The Nielsen Norman Group, established in 2000, brought together Jakob Nielsen and Donald Norman, two of the most influential names in usability. Their core business activities include producing reports on usability issues (backed by solid research) and presenting conferences that provide practical ways of improving usability.

We have attended both of their conferences in Sydney and found them very useful in maintaining and enhancing our usability skills and highlighting issues identified through extensive (and therefore expensive) research.

This year, we selected two sessions to attend in Melbourne: Web Usability 2004 and Email Newsletter Usability.

There were a number of other sessions that we did not attend, because:

• They were only offered in the United States or continental Europe (eg Interaction Design; Visioneering: Future Planning)
• They were not relevant to the scale or scope of our business (eg Usability in Practice: 3-Day Intensive Camp; Intranet Usability; Presenting Company Information on Corporate Websites)
• They were 'repeats' of past sessions that we had already covered (eg Beyond ALT Text: Designing Accessible Websites)

Unfortunately, some of the excellent sessions from last year's conference, including Information Architecture I and II (Louis Rosenfeld and Margaret Hanley [ex-Sensis, now BBCi]) and Content Usability: Writing for the Web (John Morkes [ex-Trilogy, now Expero]), were not available this year (either as repeats or new sessions with the same presenters). The variety of sessions last year required us to send three staff to attend; this year I was the only one.

This really sums up the entire Web Usability 2004 session. There was good news - we are getting better (albeit slowly) at creating more usable Web sites and ensuring users needs are considered throughout the design process. But there was also bad news - expectations are rising, and users (as always) are fickle. And of course, the issue of designers not understanding usability reared its ugly head.

The session was presented by Hoa Loranger, with Jakob making special appearances throughout the day. Her presentations tended to focus on the research the Nielsen Norman Group conducted, rather than the outcomes, and although sometimes dry, was essential to understanding the analysis they had made.

Much of the day was spent looking at user behaviour, rather than analysing the approaches used on Web sites. This was refreshing when compared with past presentations, which could be seen as somewhat more opinion rather than fact. By focusing on the users and their behaviour (no matter how unexpected or seemingly irrational), the research was able to identify the reasons why Web sites are not usable.

And the answer? In short, 1/3 of the failure is due to poor search or information architecture, 1/3 is due to inappropriate content and presentation, and the remaining 1/3 covers more esoteric technical issues and failures.

This is a fundamental shift in the approach to Web usability; in the past, recommendations focused on avoiding particular technologies (think frames or Flash) rather than encouraging better design practices (information architecture, metadata development and content management, for example).

This shift is confounded by the slowdown in the lifecycle of the average Web site - from as little as a few months in 1997 to two years in 2004. Technology has slowed dramatically (Jakob suggested that we'll be waiting for Longhorn for any major browsing enhancements, aside from Windows XP Service Pack 2), but the application of the technology is the most significant issue.

The day concluded with a look back at Jakob's past recommendations for Web usability, updating them for 2004 based on their research with users. Many larger issues have been resolved, but there are still plenty for interface designers to worry about. Unfortunately, the number of issues is still very high, and the rate of change is quite low. Hence Jakob suggested (tongue firmly planted in cheek):

We'll see.

A much smaller session than Web Usability 2004, Email Newsletter Usability focused on the results of research into end-users of email newsletters - ranging from their ability to subscribe/unsubscribe to their opinions on what makes a good email newsletter and how to minimise the impact of spam filters.

Users had extremely high success rates for subscribing and unsubscribing, although the tasks took longer than they should (a couple of minutes). Surprisingly, 'single click' unsubscribing was confusing to some users, which led to a recommendation to require confirmation before processing the unsubscribe request.

In fact, there were a few surprises in the research, perhaps due to a relatively small sample size; although there were no significant differences identified between United States and international (Australia, Great Britain, Denmark) users.

Unlike Web Usability, there were several opportunities to practically implement the recommendations that emerged from the research, including a detailed analysis of the (dysfunctional) process of subscribing to a particular newsletter.

This practical implementation was very valuable, and led to some interesting discussion amongst participants - who had very different backgrounds in email newsletters, ranging from intra-organisational (at a single site) to extra-organisational at a global level. This contrast was useful, as it demonstrated the need to ensure that the usability of the newsletter meets the needs of both users and the organisation providing the information (to minimise support costs and maximise user satisfaction).

The research provided a good background to email newsletters, but was limited in its selection (mainly larger organisations for external users) and variety of approaches to handling different issues (eg email bounces, compliance with privacy and spam legislation). As such, the presentation was skewed toward issues that didn't necessarily apply to all participants, and didn't include issues that did affect some participants. Nonetheless, the general principles were good guidelines for all participants to follow, and user testing would ensure that they meet the needs of their specific user groups.

One of the key reasons we selected this session was to obtain a 'free' single-user copy of the Email Newsletter Usability report; these reports are an excellent summary of the issues addressed in the sessions (other reports were also available at heavily discounted (30-50%) rates; we took the opportunity to purchase the report on confirmational email and transaction messages, amongst others).

I have been able to use the guidelines in these reports to review a number of our products and suggest changes to improve different aspects of their usability; this will have an immediate and potentially significant impact on our ongoing support costs.

Probably the most significant change from past years was the absence of a 'main event' - a clear drawcard focusing on speeches by usability 'experts'. The first year was a 'dream team' of usability experts (Jakob Nielsen, Donald Norman, Bruce 'Tog' Tognazinni, Brenda Laurel); the second year featured those implementing usability at a variety of .coms (Amazon, Yahoo!, Oracle).

This year the closest thing to a 'main event' was Web Usability 2004, which seemed to attract the most participants during the final two days. Perhaps this reflects the move away from the .com world in the past couple of years, but there is a place for a broad spectrum of opinions and experiences; it would be valuable to have a 'main event' at future conferences.

As always, the highlight of the tutorials was the use of actual user tests to demonstrate particular points. The videos, which combined screen recording and video footage (using TechSmith's incredible Morae), were of particular interest to those who had not conducted user tests with inexperienced users. They really demonstrated that you can try as hard as possible to create usable designs, but unless you test them with end users, there is no guarantee that they will be effective.

The numbers were the same as last year, but the conference seemed to have a more intimate feel, perhaps because of the smaller individual sessions (apparently the 3 day camp was very well attended). Most of the attendees were from Melbourne, although Tasmania again had some government departments represented (DPIWE, Education and Tourism).

Of course, conferences such as this one are a great networking opportunity, particularly for organisations like ours that are based outside Sydney and Melbourne; the conference was a great opportunity to catch up with attendees from previous years, Hobart-based attendees from different organisations, and attendees from a wide range of large companies and government agencies (including Shell International, Telstra and BigPond). It is worth noting that this year, there appeared to be a much greater predominance of academic and government participants, with less private industry involvement (especially at a small business level).

As an aside, it was a nice change to see the conference held in Melbourne this year (although a Sydney leg was added later). The facilities provided at Crown Promenade were very good, in particular the free WiFi access during the conference. The only downside was a lack of book sales (New Riders had in the past operated a stall with their range of industry books). Since we're not so Sydney-centric anymore, perhaps Brisbane would be a good choice for next time?

The NNGroup Usability Week 2004 conference was a great opportunity to find out the current state of commercial usability practice, identify issues that have emerged from user testing, and network with a variety of different practitioners. As with previous years, it was very well organised, and there were plenty of opportunities to question presenters and discuss issues formally and informally.

Although it did not have the breadth of tutorials and variety of presenters as the User Experience 2002 conference, I felt the sessions were well targeted and relevant to the needs of participants. A 2006 conference would be more successful if it broadened the range of tutorials (particularly in the areas of information architecture and content management) and variety of presenters (bring back Lou!).

For an alternative perspective, Craig Cook has written about his experience of the San Francisco Web Usability 2004 and Email Newsletter Usability sessions.

Miles Rochford is one of Human Solutions' project managers, specialising in information architecture and user-centred design.

He attended Usability Week 2004 in Melbourne on 20 and 21 May 2004, and was also present at the Sydney legs of the User Experience conference in 2002 and User Experience World Tour in 2001.