Selling hotdogs in an emergency room
By Garry Kranz
Web guru Jared Spool is an iconoclast and de facto historian of the evolution of Web design. Spool, author of "Web Usability: A Designer's Guide," founded Bradford, Mass.-based User Interface Engineering in 1988. The company provides consulting and other services aimed at helping development teams design more usable Web sites. In this special interview with The Customer Facing Technology newsletter, Spool voices his opinion on the state of Web design - challenging some long-held beliefs along the way.
What common mistakes do you spot companies consistently making during their Web design?
There are two of them. First, most designers don't have a clear understanding of their company's business goals. Every design element - every link, every button, and every graphic - has to directly serve those business goals, and serve them in a measurable sense. In other words, you should be able to tell that when you remove an element, the site no longer works and when you put the element back, it begins to work again.
The other mistake is that, as soon as a site is launched, designers move on to new projects. There's no learning going on about how users respond to the design elements. As a result, you get a lot of stuff on a site that doesn't make sense. In the five years we've been studying Web sites they have not improved. The best sites are no better now than they were five years ago.
How can companies measure whether a user's visit has been satisfying and might lead to future sales?
Imagine this scenario: you've got some magic detector that allows you to tell when someone within a five-mile radius has run out of milk. You drive to their house and take them to the nearest 7-11, and just to make sure they purchase, you give them the cash. The 7-11 store would really have to screw up to fail. That's exactly what we did on the Web. We found people who needed products. We brought them to sites that had those products. To make sure they purchased the product, we gave them the cash. We got a 30% conversion rate.
What happened?
It was just a matter of finding the product. When designers create lists, they tell us they want users to go 'pogo-sticking' up and down the list. But we found that 66% of purchases occurred with no pogo-sticking at all. The user looked at only one item on the list, and that was the one they purchased. It turns out the more pogo-sticking users do, the less likely they are to purchase.
So what's the best way to convert users' visits into sales?
The key way is by making sure that, at the list level, users have enough information to decide what item they want to purchase. That information is different for different types of products. E-commerce systems that have the appearance they were built for generic purposes, and not for specific products, fail because they don't take into account that different products require different types of design.
Aren't these more marketing than technology issues?
What's the difference between marketing and design? If the goal of the business is to sell something, everything the user sees has to be aimed at making sure that purchase happens, including design. In our testing, customers knew what they wanted and were set to buy it. Yet 7 out of 10 times, the sites screwed it up. These are major sites like Amazon, CDNow, L.L. Bean, and REI.
Is it just a matter of improving Web site stickiness?
The whole notion of stickiness is a myth. We have no evidence to suggest that sites that are sticky are any better than sites that aren't. This notion has led companies to base their businesses on complete falsehoods. Then they're confused when the business doesn't succeed. What if I told you that the success of a Web site depended on the number of "E's" on the home page? You'd question that: it's a completely ridiculous notion. There's no evidence to suggest that stickiness works. It's equally ridiculous.
What are the main issues to consider when building or rebuilding a Web site?
Companies need to understand explicitly why they're in business. They next need an in-depth understanding of why a user comes to their Web site: what makes it a successful experience or failed experience. Once they do those two things, what they do next is obvious: put together a site that meets those goals. By doing those two things, they'd have measurable criteria that enable them to tell whether the site is doing what it's supposed to.
But how can a company truly know why visitors come to its site?
There are attitudinal studies that you could do that measure a user's attitude before they visit the site, and then measure the same users after they've visited the site. These are useful because they enable you to look for changes in user behavior during that interval, and it's actually fairly reliable. There also are long-term behavioral studies you could do.
How will the Web design industry evolve? What's in store?
What's needed is an automated tool that enables companies to tell whether they've actually achieved their business goals, and allows them to know instantaneously if their Web site isn't as usable as it was an hour ago. That way, designers would be able to pull down the site and fix it.
It sounds like you're describing an entirely new layer of enterprise system, one devoted exclusively to monitoring the effectiveness of Web design.
Why not have some combined software-hardware device that looks constantly at your Web site and examines the behavior of people visiting your site? From those things you'd be able to tell how much pogo-sticking is going on. We know pogo-sticking is a predictor of a failed sale. Combine it with several hundred other predictors and all of a sudden you'd get a real accurate picture of what's happening on your site.
What about the presumption that companies who lack a Web presence are doomed to fail?
There are a lot of Web sites that make you wonder 'Why the hell is this company here?' It's like someone trying to sell hot dogs in an emergency room. Sure, there's a need for hotdogs, and sure, emergency rooms are places with lots of people traffic. But does it make sense to sell hotdogs in an emergency room, or is there a better use of space? A lot of companies put up Web sites because they have something to sell, and figure that therefore they need one. But they haven't thought it through; they never ask themselves what they want to occur and whether the Web is the right locale for them.
Garry Kranz is a freelance business and technology writer based in Richmond, Va.
MORE INFORMATION ON THIS TOPIC:
