Why Websites Suck

Gerry McGovern helps large organizations change how they manage their websites from a model based on technology or content to one based on customer tasks. His latest book is called The Stranger’s Long Neck: How to Deliver What Your Customers Really Want Online. This book explains about how to implement a “top-task” management model for websites. In this interview, he explains why websites suck and what you can do to ensure that yours doesn’t.

Q: Why do websites suck?

A: Organization-centricity. Look at Dell. It’s not the worst website, but it has a huge flaw. If you want to buy a laptop from Dell, you must first choose whether you’re a home user, business, etc. People don’t want to do that. So why does Dell do it? Because “home,” and “government,” etc., are powerful business units in Dell. At the root of practically, every website failure I’ve come across is the organization wanting the customer to fit around how it’s organized, how it thinks, and the language that it uses. By contrast, great websites organize around the customer.

Q: What causes the bad design of websites?

A: Designers are still designing brochures, billboards, or TV ads that they call websites. Bad web design is constantly trying to get your attention. It’s flashy, it’s graphic heavy, it’s in your face. Great web design is invisible. It pays attention to your needs—the tasks you need to complete. The more you notice the design on the website the worse it is.

“The Social Network” is a great film but it’s classic Hollywood. The characters all talk about how Facebook must be cool, but the designers of Facebook are much more interested in the website being useful rather than cool. There is one word to describe great web design: useful.

Q: What is your favorite website analytics tool?

A: I don’t have one. I depend on time and motion studies to understand if a website is working or not. I take a top customer task and get real people to try and complete it. I try to become as invisible as possible and observe. Patterns always emerge.

Q: How do you know how people use your website when there are so many variables such as type of customer (prospective, new, and existing) plus bots and spiders?

A: There are always top tasks. Every website has a “book a flight”—it’s just that many have not discovered it yet. I have been doing this since 1994. The conversation with clients always begins this way: “We’re different, very complex, We don’t have top tasks. We have so many audiences.”

Classic organization-centric thinking. I have done hundreds of task identification projects the top tasks are always there. They are audience independent, geographic independent. For example, what’s the top task of a health website regardless of age, sex, income, professional, geography?

Q: What is the ideal way to design a website starting from scratch?

A: Use real data to discover your customers top tasks and build around them. Often there’s only one top task, and at maximum five. Realize that when you launch—even if you’ve done the best job in the world—you will be at 60 percent efficiency. You must continuously improve your top tasks. The great website evolves through a process of grinding out incremental improvements. And remember that the tiny tasks get in the way of the top tasks. The bigger most websites become, the worse they become.

Q: By contrast, how do most organizations actually design a website?

A: It’s a project. Here’s a scenario. There’s an outbreak of swine flu.

“What are we going to do?”

“Well, we have to get a website up.”

“But can’t we integrate it into our existing website?”

“No. We have to show we’re doing something. That requires a separate website.”

“And what are we going to put on this website?”

“We’re going to tell people about all the great things we’re doing to prevent the spread of swine flu, how much we’re investing. And we’ll have lots of information about how citizens can prevent the outbreak spreading. And we’ll have pictures and speeches from politicians. And another thing we can’t use the words ‘swine flu’.”

“We have to write about the ‘H1N1 virus.’”

“Why’s that?”

“Because the pig industry is very powerful, and it doesn’t like us using the word ‘swine.’”

“But that’s what people call it. That’s how people are searching.”
“I love your youth and enthusiasm, but that’s not how things work around here.”

“But this is the Web. It’s how things work on the Web …”

Q: What is the ideal way to fix an existing website?

A: Fix the top task and work down from there. Do not do a redesign. That’s “lipstick on a pig” web design. Most redesigns are like alcoholics going into an expensive rehab without genuinely addressing the underlying problems.

Q: What would happen if companies like Amazon and Apple blew off the long tail and reduced their offerings?

A: Apple is a poster child for focusing on the “long neck” and cutting off the tail. One of the first things Steve Jobs did when he returned to Apple was to radically prune the product range. And look at how the iPod started: It focused on making top tasks—playing music—really easy whereas the other competitors were feature crazy and complex as hell.

Amazon has changed. Over the years Amazon has moved more and more to a top task/long neck design. So, if people want to buy a five-string bass guitar from Amazon, they end up on a page that’s absolutely focused on five-string bass guitars. Amazon even got rid of the tabs at the top of the page. Basically, the only long-tail link left is “Shop all departments.” Equally, Amazon knows how to really promote its top tasks such as buying a Harry Potter book when if first comes out.

The long tail is not bad per se, but its value has been massively overblown. There are a lot of hidden costs to managing the long tail, and if you don’t do it properly that tail ends up wrapping itself around the long neck and choking it.

Q: How does one measure the effectiveness of a website?

A: There are three measures and they all revolve around top tasks:
 

Success rate: What percentage of your customers are able to complete the top tasks?

Disaster rate: What percentage of your customers thought they had completed a top task but didn’t?

Completion time: How long is it taking people to complete a top task?
 

Fixing the basics is about bringing the success rates of your top tasks above 90 percent and the disaster rate below 5 percent, then the best practice is a relentless focus on the customers’ time. If they save time, you make money.

Q: How do you know if people spend a long time on your website and visit many pages because your website sucks or is engaging?

A: Unless you know what the task of the customer is, you don’t. I did some testing with a technology company recently and one of the top tasks of customers was to download software. So we gave fifteen customers a set of download software tasks. Page 1 in the process worked great. So did page two and three. But when they got to page four, the cursor went dead still for some people. For others it started whirling round in circles or going up and down the page. It was really clear that page four was confusing, and people were spending lots of time on it.

Q: What do you think of Alltop’s design? Too many topics, too many choices within topics, or is the core value the long tail of topics and the long tail of sources within those topics? I really want to know. Don’t hold yourself back.

A: Alltop is seeking to answer the questions: What’s hot? What’s interesting? There’s a particular slant to technology and the quirky stuff of life. So it has a top task, so to speak. And in this sense it’s not unlike Google News. You go to it to get a sense of what is interesting in the world today.

Much of it seems automated so the costs of production should be low. It’s hard for me to give an opinion of whether I like it or not. I’d need to check it out of times to see do I like it’s editorial angle. But I think the basic design is neat and clean. Even though there’s a lot on the page, it’s broken into mini-pages of the top stories from various sites, so it’s a fairly easy page to scan.