get the same amount of shelf space on Amazon regardless of the product. You get one page that you're building off of. I can't block you out of the aisle the way we used to and you can get the manufacturing, everything else set up through third-party outsourcing and the like.
O’Connor: If you're not on page one in the search results, you are blocked out. And the numbers are profound and they vary by category significantly. But if you're not in the upper left hand corners—number one slot in search—that's where something like 60% of the volume and some of the categories go into that one search result, not the entire first page. So constant volume in many ways is concentrating in a way we never even dreamt on a shelf. It's really amazing. So search is just one of the ways.
Kerr: You mentioned earlier—it was about getting the senior leadership team onto the new model and onto how they were going to both be in a G2 world and enter the e-commerce space, or be in e-commerce and move to the next thing. How does that play out for senior leadership teams? Is that the big sticking point? And then what can they do to be better at that transition?
O’Connor: About 20 years ago, with a large consumer electronics retailer operating in the US, I remember attending an executive committee meeting. And they had launched a website and not much was happening. They really understood the store based world and they really didn't understand a world that included stores but had this digital overlay. These were well-intentioned people that had not really re-skilled themselves. There was a new CEO within a year and an entire new leadership team and today it is one of the very few specialty retailers that has the kind of market cap and growth rates, in its entire sector. CEOs and boards have a lot to concern themselves with in these kind of transitions. The first one is can we get our assets in the right place? Can we get our investment? And if you really look at the winners, all of them have re-architected their balance sheets. So their level of investment in real estate, their level of investment in inventory, their level of measurement in digital capabilities, so the balance sheet gets re-architected significantly. Secondly, there's significant changes in cap-ex to reduce op-ex because the digital world requires a much lower operating expense sort of ratio.
Kerr: And can you explain why to everyone?
O’Connor: The pressure on profit is often coming through price transparency. And so there's a lot of margin compression. In general, very large brands can experience as much as a 50% margin reduction as their category goes from zero percent digital to 25 percent in digital—not 50 percent digital but 25 percent. So that obviously then back to the ROI equation that says you have a couple of other levers, one is op-ex, which you can only hope to bring down at the same speed of any pricing pressure, but it's often hard to do because it's about people—your largest expense is people. As we look at the transition from G3 to G4, one of the critical elements in this sort of algorithmic economy, is to leverage data and leverage, you know, analytics in effect and robotics. I mean generally all sorts of automation to reduce repetitive tasks that can be automated and redeploy those people to new tasks and new careers.
Kerr: Let's say I'm Walmart and I'm able to get you as both an in-store customer and also an e-commerce customer. Is that additive or do they sort of trade off?
O’Connor: Today we're moving into this world that's a real recognition that if you can get a customer that is doing 25% of their consumption—that is their store based consumption—with you. Let's say you're Walmart or Carrefour, you're getting 25% but you're only getting 5% of their digital consumption. What we're learning is one neighborhood at a time. If you could move everybody to an equal market share that drives up the density of your digital volume in any particular neighborhood. If you can capture significant market share of where it's discovered online, ordered and paid online, but either picked up in the store, or delivered at home or delivered office, wherever delivery wants to happen, the higher your market share is in there, the more productivity you'll get out of your distribution network.
Kerr: And so it lowers both the cost to serve but also makes to where I can say in 20 minutes you can have this product handed to you somehow, some way.
O’Connor: The model store for this is the Hema store, the [Fresh] Hippo store, which is owned by Alibaba, where their brand promise is on roughly 6,000 SKUs, 30 minutes from order to delivery within three kilometers. And so what that means to the store or wherever you're picking it from is that you basically have six minutes to pick that order and then you have about three or four minutes to get that on the back of a delivery vehicle. And then you know, on a motorcycle or bicycle or however it's going, you have 20 minutes to deliver it to that consumer's door. And that's become the standard in these high-density markets. We're redeploying our capital. We're moving from big stores to small stores and small stores that are being built first as a delivery node, knowing that 60% of the volume in that store is going to be delivered. So merchandise will come in the back of the store and it'll go into a mini-pick center instead of onto the selling floor. At the same time, when a store is designed first as a picking node and it's going to do 250,000 orders a month, enormous volumes. You can run a front end of the store that's all about fresh. That's about 40 percent of what consumers buy in supermarkets and it is what you brand yourself on today.
Kerr: Dan, this has been fascinating. One can also very clearly see how moving from G2 to G3 requires different leadership and how that can be a roadblock in a way. I want you to take us all the way out to G4 and Alibaba and we've talked about the Hema store. What are the other things that we should be learning about from this experience?
O’Connor: If you're running any big category in a large consumer product company, you're going to have an increasing number of micro brands coming after particular geographies or product segments. The second major characterization I think will be that we're competing increasingly with speed. And maybe it's speed to messaging, maybe it speed to marketing, maybe it's speed to delivery. So ready insights that drive brand entrepreneurs, this world where the data is sort of running between organizations at an ever-increasing speed.
Kerr: What about product development? Like how does a traditional manufacturer compete in this environment? Will they use the platforms? You have a billion people that are on the Alibaba system. What does that do to my product design?
O’Connor: What this data allows you to do, what the aggregated data that companies like Alibaba have, the team all have. What this allows you to do is run a very C to B kind of enterprise. You'll be able to in a very real time way connect with, ideate with, product test with consumers on scale in a way where their data can be viewed and measured and by algorithms that will help you isolate what the big trends and what the big ideas are.
Kerr: Can you give us an example of this? How would this play out if I wanted to make a tee shirt?
O’Connor: What this allows you to do is to target an item, test an item, leveraging their data and new marketing and media techniques. So things like Taobao Live are just one way that you can sort of visually connect with a consumer and say here's a product and let people respond. And what you see is you can get and measure the viewership, sort of the clickstream. You can measure the conversion rates and you can begin to see what really drives consumption. We're near the point today where large-scale platforms are not quite there, but we're near the day where as a large brand they'll be able to tell me when a product needs to be restaged. In other words, within 18 months, based on the data we see today, that product needs to be reformulated, it needs to be restaged. So the data will become predictive. Run your data against these trends and just let us know the strength and I want to know the difference between Boston and Shanghai. This whole notion of disintermediating agencies and third party information companies to going to direct data that will actually give you real time or reasonably real time, I call it near time data, and near time metrics, that someone who is a brand owner can use and begin to really drive their business in a new way is-
Kerr: It's fascinating because in the, in the G3 space where we already are, you're now able to go down to the level of this individual convenience store and say based upon the people within a half a mile of you, you should be selling more of this product and what's gone wrong. And then as you look ahead to this future, it gets ever more into just the simple ideation of there is a missing product in this space and now we know how to fill that. And we can aggregate that up. Is there a place in the future for the independently owned, fragmented retailer?
O’Connor: I believe that the very large platforms are going to make it increasingly easy for one and two store operators to digitize their stores in a way that's not dissimilar to the kind of automation that large scale retail chains have today. So what that means is the ability to really digitally capture my point of sale data, my inventory data, my forecasting data, my order data, and just to…
Kerr: You come on Alibaba's platform and you get access to all these tools for your business.
O’Connor: Exactly. So the same way that Salesforce provides CRM capability as a service. Imagine the world where—whether it was Jingdong or whether there was Amazon or whether it was Alibaba—you had basically the retail enterprise as a service. So what we're seeing is the G3, G4 leaders are actually leading the digitization of G1. And this gets back to the notion that the competitive advantage of the platform companies is the data and the insights. The competitive advantage of the local store is as a delivery node. And so if you actually are running a large digital platform, you can see the advantage of automating all these mom and pops, because then you can begin to get the data on what's happening in the local stores, what are the high demand items you can provide a supply chain for them that might be a little bit more efficient. You can definitely get them better pricing because this is the issue with small mom and pops is they don't necessarily get the very best pricing. But as importantly and really important, this avoids the capital deployment against building stores for the large platforms. So I can network 10 million stores, 6 million stores in China through this retail as a service initiative and all of a sudden I'm running a really large and in-the-community kind of distribution network. And also in suburban and rural areas, I can get The markets where there's only one or two.
Kerr: Throw in some drone delivery to those nodes. And pretty soon you've got the whole system orchestrated.
O’Connor: And the announcement by UPS that they've been licensed now to run the first drone network here in the US is really significant. It's really important for everyone to know that for, you know, brands of all type, G1 retailers are still 51 percent, 52 percent of all retail in the world today. And what I just described in the world of retail as a service, I think this will enable them to, in a sense hold share. G4 retailers, that G4 will pull more volume out of G2 and G3 so what that means is that we see weakening marketplaces and I think there's some examples of that here in the US where they've sold off some parts of the market CEO's have left. And then in the G2 world I think we see the same thing happening. And you can see this in retail centers around the US, the mall world for example. So that volume is being pulled into new places. G2 may be more impacted by three and four than G1.
Kerr: Okay. Dan, there's a lot of organizations and a lot of boards and CEOs that they feel like they're flat footed and it's this world of crazy change that you're describing. Is there any advice you'd give them as to how to get started, what to do next?
O’Connor: Everybody has to begin to sort of rationalize for the algorithmic economy—this whole G4 world. How will we create demand? Are we going to still float coupons? Are we going to move to some kind of a different and more effective approach? And, and do we still need a sales force that calls on all of these G2 retailers and G1 wholesalers? Or are we going to, in effect bring our marketing organizations and our shopper marketing organization, all these different marketing entities, into a single entity that's responsible for demand? And will we keep it as a large functional approach? Or will we move them into, so squads, for example. So the point is, as a leader, just understanding the vocabulary of what I simply described to you on the supply side and really understanding how this is unfolding is critical. On the supply chain, I would really want my leadership team to truly understand what the future of distribution was going to look like and how I was going to connect with different customers—connecting enterprise to enterprise and getting to a common forecast. The algorithmic economy starts with forecasting. It's this notion of really getting a better sense of demand and moving it from a sort of a 70 percent confidence level to a 95 percent, 98 percent confidence level on both sides.
O’Connor: Once you understand the mechanics of how these things unfold, then there's this issue of governance. Increasingly there's going to be fewer people for your people to talk to. So you're a brand, you're calling on a retailer, but there's going to be fewer people for you to call on. Secondly, as happens at Amazon today, there really isn't a role for a human-to-human connection. More decisions, increasing number of decisions are being made out of sight, machine to machine. The center of strategy today, is being able to influence the data and be able to manage through the customer algorithms to get the space planigram that you're looking for to get the pricing that you're looking for to get the promotion that you're looking for.
Kerr: What takes consumers onto these platforms? Do they stick around for a long time? You hear of platforms offering life insurance as a product area. How is this dynamic changing?
O’Connor: These large-scale platforms increasingly incorporate entertainment, payment, financial services. They increasingly surround that with all aspects of your life, all of that in a single app. And this, in the world of these sorts of super apps where everything is combined into one sort of seamless environment really raises the switching costs for consumers. Consumers go into those environments, they transact in those environments. Those environments become increasingly predictive and they make your life simpler. They help you understand what you need to buy, when you need to buy it. They help you stay entertained, they help you get tickets, they have you and they simplify your life. Switching costs are very high. In the G1, G2 world, switching costs were very low. So one of the big distinctions as we kind of moved from G2 through three to G4 is that switching costs for consumers are going way up and this allows these environments to scale even further. Scaling is all about capturing and retaining these consumers in the platforms.
Kerr: Well, Dan, has anyone ever been able to break through in one of these situations?
O’Connor: There are not a number of examples. Pinduoduo—which is for short called PDD—in China is a company that used aggregated buying where consumers can create informal networks, and then those networks through a messaging app can aggregate their volume and buy together in PDD. And what we've seen is that this has gone in a period of eight or nine years from a new company with no following at all to the second largest marketplace today in China.
Kerr: Yeah. So Dan, as you think about these Alibaba-type platforms, how much data do they have on me? Is it a, is it a lot? It seems a little scary at times. And then what's the role of privacy going forward?
O’Connor: So whether you're in Bangkok or Beijing or in Boston, permissioned data is increasingly a cornerstone in the way that commerce is done. And we've moved from the days of where a company felt they had the right to use your data in any way, to a world where permissioning is super important. What's important about the platforms is that they've actually figured out a way to have a common consumer ID that travels with you across all of these applications within that ecosystem, and this enables what's called unified data, permissioned unified data. What this allows me to do is if I have a shopper base or user base of five or 600 million people, it allows me to maintain ten thousand eleven thousand traits on those consumers. So I have deep insights not into just what they buy and not even just into core demographics, but I get a great sense of behavioral characteristics.
Kerr: How does this permission unified data translate into the United States?
O’Connor: Many retailers in the US are rewriting, re-publishing their loyalty program agreement to give them higher levels of permission to engage and use your data. They're able to get you to a new product in a much shorter period of time. And they do it in a way that disintermediates agencies in using focus groups, in other words, you're digitizing and speeding up. So you would logically ask, because we're most concerned here about the future of work, what this means to large scale enterprises that have brand people and marketing people and so on. And what we're seeing in the, in this example is tremendous, tremendous changes in the organizational structures in terms of moving away from matrix based, you know, a checks with B that checks with C and D, and you know everybody has to get their crosschecks, which has power and value. But in a speed-based environment you can't wait for two weeks. The business question again for large CEOs, big company CEOs is if that's the future, where does my capital need to be? What's my talent? What are the new cultural norms? What's the vocabulary? How am I going to talk to my people about this? And really what's my role here? Because one thing is very clear is that the value that we're creating is changing as we go from one generation to the next, as large scale brands and as retailers.
Kerr: Dan, it's been a head spinning conversation. But we appreciate you sharing with us how the retail commerce sector is changing from G1 to G2 to G3 and increasingly in to G4. Dan O'Connor is the CEO of FrontFour Ventures and also a visiting executive with the managing future of work program at HBS. Thanks Dan.
O’Connor: Thank you.
Kerr: Thank you for listening to this episode of the Managing the Future of Work podcast. To find out more about our project on the future of work, visit our website at hbs.edu/managing-the-future-of-work.