BUSINESSWEEK ONLINE : FEBRUARY 21, 2000 ISSUE
INFORMATION TECHNOLOGY Jeff Bezos: There's No "Shift in the Model" In a Q&A, Amazon's CEO and Jeff Wilke, operations chief, talk about how the e-tailer will attain profitability Will Amazon.com ever make money? That has been the overriding question from investors almost since the online superstore started in 1995. It has been of particular concern lately as Amazon has expanded both the number of products it offers -- from books to CDs, toys, and tools -- and as it has spent $300 million to build six huge distribution centers in the U.S. In its just-reported fourth quarter, Amazon finally showed some signs that it will ultimately turn a profit -- including news that its original book business went into the black. In recent weeks, it has also signed deals to give other e-tailers exposure on its site in return for hundreds of millions of dollars in payments over the next few years. CEO Jeffrey P. Bezos, accompanied by former Allied-Signal executive Jeff Wilke, now Amazon's vice-president for operations, recently talked to Business Week Senior Correspondent Robert D. Hof about the various ways Amazon aims to make the entire business profitable. Q: Why are you focused so much more on profits these days? Is it because more investors seem to be demanding to know when Amazon will turn profitable? Bezos: I'm not sure if there really is a new emphasis. Maybe people have become sensitized to it. Some of our businesses are getting more mature, and they reach inflection points when they actually do need to be managed in different ways. With U.S. books' profitability in the fourth quarter and expected profitability in 2000, there are some changes you would expect to happen. One of them is a stronger focus on operational excellence -- which means treating customers right but at lower cost. Q: What constitutes an inflection point? Bezos: It's the size of the business -- number of transactions or sales dollars. If you have a billion dollars in revenue, a hundred basis points in operational efficiency is a significant amount of money -- it's $10 million. Wilke: On the operations side, it's also related to cycles of learning. When you get good enough at understanding a particular set of customers and a particular set of products, then you can start shifting what you're focusing your attention on. You can really home in on efficiency. Bezos: And there's a different inflection point which has to do with the number of customers. We've also reached a tipping point in the number of customers. When you have earned a relationship with a large number of customers, you get to do things on their behalf -- always on their behalf, or it doesn't work -- that you can also monetize. That's what leads to these deals like Drugstore.com, Living.com, Ashford.com, NextCard, and so on. Q: Do these deals indicate a shift in Amazon's model from being a retailer to being a landlord? Bezos: Our No. 1 goal has always been to be Earth's most customer-centric company. Within that, we're building this universal selection -- a place where people can come to find anything and everything. There are starting to be high-quality companies out there who are already doing a great job serving certain segments, like Drugstore.com and Living.com. If we can incorporate those in a seamless and integrated way into our Web site, we've accomplished the goal of both being customer-centric and having universal selection. It's the third leg in the toolkit of parts that we have to help our customers. There are things like zShops [small merchants selling on Amazon's site]. There are our retail stores. And there are these partnerships. That is how we will go about assembling Earth's most customer-centric company. Q: Why are these partnerships coming so quickly now? Bezos: You saw this with America Online, too. The first deal that they did when they reached their tipping point was the Tel-Save deal. What happens is, when you reach a certain critical mass of customers, and you have relationships with them that you've earned over time because you've done a good job for them, you very quickly have a long line of people who want to associate with you. That doesn't happen when you have a million customers. I don't know why, but it seems to happen, as it did for AOL, somewhere around 10 million or 15 million customers. Q: When did people start calling you? Bezos: They've always called, but it's also whether they will do a deal that is actually good for customers. Take our NextCard deal. We offer our customers an Amazon.com-branded credit card. We could have done that deal at any time. But what happens is, when you reach that tipping point, you can negotiate better terms for the customer. Think of it as cooperative buying. When we have 17 million customers, obviously we can cut a much better deal than when we have a million customers. All of a sudden, we can get really great deals for our customers -- and at the same time have it make sense from a business point of view for the company. It's something that we knew, if we were successful, would happen. It wasn't like all of a sudden, one day, we said, "My goodness, look what's happened!" Q: Don't these deals represent a shift in Amazon's business model? Bezos: It's not a shift in the model. It's something we had always thought about. For at least a year, we've been talking about ourselves as a "platform." It's a foundation or a workbench from which you can do a lot of things. In our case, it consists of customers, technology, e-commerce expertise, distribution centers, and brand. Let's say we decide, as we did recently, to launch into the tools business. If we had no brand and no customers and no distribution centers, and no customer service infrastructure, and no technology, and no e-commerce expertise -- well, gosh, it would be hard! It would be downright daunting. But given that foundation, then it's actually pretty easy to enter into a new business. Q: What other sorts of businesses might you be able to launch off that foundation? I've heard you mention membership clubs. Bezos: I've mentioned it many, many times, but always the same way -- which is a completely hypothetical response to a completely hypothetical question. We have no intention or plans of doing any kind of membership clubs. Q: Never? Bezos: It's a good idea for a certain contingency that will evolve over several years. People ask about perfect price information online. The analogy that often gets drawn is the electronics boutiques of the 1970s. What happened is they had highly trained salespeople, and they charged high prices. And people would go into electronics boutiques, use the salespeople to make purchase decisions, then go into the big-box electronics retailers to make the purchase and save money. People point out that we could conceivably face the same sort of problem at some point in the distant future. The problem would be that we've put a lot of energy and attention into customer discovery -- helping people make purchase decisions. So wouldn't it be possible for somebody to use our technology to make purchase decisions but then go purchase elsewhere? But the basic flaw with that reasoning is that it misunderstands scale economies online (which says that online merchants will have the lowest prices because their costs are lower than those of physical retailers). Then you can say, even if a particular company, like Amazon.com, doesn't achieve the scale to have the best discovery features AND the lowest price, which is possible, then you can still have membership plans. So you can see how hypothetical this is. Q: Do you think the value of your customer base -- which is so much more important online where there's no geographic proximity to keep people coming back -- has been recognized? Investors don't seem to focus on that, given the lower stock price lately. Bezos: I never worry about issues like that. Assets that aren't recognized on the books are quite common -- things like brand names -- and they pay dividends in other ways. Brand name becomes financially visible in reduced customer-acquisition costs. And ours, as far as we can tell, are the lowest in the industry. So I never feel like we don't get enough credit for that. But customers are more important to us than they are to most companies, or any other company. Q: How do you decide which products you will sell yourself and which you'll offer through partners? Bezos: We have a lot of criteria, but No. 1 is how do we do the best job for the customer. Q: I imagine size of product makes a difference, but what other criteria are important in that decision? Bezos: If it was an airplane, size might matter. But we sell 60-inch televisions. We ship 600-pound table saws. Everything you do has to reinforce the customer experience. What you can't do is simply harvest those customer relationships. You have to make sure that everything you do reinforces the strength of the relationships. Take Drugstore.com as an example. That is a very complicated business, because you have to be regulated in all 50 states in a very careful way. You have two payers because you pay the $5 copay, and the insurance company takes care of the rest. That leads to a different set of technology systems to make that work. So it becomes clear very quickly that because they're up and running and they have that customer experience nailed, it would be much better for our customers to offer them that experience than to put our energy and time into trying to replicate it. Wilke: We don't view it as our birthright to handle everything that the company chooses to offer. If we have some [new product] that fits into what we've got and we can move it efficiently and we can support the total experience, then we ought to do it internally because we're probably going to do it at a lower cost and offer a more complete package. But if we can't do that, and we'll be in the way of doing it fast, then we should step out of the way. Q: You have a long way to go before you actually offer "anything and everything you might want to buy online." How soon are we likely to see that? Bezos: I wouldn't want to speculate on that. To really get to the vision, which is truly anything and everything that you might want to buy online, that is a lifetime mission. That is a multidecade proposition. Wilke: Thank god. Bezos: We'll make very rapid progress. We'll do more this year than we did last year, for example. Q: You mean more of your own online stores on the site, as opposed to stores of partners? Bezos: That is true. Q: How will you deal with potential conflicts? Won't some partners' products overlap? Bezos: I sort of doubt it. I'm not too worried about that. Q: In focusing on profits, you've made several moves to improve the efficiency of operations -- including a recent layoff of 150 people. What did the layoff accomplish? Bezos: We have to be constantly evaluating the fit of our mission and our people. Q: So, some people just didn't fit? Bezos: There were positions that needed to be eliminated. Q: Positions, or people whose skills didn't match up? Bezos: It's positions. On a personal level for everybody here, that's distressing. But it's also part of trying to make the best decisions that we can. Q: Are there more layoffs to come, given that you're at 7,500 people, in a fast-changing industry? Bezos: I wouldn't want to speculate on that. We're constantly evaluating everything of that kind. That was 2% of 7,500 people. If that were done on a daily basis, it would have been like two people a day throughout the quarter -- which for a company of 7,500 people is not unusual. Q: Folks inside Amazon say some departments such as customer service are not very well-run. Bezos: What?! That's ludicrous. I have no idea where that comes from. In fact, that's the thing we are most widely recognized for -- excellence in customer service. Q: Still, you said recently that there's a lot of room for improving operational efficiency? What processes can you improve this year? Wilke: We have a large distribution center network that we built mostly last year, with a huge amount of information technology and nice automation in the right places. Now, we put people into it for the first time, and...people who are just learning how this thing works during the holidays are ramping up to a level that the company -- in fact, no one -- has ever seen. That creates an enormous opportunity to learn. What we did was get through the holidays with great performance in terms of the customers' experience, and we did it by using the best of the folks that we could, but they hadn't been on the job very long. What we'll do this year is focus less on making sure that we have enough people out there, and focus more on making sure that the people out there know what they're doing, have the right skill sets, the right tools and processes, and that we're as efficient as we can be in everything we do. We want to do everything we can to make sure that our processes are repeatable, and that there's no variation. Bezos: 1999 was a year when we added 3 million square feet of distribution center space, up from 300,000 square feet. Getting those distribution centers up and mechanized and running at all is hard, and it was on a very compressed timeframe. But in this year, we don't think we're going to have to open any new distribution centers in the U.S. So this can be a year of honing. Q: What things can you hone in the distribution center? Wilke: If we can shrink the time it takes us to get a toy, a book, a DVD, any product, in half, we lower the inventory we have to carry by a huge amount. Most of what you carry is to cover the fact that it takes you a couple of weeks in some cases to get product. The other thing that happens is, if I can get the lead time, from the time I place an order to a vendor to the time I ship the product out the door, to one day -- and I offer in the distribution centers 24-hour availability -- what is my inventory? The answer is zero. So the more we reduce inventory, the more we become a cross-docker, a matcher of items that you've ordered, and less a warehouse to carry stuff that we ordered weeks ago. Your order spends several hours in our process, and in that time, there are several people that have the opportunity to touch and influence that order. And we check it throughout the process -- picking, sorting, packing, and shipping. If we have to check all along the way, and make mistakes, and catch those mistakes by our checks, we're introducing inefficiency into the process every time we do that. So if we can get picking to be 100% perfect, we'll never have to go back later in the process and send that bucket of goods all the way to be repicked. That saves time, saves money, and makes our people more efficient. Same thing in the sorting routine: If I can make sortation perfect, and every time that chute closes it's got all the items that you ordered, we don't have to check it down the road, or when we do check it, we won't find any errors, so we don't have that repeat flow of goods. When you think about just a few percent of the yield having to go back into the shelves and be picked again, the benefit in productivity is enormous. On the shipping side, we have a really tight information link between the labels and your orders, so there aren't really any errors that are created there. Q: You've said you can make rapid progress on this in 2000. How rapid? Wilke: The processes are mappable. It's the classic "measure, analyze, improve, control." In the world I came from, I had to worry about chemistry, physics, and electronic interaction. All we have to worry about here is the flow of stuff. We understand what to measure. It's complex, but it doesn't require the same amount of mathematical rigor. We can reduce the variability with some basic tools, and we can link that all together into a process that everyone understands, and then you lock in the gains by controlling what you've created. We should start to see gains within about six months. So way before next holiday season, we should have a process that's much more controlled than what we have now. So the customer will see the same level of performance we've always had. We'll just do it more cost-effectively. Q: What key metrics will show you're succeeding or failing? Wilke: Our operational performance this year will be evaluated on the bottom line. If it doesn't show up on the bottom line, we're not working on the right stuff. Q: Do the losses that the companies in which you've taken equity positions concern you? Bezos: We take equity positions in these companies, which is clearly a completely discretionary thing to do, because we think that they're going to be successful. And we think we can help make them successful. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ BACK TO TOP					RELATED ITEMS Suddenly, Amazon's Books Look Better CHART: Profits on the Horizon? TABLE: Amazon's Plan to Stem Losses ONLINE ORIGINAL: Jeff Bezos: There's No "Shift in the Model" INTERACT E-Mail to Business Week Online

Copyright 2000-2009, by The McGraw-Hill Companies Inc. All rights reserved. Terms of Use   Privacy Notice