Why E-commerce Didn’t Die With the Fall of Webvan

The Internet grocer Webvan died a nasty death along with many other online delivery services—or did it? HBS professor John A. Deighton describes how the forces that propelled it are here to stay.
by Martha Lagace

Times were hard for Webvan this year. Like other online grocers and delivery services that hit the screen in 2001—among them, Homegrocer, Kozmo, and Streamline—Webvan finally called it quits in July after two years in business.

Webvan may be gone, but the forces that propelled it and other online delivery services should not be discounted too soon, according to HBS professor and marketing specialist John A. Deighton.

As Deighton explained in the article "Who Wanted Webvan to Survive?" published last summer in The Boston Globe, the biggest fans of Webvan were always upstream. Packaged good manufacturers were eager to create a direct-to-consumer link like the one Dell found to sell computers. Such a link would minimize wasteful and inefficient marketing practices, which to date have mostly focused on broadcast advertising.

"In the long run, the combination of efficiency, intimacy, and community offered by direct-to-consumer marketing will have an impact on every industry and intermediary," Deighton predicted.

Supermarkets may have "dodged a bullet" for now, but direct-to-consumer marketing is here to stay, Deighton says. In an interview with HBS Working Knowledge senior editor Martha Lagace, Deighton explains how it continues to carve out territory in the marketplace.

Lagace: Were you surprised that Webvan failed?

Deighton: No. The purpose of writing the case was to have classes explore whether there was merit to the business. Most times that the case was taught, the general consensus was that the economics of the business were very shaky.

Q: How so?

A: There were two parts to the economics. You could say there was the part of the business that ran in the digital world and there was the part that ran in the physical world. In the digital world, the economics were terrific. The business was highly scalable but, in addition, there was a great market for the digital information. So that part of the business looked to be very well-founded.

But in the physical world, it was a business of delivery: trucks and traffic jams and scheduling conflicts. And it was essentially being cultivated neighborhood by neighborhood. Unless you could get really high density, as the U.S. postal service does, and until you got very close to complete saturation, the business was just a headache.

Q: In your Boston Globe article, you wrote that buyers are taking to Staples.com and Merck-Medco. Why didn't they embrace Webvan, Kozmo.com, and grocery deliverers such as Homeruns to the same extent?

A: There's a simple answer. It's the idea that there's a direct migration path from direct mail businesses to Internet businesses, and that some of the most conspicuous successes in online commerce came from companies that already had a pretty good catalog or direct mail business.

Staples and Dell migrated successful catalog businesses onto the Web. Merck-Medco had done a terrific job of migrating patients with chronic ailments—patients taking prescription drugs for chronic conditions—from pharmacies to a combination of direct mail and telephone. The idea was that if you have a chronic condition, you're in a replenishment mode. You have to fill the prescription regularly over a number of years.

So it made a lot of sense for Merck-Medco to build a centralized fulfillment center instead of selling from a large number of decentralized pharmacies, and then encourage those people—essentially by cutting the size of their co-pay—to place their orders by phone, renew their prescriptions by phone, and then receive them a few days later by mail.

They induced about 12 percent of the pharmaceutical market to buy that way. Once that was done, it was comparatively straightforward to use the same inducements to persuade people who had Web access to do that over the Internet, saving the cost of telemarketing centers.

Q: With grocery delivery, the products are perishable and you need to wait at home for them to be delivered. In that light, do you think there's a future for online grocery shopping?

A: Home-delivered groceries? Never. Online communication, auto-replenishment services, direct-to-consumer promotion? Definitely.

The argument in the Globe article is that if a manufacturer can go direct, then the manufacturer has an instrument that combines advertising, distribution, market research and all of the marketing functions in one. That's terrifically attractive to the manufacturer because the great area of waste in any industry tends to be in its marketing.

In the period since the Internet boom, there have been a large number of very conspicuous successes.
—John A. Deighton

Physical distribution of groceries in the United States is a marvel of logistics. It's amazingly efficient and it allows supermarkets to run making a dollar of profit for every hundred dollars of revenue. The stage from manufacturer to the supermarket involves efficiencies of something like thirty dollars expended for every hundred dollars of revenue. It's very much less efficient.

When you bring things together in this direct model, you make the marketing investments directly accountable. You tell whether they're working or not. You can shift expenditures from unresponsive marketing methods to more responsive ones. And you improve efficiency, obviously.

Most packaged goods manufacturers have been looking for ways to build more direct channels to their consumers. Webvan was perhaps unrealistically imagined to be such a channel. The difference between Dell and Proctor & Gamble is that Dell can supply all your computing needs, but Proctor & Gamble can never aspire to supplying all your home provisioning needs.

And so the dream of direct-to-consumer would have to involve direct relationships with multiple manufacturers—or seem to; I don't know if it has to require that. And no one has found the model to do that.

Q: Do you think part of Webvan's demise was a problem of technology, that the technology was not sophisticated enough to handle all the kinds of transactions?

A: No, actually, that's the easy part of the problem. And that's proven by the fact that frequent shopper programs are able to deliver coupons and in some cases, samples, to regular supermarket customers.

Q: How about those coupons some supermarkets give you that are printed on the back of your receipt?

A: In those instances, the coupons that you're given are a function of what you've just bought, and in that sense they have no memory of what kind of customer you really are. So they can be [only] marginally effective.

I had a student who told me that every time he bought a Nestlé chocolate bar, he got a coupon for Hershey. And when he redeemed the coupon for Hershey, he got a half-price coupon for Hershey. When he redeemed that, he got nothing. So he said, 'I very quickly learned that my chocolate consumption should be Nestlé - Hershey - Hershey - Nestlé - Hershey- Hershey.'

The machine, of course, had no ability to learn what he was doing, whereas a frequent shopper program can do much more subtle things.

If you were a regular user of, say, a skin cream, and a competitor wanted to launch a competing skin cream, [it] could literally coupon heavy users of the competitor's cream versus light users. Of course, then the incumbent can do the same in return. At that point the supermarket becomes essentially an advertising medium.

The situation I'm describing is run by a company called Catalina. Catalina's popular with manufacturers, and that shows the basic need that manufacturers have to reach their customers directly. But it's not very sophisticated.

The one advantage that Catalina has is that it is national. It can reach the entire United States, every single skin cream user, whereas frequent shopper programs have to be aggregated market by market because we don't have any national supermarket chains. Except for Wal-Mart. That's a very interesting story in itself. Wal-Mart is, I think, the number five grocery retailer in the United States. The great advantage it has is that it's national and to a marketer that's very important.

The conclusion that follows is that perhaps the dream that Webvan was feeding—living off—will actually be realized by someone like Wal-Mart, managing only the information side of the business and not getting itself entangled in the physical or home delivery side of the business.

Q: Priceline recently announced a small profit. Do you think this is giving heart to other online businesses?

A: Priceline failed in the grocery business. But I'm not sure I would develop much out of that [the profit]. There are plenty of profitable online businesses. Merck-Medco and Staples are big online success stories. There are lots of unglamorous, incumbent businesses that have used the Internet to take an enormous amount of cost out of their system and run big parts of their business highly profitably. They just don't make the headlines. The only interesting thing about Priceline is that it is a dot.com startup.

The funny thing about the way the press covers the Internet and e-business—and maybe the way the stock market got enthusiastic about it—is that we had this period from 1996-97 through the first quarter of 2000 when the press couldn't get enough of the Internet. Everything was hyped and the stock market was excited. And, of course, then the subsequent failure of many of those business models was seen as some kind of divine retribution for being overly enthusiastic about something that had feet of clay.

And yet in the period since that boom, there have been a large number of very conspicuous successes. Dell is doing 18 million dollars of revenue a day on the Internet. There are these colossal shifts in spending to the Internet. And yet the public seems to have no interest in that sort of activity. The apparent explanation is that these successes are happening within incumbent firms and they lack the glamour and the smell of revolution that was associated with these garage-based startups.

To an economist or to someone from Mars or to someone studying the phenomenon, they should be just as interesting. But because they don't hold out the scent of gold rush profits for entrepreneurs, they just don't seem to get the front-page coverage.

I got a call from someone from Wired magazine just the other day on what's happening as people go back to school. The story is that people are going to take the same old courses they were taking ten years ago. It's going to be banking and consulting. And why? They would say, 'Because the Internet failed.'

But it didn't fail. It just went mainstream.

About the Author

Martha Lagace is senior editor of Working Knowledge.