How We Sell

If shopping is the great American indoor sport, why isn’t it more fun? Why isn’t it easier? Why isn’t it better? Paco Underhill, founder of Envirosell Inc., decodes the secrets of retail design to explain the rules behind how we shop.

New York City’s Chelsea neighborhood, wedged between Greenwich Village and midtown, is a square-mile shrine to cusp-of-the-millennium mainstream retailing. Barnes & Noble, Victoria’s Secret, and Restoration Hardware (see inset photographs) have all established big, boisterously appointed outposts amid turn-of-the-century office buildings. The Body Shop, J. Crew, Banana Republic, and the Gap have also arrived. abc Carpet & Home, a rug shop-turned-home-furnishings megaboutique, has become a consumer destination unto itself.


Two floors above Chelsea’s mercantilist fray, tucked into a former hotel, you will find the headquarters of Envirosell Inc., home to managing director Paco Underhill and 25 of his colleagues, who are perhaps the world’s savviest students of the art and science of America’s favorite indoor sport — shopping. Its offices are, to be generous, utilitarian: white walls, plain windows, and a hodgepodge of oak-laminate cubicles situated beneath bulky air-conditioning ducts. From the looks of things, this is a company of the retailing business, but clearly not in the retailing business.

The single distinguishing feature of Envirosell’s two floors, in fact, is the collection of videotapes. There are thousands of them: Some rest on shelves that rise from floor to ceiling, some are piled into plastic tubs, and others are simply stacked on top of one another. Each is carefully labeled: “Chicago/White Hen, Gatorade/July-94”; “Costa Mesa/Pepsi Drive-Thru/11-6-98”; “Deerfield/Blockbuster/ 12-9-95” — city/store/date, videotape after videotape.

“This,” says Paco Underhill slyly, pointing to the massive collection of tapes, “is my retirement annuity.” He’s not kidding. Outside of Hollywood, this could well be the most valuable film library in the nation. The tapes are meticulous diaries of retailing sociology and, if you will, markers of cultural history. They tell us a bit about who we are, and they show us a lot about how we shop.


Underhill, 47, founded Envirosell in 1978, a few years after hearing urban anthropologist William H. Whyte Jr. speak on the subject of the mechanics of city planning. It was an epiphany for Underhill, who soon realized he could apply Whyte’s tools to the retail world. Equipped with a movie camera and a notebook, he could record how many steps shoppers took past a store entrance before pausing, which direction they turned, and where their eyes came to rest. He could track how often parents reached for a certain brand of cereal, with which hand, where the box was located on the shelves at the store, and whether their children’s protests changed the parents’ decision.

This is what Envirosell does — in malls, in restaurants, and at kiosks around the world. Underhill’s researchers patrol stores at the behest of Samsonite, Hewlett-Packard, Denny’s, Toro — and dozens more merchants, retailers, ad agencies, and financial-services institutions, not to mention the U.S. Postal Service. For each assignment, Envirosell researchers position video cameras at crucial points — at the store’s entrance, say, or near the cash registers — and then they discreetly cruise the aisles, mapping the routes of shoppers and taking notes on their behavior.

After returning to the offices in Chelsea, the researchers spend hours poring over the videos in fast-forward mode, studying shoppers as they zip past displays and through checkout lines. Every so often, they spot something: an abrupt turn past a table stocked with products, perhaps; an overly long pause in an aisle; or an awkward reach for an item that’s been stocked on the bottom shelf — some small gesture. At this point, they slow the tape, watch it again, and jot a note. They have captured an insight.


During more than 20 years and thousands of tapes, Underhill and his crew have turned up some interesting insights: First, shoppers almost always drift to the right; they walk to the right and look toward the right, which is why merchandisers position new products just to the right of known top sellers. Second, people don’t read more than three or four words of a sign in a shop window. Third, shoppers walk past banks quickly. Fourth, mirrors remind shoppers that they are being watched. And fifth, if you put chairs in a women’s clothing store, men will sit and women will shop longer.

These are not opinions; they are facts. And if you are a retailer or a consumer-products marketer, they are incredibly valuable facts. But they are also significant observations about business, work, and the act of finding — or being — a customer. That is why Underhill’s new book, “Why We Buy: The Science of Shopping” (Simon & Schuster, 1999), entered its ninth printing in August. His book is a remarkable business tool, a distillation of all those notes and tapes, packaged in a way that is useful, witty, and loving.

For this, Underhill, a quiet, almost awkward man, has become a minor phenomenon. In the months since his book came out, he’s been featured on nearly 20 National Public Radio segments. He has been interviewed by country-music stations, the Christian Broadcast Network, and a flagship polka station — which, as he notes, “doesn’t usually do books.” Pacomania reigns.


And, at the same time, Envirosell is booming with new clients, new assignments, and new opportunities to add to the stock of tapes that feed Underhill’s ability to decode the continuing evolution of shopping behavior. Despite his newfound fame as a successful author and shopping guru, Underhill still makes it a point to spend time “in the field,” keeping his eye engaged in anthropological observations. On any given day, he and his researchers are more than likely holed up in a store somewhere, shooting tape and looking for new insights. Fast Company caught up with him in one of his rare moments back at his office.

How has your book and your work contributed to people’s understanding of retailing?

The point is that we either love or hate shopping. This book holds up a mirror to the shopping experience, so that people can read it and see themselves. And that’s what brings interest. What we do sure ain’t rocket science. I mean, it’s not as if we invented some magic black box. On the other hand, there’s a very distinct logic that runs through much of what we do. We can tell you, “If you do this or that in a store, here’s what the consequences will be.”


Why are those consequences particularly important now?

There are a couple of critical issues here. One is a basic realization in the business world, whether we’re talking about retail goods or consumer products: The marketing engine that we have built with some care during the 20th century still works, but it doesn’t work nearly as well as it needs to. As more people realize the degree to which it does not work, they are reaching out to different sources for information or inspiration.

Increasingly, too, in our time-pressured culture, retailers are recognizing the importance of purchase decisions that are made or that are heavily influenced at the point of sale. More and more, people today tend to park their brand predisposition at the door. As we approach the 21st century, the notion of somebody being “a Chevy man,” in the way that there were loyal Chevy men in 1957, is almost passé.


So we can measure brand awareness, brand image, and purchase predisposition. But even with all of those measurements, a lot of people walk in the door of a CompUSA thinking about buying one brand, and they walk out having bought another. I go into an auto showroom to look at a Ford Mustang, and if I don’t like the salesperson, I get frustrated. So I leave, and then I head to the Nissan dealer. These days, we’re all quicker to bail.

And I think that we consumers are eminently more cynical than we used to be. We’ve been flimflammed, manipulated, and marketed to, and we’re starting to see through the facade. We are more experienced. Our tempers are shorter, and our patience is thinner. As consumers, we are becoming aware of our power — the power to pick up our toys and go home, or to go somewhere else to play.

Are you saying that the old tricks simply don’t work anymore?


They still work. No matter how much smarter or more experienced we have become, there’s still something called “retail magic,” where someone is able to put something out there in a way that makes us fall in love with it and have to have it. What is different now is that the same old tricks don’t work with everybody. The older we get, the more experienced we become. And while people from generation X and generation Y are definitely manipulable, they are still more streetwise than you and I were when we were their age.

Also, while we live in a world in which there are global brands and identities that almost everyone recognizes, we also see some measure of heterogeneous complexity. There are a lot of subpopulations to sell to, more than there used to be. On the same sidewalk, you’ll see some people who wear brand names on the outside of their clothing, others who think the label is best kept inside, and others who don’t care. So stores need to find subtle ways to broaden their reach. One way to do that: Stores can adjust their music. Old Navy, for example, does a great job of playing contemporary music, “oldies,” and some jazz, and through its music selection, the company makes a cross section of people feel welcome.

Which shoppers are most ignored by merchandisers?


There are a few ignored groups. First, for all the attention we give seniors, that shopping segment has not been catered to in the way they could be. The aarp may negotiate discounts, but it’s not negotiating focus. Marketers are just starting to make progress in senior housing and senior-friendly furniture and athletic apparel. I still have a tough time getting a generation-X marketer to believe that going to Naples, Florida to study a large senior market may be a great head start on learning what the marketing problems are going to be in 2010. So the first group I would focus on is seniors.

Women are probably a close second. Women continue to be neglected in traditional male arenas, from cars to technology to hardware — this despite the amount of purchasing power that they wield. Ethnic groups are third on the list. Reaching out to an ethnic market is often cheap and effective because the ethnic market has fewer choices. We did an interesting case study here in New York City, where we were asked to look at the dynamics of a newsstand located at Greeley Square. The client was surprised that the newsstand sat at one of the major corners of New York’s Koreatown, and yet it stocked no Korean-friendly products, except for weekly papers. There were no Korean magazines, no Korean soft drinks, and no Korean beer. No one had thought about selling green tea.

The issue here: Often, marketers have no real sense of the ground. Where is my store? If there’s an ethnic community adjacent to it, maybe it makes sense to cater to them! Another way to think about the “ground issue” is to recognize who’s coming onto your territory. This country has become a huge tourist destination, and yet there’s little that we do as marketers to cater to that offshore purchaser. I challenge you to go look for charts converting American sizes into European sizes at the Gap or at Disney stores. If you make even a minor gesture like that, you’ll immediately reap rewards.


Restoration Hardware, for example, is a wonderful store. At the same time, it frustrates me. There’s no thought given to the international shopper: Once it had a beautiful brass frog that I wanted for my garden. (I’m one of the rare New Yorkers who has a garden.) And I wanted to buy one for my business partner in Milan. When I asked about a European hose coupling, the store manager had no idea what I was talking about. Yet, during the Christmas season, a third of the people walking into that store come from someplace where American-size fixtures aren’t used very often. Having couplings available for European hoses might have meant that a lot more of those brass frogs got sold. The store is controlled by its buying office, in Marin County, California, and I don’t think they have a good sense of what’s going on in urban settings.

Is there a difference between the way men and the way women shop for their families?

First, when it comes to shopping for the family, most wives are better trained at it than their husbands are. Wives have a better sense of discipline, and shopping is more of a chore for them. Wives approach shopping as a money manager, as opposed to husbands, who approach it as a moneymaker. It’s easier for husbands to slip into drunken-sailor mode: “That’s a good deal. It’s a curious product. I want it. I’ll figure out how to use it when I get it home. Salsa-flavored ketchup, just what I always wanted!” In general, kids have an easier time manipulating Dad. Mom is better at saying, “No.”


Do you see the gap narrowing between how men and women behave in stores? Are we more alike than we used to be?

I think that the whole question of gender and sexuality is an issue that we’re eminently more comfortable with in 1999 than we have been in the past. But it’s a very curious issue — whether it’s looking at the men’s fashion industry and at the relationship between straight and gay men, or that Gapbody has been launched, which, as I understand, is a store that sells underwear for men and women under one roof. On the other hand, some successful merchants, like Banana Republic, have separated their stores: a Banana Republic for men and a Banana Republic for women. Certainly, almost all men in those unisex stores like the Gap or Old Navy have had the experience of looking at something and realizing, “Uh-oh, this is not meant for me.”

There is that terrifying prospect of being caught on the wrong side of the store.


In the women’s-apparel business, it’s interesting to see the number of clearly not-petite women shopping in the petite section. Who can explain that? Gender differences are going to be around for the foreseeable future. Testosterone is here for the foreseeable future. And there are still going to be significant numbers of men who don’t buy their own underwear and who instead focus on buying beer, gas, and guns. But many of us are set in the midst of a much more profound gender revolution.

Despite all of the changing attributes of shoppers, are there any new developments that have actually made it easier to operate in the current retail environment?

For most retail stores, there’s an absolute predictability about who is in the store and when they will be there. You can tell who’s coming in and what opportunities they present. If I’m running a bookstore in a mall, I know that senior citizens come through early in the morning; stay-at-home moms shop during the day; kids come by after school; and professionals shop after work. That information ought to make it a lot easier to design retail spaces that appeal to each group at each time. For example, why not devise a rotating display facing the mall corridor? You turn it one way from 10 am to noon, rotate it from noon to 2 pm, then turn it again from 2 pm to 6 pm. You simply put different books on the display sides depending on who will be walking past.


There are a number of things that you can do — small ideas that are fundamentally clever, that cut through the clutter, and that work. Instead of using a lot of small, clever ideas, everyone’s looking for that one perfect moonstone. Here’s another example of a small work of art: There’s an elevator in a hotel in New York’s financial district with a mirror and a sign underneath it that says something like, “You look famished.” Then it lists all of the restaurant options that can be found in the hotel. It’s brilliant!

There’s an mtv logic here, involving the use of icons and words to get across a complex thought. It’s a technique that applies the same principles to marketing that Reginald Marsh applied to the world of graphics with his giant-size murals — that somehow there was a way in the course of one frame to get across the image of something infinitely larger. The idea of a mirror and “You look famished” uses the power of a reflective surface. A mirror has always been a magical symbol. Someone has just twisted it wonderfully.

A lot of what you’ve said seems like common sense. Why don’t merchandisers get it?

In part, it’s because the merchandising business is in a cowboy stage. It has very little accountability built in. The point-of-purchase industry is now maybe four times larger than it was 15 years ago, and, even at that, it’s desperately looking for its place at the table of marketing. I had lunch today with a senior executive at McDonald’s, and I asked her, “How much do you spend testing commercials? And how much do you test your merchandising and promotions?” Although McDonald’s does test some point-of-purchase merchandising, it spends far more on checking commercials.

The dollars spent on merchandising crept up without anyone keeping close track. And lots of people talk strategy, but not many of them know about tactics. We’ve made huge strides in the past 10 years, and yet, until recently, there wasn’t a major business school in the country that specifically offered a course in merchandising.

Why hasn’t technology solved all of these problems?

Technology is a troubling issue, because as much as it facilitates, it also confuses. Over the past 10 years, a lot of people have knocked on our door with technology, wanting to back it into some application. Consequently, we’ve spent a lot of time working with technology. We found that almost all of the interactive devices that we’ve studied over the past decade have failed. In some cases, we have tried to apply technological solutions to human problems. And almost universally, the designers of those solutions have had an imperfect understanding of the environments that they were studying.

Time after time, we’ve seen merchandising approaches that involve putting computers on the retail floor — and no one has considered that somebody on the floor has to reboot the computer on a regular basis. And it’s amazing that, when ideas for introducing technology into the retail space come up, no one seems to realize that eventually any flat surface in a commercial space is going to get something spilled on it. We sell an awful lot of soft drinks in this country. Sugar water isn’t good for computers — and it finds its way into almost every store.

You make it sound as if the retail environment suffers from a number of disconnects: the disconnect between buyers at headquarters and salespeople on the shop floor, and (the other way around) the disconnect between the brand manager who designs a marketing campaign and the shop-floor clerk who sets it up in the store.

We ran a study a number of years ago that looked at hair-care products for a major American health- and beauty-aides maker. We went to different stores and took pictures of shelf sets — what products were located where. We enlarged about 15 of those pictures for different stores, and we took those enlargements into our client presentation. Everybody went nuts over the photos! “That’s not the right shelf set! We didn’t pay for that! We’re supposed to get 16 facings, and we have only 12! How did that brand get on that shelf? It shouldn’t be there!”

It was amazing to think about this disconnect between our clients, sitting out in their exurban campus in New Jersey, and the stores themselves, where the objects of their attention were sold. I felt as if I was dealing with a puppet master who was 10 miles away from the stage where the puppets were performing. This is not to say that all of that planning is bogus, or that the execution isn’t generally pretty good, but it’s by no means perfect. As we go into the 21st century, stores are struggling to control labor costs. As that happens, and as fewer people work the floor, more stores are going to get the merchandising execution wrong. In the grocery-store trade, it’s been estimated that 25% of what’s shipped never makes it out to the floor. A quarter of the marketing vehicles that are sent off to your local Ralph’s Supermarket get in the back door and make a one-way trip to the Dumpster. That’s a frightening number.

What lessons from your store research can you translate to commerce on the Internet?

I’ve been asked that question innumerable times, mostly from e-commerce companies. The Net is desperately looking for answers. In one sense, we can’t translate much from the physical retail world to the Web. But in another sense, there’s a framework in stores that could be powerful in e-commerce. Stores have a certain predictability: I enter using a doorway, after which I follow a prescribed pathway. Merchandise is presented in a certain way that is more or less consistent from store to store. I get assistance in a certain manner. Finally, goods are exchanged for units of value using an established procedure. Everyone understands those rules; it’s a basic system that we’ve all grown up with, and it works at Calvin Klein as well as at the local Piggly Wiggly.

The Net, on the other hand, has no prescribed doorways. People can drop in as well as drop out. There is no established system for telling me, “This is the product.” And most important, where the rubber meets the road in e-commerce is that there’s no universally understood way of getting to the transaction. I was talking to the research director for one of the most progressive retailers in the world. He was tearing out his hair with a Web designer trying to figure out what their online shopping cart was going to look like.

Then there are some very basic issues: There are visual acuity problems and basic communication issues. Lots of stuff is designed on a 21-inch screen but is seen on a 15-inch screen. Above all, there is a basic kindness shown toward people in most stores that hasn’t as yet been translated to the Web.

Even considering all that, of course, e-commerce has real power.

E-commerce has succeeded where the traditional connection between manufacturer and retailer has failed. Books, music, movies, porn, and stocks are five categories in which the relationships between manufacturer and distributor are fundamentally flawed. I say books, because the bookstore is still trying to sell them as consumables, and the market is buying them as durables. In the world of music, manufacturers are eminently closer to the artists than they are to the consuming public. And in 1999 music is still packaged the same way that it was in 1959 — only the lp format has been shrunk down to the cd. To me, that’s a disconnect.

The real opportunity online is with people who are ill-served today. Why isn’t somebody figuring out where the disconnects are between present-day retail and present-day consumers? Is there a search engine that finds products for large-size or tall people? There are enough of them out there. The future of the Net is going to be great, but in its present form, it’s trying to capture everything rather than focusing on exactly where it needs to go.

So what does that say about what the Web should be doing?

The Web should focus on creating something better, rather than being smart about recognizing distinct niches. Or it can play an important role by forming a bond between the cyberworld and the physical world so that they support each other. I love that you can go into a Borders and order a book from its warehouse via an Internet kiosk and have the book delivered to your house. And if it’s not the book you want, you can return it to the store. But the Victoria’s Secret Catalogue, the Victoria’s Secret store, and are three separate businesses. That’s really dumb. When you see that kind of division, you know that the Web opportunity is simply being missed.

So your forecast is, We’ll always have stores.

Yes, we’ll always have stores. Part of what will come out of the Web, though, is an interesting muddying of what is physical space and what is cyberspace. What if I can walk to the market here in Union Square, go shopping for vegetables, and then drop off my order at a cybercafé? Or do the same for laundry soap or toilet paper? That lets stores shrink their footprints dramatically — to get out of unprofitable commodity categories, to present the public with a relatively small front end for a much bigger warehouse and distribution operation. Part of what we’re looking at, then, is the Net’s future as an integrated part of our brick-and-mortar existence. Just as I can use my store to drive traffic to the Net, I can use the Net to drive traffic to my stores.

Everyone is out there searching for magical technology solutions, and I’m sure there will be better ones than those we have now. Technology may have answers. But let’s start with the basics and see whether we can get them right! I’m still amazed by how far businesses get down the e-commerce road without understanding their consumers. What will always be true is that having a good pair of eyes and spending a little time out on the floor are valuable. As much as technology may be able to capture transactions and a mechanical system may be able to generate numbers, it’s hard for such technologies to capture the rhythm of real life.

When you go shopping, what grabs you? How do merchandisers reach you?

To some extent, my approach to stores has been spoiled. Everywhere I go, I’m engaged in deconstructing the store. I love doing it — don’t get me wrong. But it often makes me angry. When I see something done wrong, and when the solution is so obvious, I want to grab the manager and shout, “You’ll make more money if you just turn it around! Why did you do it this way? It doesn’t make any sense! Don’t you know that about 85% of the world is right-handed? Put that on the other wall!”

Keith H. Hammonds ( is a senior editor at Fast Company. You can reach Paco Underhill on the Web ( or by email (

Sidebar: The Paco Principles

In a short period of time, Paco Underhill’s “Why We Buy: The Science of Shopping” (Simon & Schuster, 1999) is fast becoming a business best-seller. Here, adapted from the book, are some of Underhill’s key lessons on shops, shoppers, and shopping.

Bring ’em in; then keep ’em there.

The amount of time shoppers spend in a store is perhaps the single-most-important factor in determining how much they will buy. The majority of advice we give to retailers involves ways of getting shoppers to shop longer.

Honor “the transition zone.”

On entering a store, people need to slow down and sort out the stimuli. Which means that whatever is in the zone they cross before making that transition is pretty much lost on them. If there’s a display of merchandise, they’re not going to take it in. If there’s a sign, they’ll probably be moving too fast to absorb its message. If the sales staff hits them with a hearty “Can I help you?” the answer is most likely going to be, “No, thanks.”

The hand bone’s connected to the wallet. The fact that most shoppers have two hands is well known. But the implications of that are often ignored. It’s hard to overemphasize the importance of the hand issue to the world of shopping. A store can offer the finest, cheapest, sexiest goods, but if the shopper can’t pick them up, it’s all for nothing.

Mirrors versus banks. People slow down when they see reflective surfaces. And they speed up when they see banks. Bank windows are boring and so are banks. Mirrors, on the other hand, are never dull. So never open a store next to a financial institution — by the time pedestrians reach you, they’ll be moving too fast for window shopping. If you can’t help being next to a bank, make sure to have a mirror or two on your facade to slow shoppers down.

Take men. Please. Men always move faster than women do through a store’s aisles. Men spend less time looking, too. In many settings, it’s hard to get them to look at anything they hadn’t intended on buying. Men also don’t like asking where things are — or any questions at all, for that matter. If a man can’t find the section he’s looking for, he’ll wheel about once or twice, then leave the store without ever asking for help.

Behold the geriatrics. By 2025, nearly one-fifth of the American people will be 65 or older. All of retailing is going to have to cater to seniors because they’ll have the numbers and the dollars. But we’re going to need a whole new world. For starters, the words on packages will have to be large enough so people can read them.