Amazon China on ‘transformative’ technologies
Doug Gurr reflects on how China differs from Western markets and what role data informatics will play in the future of retail.
Doug Gurr had been a global vice president at e-commerce giant Amazon for less than three years when he was asked to take on a new challenge: lead the company’s efforts in the fast-growing, hypercompetitive Chinese market. In September 2014, Gurr relocated to Beijing from his home in the United Kingdom. Today, he is in charge of an operation that employs approximately 5,500 people.
Gurr recently spoke with McKinsey’s James Naylor. Excerpts of the conversation follow.
McKinsey: During your entire retail career, you’ve championed investments in technology and innovation. Are you finding the technology in China to be different from what we see in more developed markets?
Doug Gurr: Yes. The technology in China is phenomenal. You can see multiple ways in which the country is leapfrogging. For example, there’s not much of an established physical retail infrastructure, so people are going directly to a purely online world. They don’t go to a physical store at all—they simply look online and then purchase. If you talk to a group of Chinese women between the ages of 20 and 25 and ask them where they shop, they’ll just look at you like you’re a bit stupid. “I’ve never been to a store. Sure, I buy fresh food at stores—but for anything other than that, why would I ever go to a store?” The world in which consumers get their insight and information exclusively online is very different. Social shopping, for instance, is an enormous phenomenon in Asia.
Already, 4 of the 15 largest Internet businesses by market cap are Chinese. The pace of innovation and the quality of the mobile experience in China in many ways far outstrips what you see in the West; it’s gone down a divergent path. Again, there’s a lack of physical infrastructure, so in areas like banking China is leapfrogging—going directly from cash to pure mobile e-banking.
Another example of leapfrogging is the use of geolocation. There’s very little mapping in China, and there are many areas with no street addresses, but China has solved these logistics problems with geolocation. You wouldn’t have thought you’d see bicycle rickshaws with better point-to-point geolocation and better GPS-enabled devices than you see anywhere else in the world. It’s amazing and exciting—there’s a blend of rough, old-fashioned ways of doing things coupled with technology that is way ahead in terms of the use of data informatics.
McKinsey: Do you think the rest of the retail world will eventually look like China? In other words, will physical retail become irrelevant everywhere? What do you think will happen to the retail value chain?
Doug Gurr: One of the megatrends of the next five to ten years will be e-commerce moving from a primarily national to an international business. That’s where a lot of the energy in China is going. It’s about refining the answer to a simple question: What is a retailer for? The job of a retailer is to connect a product anywhere in the world to a customer anywhere in the world—to provide that customer with the best information to aid discoverability and guide purchase decisions—and to do it with as little friction and as quickly and cheaply as possible.
No one is saying that physical retail will disappear. Physical space is a fantastic way of discovering products. But it’s also time consuming and expensive compared with a truly optimized, truly evolved digital-discovery experience. So I think there will be a role for physical stores, but I wouldn’t be too sanguine that such and such a format will necessarily survive forever.
McKinsey: Say more about e-commerce moving from a national to an international business. What are some things Amazon China is doing on that front?
Doug Gurr: Our primary focus is on cross-border e-commerce. We already do a pretty good job of helping Chinese businesses sell around the world; we want to do an equally good job of helping Western and Japanese businesses meet the growing demand from Chinese consumers for high-quality, authentic international products. We know the demand is there and that many international brands would love to sell in China but find it challenging to navigate the Chinese e-commerce landscape. To that end, we’ve launched a number of new services. One is the Amazon Global Store. We provide translation, listing, regulatory compliance, local-language customer support, local marketing, global logistics, and so on, so that brands can launch in China with no more effort than selling in their local market.
In just six months we’ve been able to launch over three million unique products for our brands, with no cost or effort on their part—literally with the click of a button. We’re rapidly expanding this and other services, and we’re starting to explore partnerships with many brands—US, European, and Japanese—that are interested in the complex but compelling opportunity presented by the Chinese consumer.
McKinsey: What role do you think robotics, automation, and technology in general will play in the future of the retail industry?
Doug Gurr: It’s transformative. You can take a view of retail organizations as decision-making machines. We have to make hundreds of millions, even billions, of decisions every day. How much do I price? Is this product safe? How much inventory should I hold at a particular location, at a particular store, at a particular moment of time? When should I replenish?
You read a lot about whether machines are better at making decisions than people are. I think it’s kind of irrelevant unless the machines are materially worse at it, which they’re not. The point of giving the decisions to machines is that you have scale. If you put human beings in the middle of every decision, you slow down. I used to work for a physical grocer, and every morning we’d argue about how much bananas were going to be that week. Today, it’s unimaginable for me to spend time setting prices.
Retailers need to understand a whole bunch of new tools and technologies. Of course, they’re not perfect; we’re at an early stage of these technologies. For example, we’ve all received those personalized marketing e-mails that aren’t quite right: “My dog died, so don’t send me e-mails promoting dog food.” Or, “I already bought a TV; please stop telling me to buy another one.” There’s a lot of clunkiness because we’ve only been in this game for a few years and we’re not yet very good at it. But personalization is so powerful that even at this early stage, if you compare a machine-based process with a pretty refined state-of-the-art manual process, technology wins every single time—and not by a slim margin.
I like to say that the only things people should do are things that only people can do—that is, making complex decisions that can’t be automated because they’re high uncertainty, they’re hard to reverse, there’s not enough data, or they’re judgment calls. That does two things for you. First, it allows decision making at genuine scale. Second, it makes jobs more interesting for people.
My personal view is that the transformative technology of the early 21st century will be data informatics, and I think it will happen much faster than most people assume. It’s a classic distinctive capability. I think in ten years’ time, in any business sector, the performance gap between an organization that invests in data informatics and one that doesn’t will be huge. I would argue that the performance gap is already substantial, and it’s only going to get bigger.