Dangerous Usability Findings

0

Today I saw a slide in a meeting that said: Usability Finding: People don’t like to read on the web, only skim.

In short, I agree, but with conditions. People SKIM to find what they want to READ so they can decide what to BUY.

And that was my problem. The conditions were not communicated. I feared the marketing audience would internalize this statement as a general ‘online tenant’, and therefore de-value online copy.

Since I no longer manage the site, I started to bite my tongue. However, without further explanation, I feared this axiom could have far-reaching consequences. So I spoke up.

For four years I managed Dell.com Consumer, and participated in 30-40 usability studies, with all types of methodologies. I captured pages of insights and ideas from these studies, and they were always valuable in gaining empathy for the customer. But I also got data from surveys, financials, web metrics, previous studies, and my experience. I never made a major decision without these balanced data points, or at least a real-life test.

In this study, if the task presented were to BUY a computer, I guarantee people would be reading much more diligently. I’ve seen studies where some people (i.e. “the readers”) read everything line for line, just to make a navigation decision. The littlest thing can stop them. The majority of participants, however, skimmed to find what they were looking for. Regardless, you can’t generalize behavior in usability to real life. You can gain insight and direction.

In short, I agree visitors do skim. That’s why writing in reverse pyramid format (answer first), formatting, chunking copy, bullets, and succinct relevant copy is important. However, results on a selling page would suggest that long copy pages sell better. Customers want the details when they’re considering buying, and that benefit statements drive conversion. And you are reading this!

I think the team understood this and simply mis-communicated (I think). Certainly they agreed they would not make major changes without an A/B test. But, it made me realize how poor conclusions can be drawn by generalizing usability findings or communicating them with incomplete context. Be careful out there.

Leave a Reply

© 2004 Decker Marketing. All rights reserved.
Proudly designed by Theme Junkie.