Jakob Nielsen's Alertbox, June 12, 2006:
Summary:
Newsletter usability has increased since our last study, but the competition for users' attention has also grown with the ever-increasing glut of information.
Only two years have passed since our last study of email newsletter usability. For almost any other topic, conducting a new study so soon would be a waste of time. Usability guidelines change extremely slowly because they depend on human nature, which tends to change at a rather glacial pace. Newsletters are one of the few exceptions to this rule because their environment -- the inbox -- grows ever more crowded each year.
To assess email newsletters' current usability, we conducted a new round of user research (described below). As it turned out, all of our original findings continue to hold. The 127 design guidelines from our previous newsletter report are still valid, though we refined some in light of the new research. The main change is in volume: we added more guidelines based on additional insights derived from eyetracking studies of how users process their inbox and read their newsletters. We now have 165 guidelines for email newsletter usability.
Our main conclusion remains the same: Email newsletters are the best way to maintain customer relationships on the Internet.
We tested 117 email newsletters, as follows:
Finally, we included a field study component wherein we observed users in their offices during a typical workday. This ethnographic approach allowed us to study how participants used newsletters and news feeds in an environment with many competing information sources and demands on their time.
Despite the improvements, however, much remains to be done. Users should be able to subscribe to a newsletter in less than a minute, for example.
Also, in the latest study, the success rate for subscribing to newsletters was 81%. While higher than the 66% we found when testing a broad range of websites, a success rate of 81% implies that a newsletter with 50,000 subscribers could gain an additional 11,700 subscribers by improving the usability of its subscription process.
People were highly inclined to skip the introductory blah-blah text in newsletters. Although this text was only three lines long on average, our eyetracking recordings revealed that 67% of users had zero fixations within newsletter introductions.
Eyetracking heatmap of users reading a newsletter. Notice the emphasis on reading the first two words of the headlines.
The areas where users looked the most are colored red; the yellow areas indicate fewer views, followed by the least-viewed blue areas.
Users often deliberately trade off newsletters against each other to reduce their email volume. A good newsletter might be booted if a better one comes along. People are conscious of the need to protect their time, and they try to identify the best newsletters for their various information needs.
Many users are getting quite skilled at managing email, since it's such an important tool for both business and personal use. (Email is the Internet's true killer app -- websites are a poor second as far as most users are concerned.) On average, users maintained 3.1 email accounts each, using different accounts for different purposes.
As time passes, users are also accumulating "old" email accounts that they rarely use but might check sporadically. If your newsletter arrives in such an account, users might not see it while it's fresh, so you could see a "slow tail" of clicks and orders coming in from such users. There's not much you can do about this, other than to ensure that your newsletter's links and special offers don't expire too soon.
Several participants used special accounts to sign up for things that they feared might be spam or be generally less useful. It's imperative to convince users that your newsletter belongs in a higher-priority account. You can accomplish this by following the guidelines for giving a credible impression in the subscription interface of your newsletter's value and publication frequency.
Now that most email services provide gigabyte-sized storage, users are archiving more newsletters than they used to. On the one hand, doing this adds to their information overload. On the other hand, it enhances the long-term value of newsletters compared to other Internet media forms. The reason for this is that stored newsletters become part of users' personal information spaces and thus will be found when they search their own environments. Although your newsletters don't need full-fledged SEO (search engine optimization), you should consider how users might want to retrieve old issues in the months or years to come.
First, the labels obviously violate the guideline established in our very first report on email newsletters in 2002. It's extremely well documented how to phrase the two options so that mainstream users will understand them. There's really no excuse for the geeky labels seen here.
Second, as we reported in 2004, the horizontal layout of the radio buttons is simply begging users to select the wrong one (as detailed in guideline #6 for checkboxes and radio buttons). Sure enough, this is exactly what happened in our study: several users who were technical enough to understand the difference between "HTML" and "Text" thought the "HTML" label applied to the circle to its right, and selected accordingly.
Actually, "thought" is the wrong word: clicking happens so fast that people don't really think. If users thoroughly analyzed these radio buttons, they'd probably use them correctly. However, nobody ranks "understanding radio buttons" as a main goal in life. People would rather spend time with their kids, so they click website features as fast as possible, meaning that designers need to prevent errors by avoiding error-prone designs like the one above.
Some users were familiar with the general idea of feeds, even if they didn't know the term "RSS." This was typically because they were receiving some feeds on their My Yahoo! pages or other personalized portals.
Users had very mixed feelings about feeds. Some people liked viewing information from multiple sites in a single, centralized location rather than having to go to each site. Some users also liked scanning a list of headlines without seeing content they didn't ask for. Finally, some users appreciated the ability to determine when they would view their news items. This is in contrast with newsletters, which arrive at times beyond the user's control.
On the other hand, many users had negative feelings about feeds. People already suffering from information overload resent having to go to yet another source of information. In contrast, email newsletters arrive in a tool that users are already using, so they don't impose the burden of having one more thing to do. Email is also easier to archive for later use, whereas feeds have an ephemeral nature.
Several of our participants had feeds on their My Yahoo! page that they'd stopped using. Many previous studies have found that users are reluctant to spend time customizing portals, so it shouldn't be surprising that some users decided to stop looking at that part of the page instead of editing their preference settings.
Finally, some of our users resented the fact that news feeds are divorced from the context of the publisher's website. They preferred the serendipity that came from visiting a full-fledged website that offered additional content beyond the current headlines.
News feeds are definitely not for everybody and they are not a replacement for email newsletters. Feeds can supplement newsletters for sites that cater to users who prefer a centralized view of headlines. This will primarily be newspaper sites and other sites that focus heavily on news and breaking stories, as well as sites that target Internet enthusiasts. For sites that target mainstream business users or a broad consumer audience, news feeds may be less important. Such sites might be better off emphasizing higher-quality newsletters and a choice of publication frequency.
Our eyetracking of users reading news feeds showed that people scan headlines and blurbs in feeds even more ruthlessly than they scan newsletters. When you appear in somebody's newsreader, your site has a diminutive footprint that's rubbing shoulders with a flood of headlines from many other sites. Under these conditions, users often read only the first two words of a headline, so it's crucial to have brief headlines and to start them with the most information-carrying words.
Feeds are a cold medium in comparison with email newsletters. Feeds do not form the same relationship between company and customers that a good newsletter can build. We don't have data to calculate the relative business value of a newsletter subscriber compared to a feeds subscriber, but I wouldn't be surprised if it turns out that companies make ten times as much money from each newsletter subscriber. Given that newsletters are a much more powerful and warm medium, it is probably best for most companies to encourage newsletter subscriptions and promote them over website feeds.
Copyright © 2006 by Jakob Nielsen. ISSN 1548-5552