Design for the Scent of Information

Jared Spool is CEO and co-founder of User Interface Engineering (UIE), a think tank. They advise clients how to design their browser-based systems for the best information retrieval experience possible. This includes web sites, but also knowledge management systems –nowadays better known as enterprise content management systems.

Imagine yourself looking for inkjet drivers on HP’s website. That’s not too hard is it? Chances are that you found HP’s site pretty much supporting your efforts without the need for passing by the search box. But have you ever wondered why you thought HP’s –or your own favourite vendor’s– site did such a good job?

The chaps at UIE have. In fact, that is what they do for a living. They invite people like you and me in their labs, and let them perform information lookup tasks. They watch our every move and if necessary they’ll ask us what we thought when we were hunting for that one piece of vital information.

If I would have to compare UIE with anything else, Nielsen’s usability research firm would be the first to come to mind. But UIE is not about usability, it goes way deeper. In fact, it goes right to the core of what we do on the Web, on intranets and extranets: we find ourselves hunting for information.

Jared Spool was kind enough to set up a chat with me on the subject of his latest report. Initially, I was going to meet Jared in the flesh, at the KCCEurope event in Amsterdam for one of the paper magazines I wrote for (Computable in the Netherlands, for the curious), but Jared couldn’t make it. As he was the main reason why I wanted to go to KCCEurope, I rang him and asked him if he would be willing to answer my questions on the phone.

Jared did better; he proposed to Skype so that it wouldn’t cost me a fortune, and he offered to send me his latest report, the one on designing for the scent of information. And so I had a good and interesting read, and we both had the pleasure of hearing each other’s voice as if we were sitting in the same room on the 9th of November.

The report

I used the word “hunting” a few times in the previous paragraphs. I did that on purpose. It turns out, UIE found that people searching a large information space, act as “informavores” on the hunt for information. The path towards the information a user needs must have a “scent” to it –it must make us feel confident that with every step we make, we’re getting closer to our objective.

And so, the report tells about user that have been subjected to UIE’s questions on how they searched for specific information. The word search, however, is perhaps a bit misplaced: “Browsing is actually more reliable to find what you’re looking for,” said Jared Spool in the interview.

The example used in the report covers HP’s web site and a user who is trying to find the latest drivers for his inkjet printer. When he lands on the home page, a button clearly labelled “Drivers” takes him to another page which contains a small form with a drop-down menu holding all of the types of drivers available. The user selects Printers and without hesitation clicks the Next button. Then he is presented with a form asking him all kinds of questions, but as all of these questions are related to narrowing down his search, he fills in the blanks without complaining and finally he gets what he wants.

Most of the interface elements were above the fold –the fold is the invisible dividing line between the first screen full of information and the rest of the page that is hidden from the user. To reveal the data below the fold, one must scroll down.

Throughout the first example, the UIE team comments on glitches on the various pages which don’t seem to stop the user in his steps. “That’s because the scent is always there. Every choice made leads closer to the goal. If one of these steps would lead away from the objective, the user would start feeling confused and would eventually become frustrated,” according to Jared.

The report holds more than just the conclusions of various user research studies. For example, it will also tell you what not to do. A good example is dividing up a page with horizontal rulers. If such a ruler happens to fall right above the fold on some users’ screens, they would psychologically be challenged to scroll down further.

Jared told me it doesn’t matter whether the web designer was a professional. “It will happen to any designer who doesn’t start the design with the content instead of with the home page,” he said. According to Jared and his team, designers should first start with the content pages and then design the home page(s) from there. That should make them answer questions like “What is the user doing her?”, “Which page is most important to them?”, “How will they find the page?”, “What are the trigger words?”, etc.

“The worst is that this same type of faults also happen with Enterprise Content Management System interfaces, where it is common the customer gets stuck with a browser-based interface that cannot be changed,” Jared said. According to him, the fact that you’re in a company, and perhaps more knowledgeable of the information scattered around the intranet, doesn’t help.

“The problem with large information systems is that it is extremely difficult to know how others will get to the information they need,” Jared said. You might argue that, if the interface itself fails, there is always the search engine as a rescue technology. But unfortunately, search engines and their relevance ratings don’t pass the test at UIE. “Relevance ratings don’t mean a thing to most people. What’s the difference between a 80% hit and a 60% one?”

The UIE report and Jared’s answers to my questions made it clear to me: they show designers and professional web developers should rethink their habits. Not just the habit of starting a design with the home page, but also the habit of thinking in myths; like the one that says people hate to scroll.

They should put the user on central stage, and think through their whole design concept from his point of view. They should try to avoid forcing a user to find information using the search engine. Browsing is better.

One thing UIE and Nielsen’s usability research has in common is their view on graphic elements. Both tell us not to use graphics because of the looks, but only when it adds value to the experience –like in navigation or to explain a feature. It’s only one of the other issues you’ll likely be surprised to know about when having read the UIE Designing for the Scent of Information report.