Essays & Articles


Usability Is Not a Luxury

On the Internet, Usability is Not a Luxury�It is Fundamental to Customer Relationship ORIGINALLY PUBLISHED BY JAKOB NIELSEN AND DONALD NORMAN IN INFORMATION WEEK1 If the customer can’t find it, then the customer can’t buy it. This simple statement explains why usability is the lifeline of e-commerce.

Try an experiment.

Imagine that you want to buy an ink jet printer for your home office. Go to HP, (, Canon (, and Epson ( Frustrating? Yup. In fact, here is the experience of one of the authors (Norman). HP: Printers and printing supplies are the main source of income for HP, so you might think that so many people turn to the HP website for information, that ink jets would be there. Nope, nary a word about printers: the consumer has to guess. I guessed “Home & Home Office.” Worse, on my way to “Home & Home Office,” I was waylaid by a new window with a questionnaire that blocked the screen. Hey folks, I want to buy a printer, not answer your questions. If you want answers to those questions, make it an option I can select, not a bold intrusion on my workspace. In real life, I would have quit at this point: First, they made me unsure that I was embarking down the right path. Second, they thrust an irrelevant questionnaire in my face. Why stay and be insulted even more? But this was an experiment for Info Week, so I stuck with it. Hey, I got closer: the next page had “color printers” as an option. I clicked once more (click three, counting the one to get rid of the questionnaire). Oops, now I was confronted with meaningless pull-down menus that ask which of numerous unintelligible product numbers I want to learn about. I gave up: I don’t know what those product numbers mean. Maybe HP’s product managers understand the numbers, but not me. Do I want a DeskJet 970 Cse or a Cxi? Or a 952C? Isn’t it HP’s job to describe their product line and recommend to me which model I want? Instead, I have to laboriously use the pulldown menus to look at each printer in turn, try to remember the characteristics, and then decide. Not a chance. Goodbye HP — off to Epson. But no, even after I have decided to leave, HP enacts its final revenge: turns out HP had opened up a new window, so I couldn’t even back out. I had to use yet another mouse click to get rid of that unwanted window. HP — did you ever watch real customers try to use your site? I doubt it. Epson: Two clicks and I had a list of all their printers, with a short description. One more click and I had technical information on the one listed as “Printer Boasts Blazing Speeds,” and from there just two more clicks got me to Epson printers at Microwarehouse. But, um, I had just decided to buy the EPSON Stylus Color 900, and that particular printer wasn’t listed. Click, back to PC Mall: yes, they show the printer. Epson did OK, but not Microwarehouse: PC Mall would have won my business. Canon: if HP was bad, Canon seems to have little interest in selling printers at all. My experience with the printer companies was so bad that, just out of curiosity, I decided to try my favorite book site. “Hello Donald A. Norman” the first page said, “We have recommendations for you.” I ignored the recommendations, even though they were tempting, and typed “ink jet” into the search window prominently displayed near the top of the screen. Poof — in one click I got a page that showed a formatted list of ink jet printers, described them, told the price, shipping time and a customer rating for each item. Is it any wonder that I continually return to Amazon, continually buy from them? Now why is it the printer companies can’t do as good a job of making their products available to customers as a bookstore? All I can guess is that the printer companies don’t care about their customers: they sell mainly to distributors. Amazon does care — it shows. But the printer companies are forgetting something; let the distributor or point-of-purchase site do the recommendation, and you have just lost control: the recommendation could very well be for your competition. (Note added June 2001: The text above accurately describes my experiences at the time the study was done. Since then, all the companies have made major changes to their websites. Sometimes, these changes were even improvements. To see where they stand today, do the test yourself.)

The Web puts user experience of the site first, purchase and payment second. On the Web, people first experience the usability of a site and then buy something. Get a good experience, and the person is apt to turn into a frequent and loyal customer. But the web offers low switching costs: so, only if the site is extremely easy to use will anybody bother staying around. After all, there are eleven million other sites to go to on the Web.

Switching costs

The real difference between a person’s behavior on the web and the same person’s behavior in the physical world of real stores results from switching costs — how much effort it takes to switch from one vendor to another. In a physical store, the costs of switching are high. The person has driven to the store, entered the building and walked deep into the interior. Even when faced with dwindling supplies, inattentive or rude salespeople, and lines at the checkout counter, the purchaser is apt to stick with it: the cost of leaving, going to another store, and then possibly encountering the same behavior, is simply not worth the effort. Of course, in the physical world, many people then never return. (On the west coast, many in the technical community refuse to purchase at Fry’s except as a last resort, because the user experience is so bad, complete with pushy clerks who often know less about the technology they are pushing than the customer.) On the Internet, switching costs are low. If people fail to find what one wants, well the competition is only a mouseclick away. Get a pushy questionnaire in the face, and not only will the person not answer it, but the person may turn away in disgust and go to another vendor, never to return. Vendors such as try to overcome switching costs in several ways. First, they make it easy to find the item of interest. (See the box). Second, they make it worthwhile to return, for the more items purchased, the better their recommendations of items that are suggested. Both of us have at times purchased items from the recommendation lists, even though we did not start off intending to buy those items. Third, Amazon uses its affiliates list to make one part of the family, earning money by recommending the site to others. And fourth, the purchase process is as easy as any on the web, again fighting the low psychological cost of switching with low psychological cost of purchasing.

A low tolerance for difficult designs

Studies of user behavior on the Web always find a very low tolerance for difficult designs or slow sites. People don’t want to wait. And they don’t want to learn how to use a home page. There is no such thing as a training class or a manual for a website. People have to be able to grasp the functioning of the site immediately after scanning the home page for a few seconds. Of course, there are no rules without exceptions, and there are a few cases where a website is so useful that people are willing to spend some time learning the site. Extranets sometimes fall in this category. But it is dangerous to assume that a site will be so useful that people will give it more leeway than they give the other eleven million sites. Arrogant as well. Most brands are not nearly as treasured as corporate executives would like to believe, especially when switching costs are factored in. Take Coca-Cola, one of the most famous brands in the world. And yet, suppose you are on an airplane and ask the flight attendant for a Coke. If you are told “we only have Pepsi, will that be OK?” What’s the chance that you will say “yes”? We would bet it is pretty high.

Multiply two numbers: visitors and purchasers.

Success on the Internet depends on multiplying together two numbers: The number of people that will visit a home page times the proportion that actually buy anything — the percentage that become customers. It is expensive and difficult to get people to the website in the first place. This is why advertising budgets are so high. Increasing the second number is a lot easier. To double the success of a site, either double the visitors or double the conversion rate. Doubling the visitors could very require doubling the advertising budget. Doubling the percent who purchase requires initiating a human-centered design process for the website. Considering that many sites have conversion rates on the order of one or two percent, it is hugely more cost-effective to focus on the second number in the formula. Today, the vast majority of websites have miserable usability. Usability studies typically find a success rate of less than 50%. In other words, when the average person is asked to accomplish a simple task on the average website, then the average outcome is failure. Why is the Web so successful if the average user experience is one of failure? Mainly because of the few sites that do work. Even though 90% of current websites are incredibly bad, people only spend about 10% of their time on these bad sites. As soon as they discover that the site is filled with bloated graphics and little useful information, they go elsewhere. Worse, they are unlikely to return. If a site crashes their browser, they don’t go there again. If they can’t find the product they want, they will go elsewhere. And then they are apt to stick with the site that they know works. In other words, the initial visit to the site is triggered by the advertising budget or other links. When people have a positive user experience, they are apt to return. When they have a bad user experience, then they are likely never to return. Over time, people gravitate to the sites that treat them well and that are easy to use. Sheer Design Darwinism: survival of the easiest.

Discarded shopping carts

Suppose a company has a good advertising budget and a well designed site: is this sufficient? No. Studies show that a remarkable percentage of people visit a site, put items into the “shopping cart,” and then leave the site without ever finishing their purchase. Why would people discard their carts? After all, they have done the hard part: they found the site, figured out which items to purchase, and then when all they had to do was pay, they left. There are two reasons why people discard carts. Most shoppers like to do comparisons among the alternatives available. Make a tentative selection, then compare it with a few other alternatives. On many websites, this is not easy to do. One way is to remember to open up new windows for each of the items to be compared: most users are not comfortable doing this. The other way is to dump all the comparisons into the shopping cart as the simplest way of keeping track of the alternatives. In other words, If the site does not make it easy for people to compare, they will invent ways to do so, including the shopping cart. The second reason for discarding the cart is much more fundamental; the payment process is too onerous. Are customers asked unnecessary questions? Does the process get in the way? Is it overly difficult? Is the website secure and customers assured that their privacy is respected (and then, does the site actually do so)? Do the site provide information about stock availability, shipping and other extra costs before the purchase process or is the customer surprised mid-way through? Making these final steps as pleasant and easy as possible is just as important as any of the other steps. The customers are already at the site. They have already decided what items to purchase. If they leave at this point, it is the site’s fault. From a business point of view, this is inexcusable.

How to provide good user experience

The good news is that it isn’t difficult to have superior user experience. Since most sites are so bad, it does not take that much to stand out and be one of the easiest sites on the Internet. Get rid of the spinning logos and boastful marketing and focus the site on people’s needs and plainspoken language in a layout that it easy to scan and fast to download. Make the site architecture fit the way customers think, not the way the company’s product line is organized. Avoid the HP error. Providing a customer-centered organizational architecture is important, but it can be surprisingly difficult. It means that competing lines within a company might very well have to cooperate in the structure of the web. Remember, the customer doesn’t care how the company is organized; the customer just wants to know which product to buy, what its characteristics are. And then the customer wants to be able to buy it, right then and there. If a company does not wish to compete with it’s distributors, they should make it easy to go to consumer sites and get to the item in question immediately, with no further search. Although doing website properly is easy in principle, it takes the right staff. Professionals in user experience understand people. They know how to do field research, interaction design, develop rapid prototypes and do rapid tests. They work closely with graphic designers, web coders and marketing. It is critical to have the assistance of designers who understand people, that the group has an appropriate budget, and that it has sufficient authority to execute upon their recommendations. A deep understanding of needs requires field studies of customers. Developing the perfect website is a major undertaking, especially because of the need to overcome the primitive technical limitations of the current Internet. Luckily, the site doesn’t have to be perfect: just sufficiently better than the competition. How to get a better site? Observe real customers as they actually use the site. A professional user experience team will go to the customer’s home, office or normal place where they will access the site. Forget focus groups. Focus groups reveal what customers think, not what they do. They reveal what customers think they think, how customers think they behave, not what they actually believe, not what they actually do. Don’t trust what customers say, trust what they do. Quick and dirty user testingAlthough nothing can substitute for a professional user experience group, a lot can be accomplished quickly, easily, with little time, effort, or expense. The main secret is to observe real people doing real tasks. Ask if you can visit customers in their offices or homes. Even though friends, colleagues and spouses are not likely to be exactly like your real customers, you can learn a lot by observing them. But watch them do real tasks, not some silly artificial task that you made up. Watch as they answer questions that are of interest to them, as they search for products, as they try to make purchases — real ones. (If you have to, offer to pay (“you told me you need a printer: I’ll pay you back for it if you will let me watch how you buy it.”) Watch. Don’t interrupt. Don’t offer to help no matter how great the temptation. Don’t ask why they have done something until after the session is all over. (Sit behind them so you aren’t in the way.) If you need to bring people to your location for the observations, Invite one person at a time so that you can observe that person’s natural behavior without the bias introduced by the presence and comments of other people. Watch each perform a set of typical tasks with the site. For example, on a corporate site, it might be reasonable to ask customers to research a product purchase and determine which member of a product family best fits their needs. On an airline site, ask them what they typically do at a travel site and then watch as they try it on yours. Although it is best to observe representative customers, tests of colleagues, friends or spouse are better than no tests at all If your friends, colleagues or spouse have trouble with the site, beware.) Websites are so bad that the major problems can be found with very simple tests that only take about a day to run.

Usability is not a luxury

Usability is not a luxury on the Internet. It is essential to survival: it is the key technique for superior customer relationships. Remember, because switching costs are so low, attention to usability increases the percentage of those who actually complete a purchase after visiting the site. It’s a lot cheaper to increase the human-centered design budget than to double the advertising budget. The Internet obeys a kind of Sheer Design Darwinism: survival of the easiest.

About the Authors

Jakob Nielsen, Ph.D., is a principal of Nielsen Norman Group (, a user experience company. His latest book is Designing Web Usability: The Practice of Simplicity. Donald A. Norman, Ph.D., is a principal of Nielsen Norman Group and President of Cardean Learning Systems. His latest book is The Invisible Computer: Why Good Products Can Fail, the Personal Computer Is So Complex and Information Appliances Are the Solution.