Ten Best Intranets of 2003

by Jakob Nielsen on October 13, 2003

Summary: This year's winning intranet designs emphasized workflow support, self-service content management, and offloading tasks from email to collaboration tools. On average, companies spent three years between redesigns, and one year on the redesign itself.

This is the first year in which several submissions had implemented many of the usability guidelines and best practices for intranet design that we've featured in our previous intranet design annuals and intranet usability test reports. We didn't start publishing these reports until late 2001; considering the turn-around time to redesign a big intranet, it's understandable that our previous recommendations have only now begun to impact real projects.

It is in the nature of awards that only a few winners can be singled out for praise, but this year's intranet design competition proved that excellence can now be found in many intranets. Every year it becomes harder to select the winners, and harder to be unable to acknowledge the many great designs that didn't quite make the top ten.

The Winners

The ten best intranets of 2003 are:

  • Amadeus Global Travel Distribution, Spain
  • ChevronTexaco
  • Design Matters, Inc., a Web design agency
  • FIGG Engineering Group, a consultancy specializing in bridges
  • Fujitsu Siemens Computers, Germany
  • Landor Associates, a brand strategy consultancy
  • Mayo Clinic, a non-profit medical center
  • North Tyneside College, U.K.
  • United States Coast Guard
  • Wachovia Corporation, the 5th largest US bank

We also awarded an honorable mention for excellent redesign of a previously winning intranet to silverorange of Canada, a 2001 winner.

Winning Characteristics

Most of this year's winners are big organizations with tens of thousands of employees, though we don't have any 100,000+ companies like we did last year. Big companies can afford substantial investments in intranet usability, since their ROI is enhanced when their many employees become more productive.

Still, good design doesn't always require big budgets. Our winners include three fairly small consulting companies, of which only one has the "unfair" advantage of being a Web design agency. Such knowledge-intensive companies benefit disproportionally from the increased communication and enhanced awareness that a well-designed intranet can bring.

North Tyneside College provides the ultimate proof that good intranets depend more on will, talent, clarity of mission, and commitment to usability than it does on having lavish funding. This winning intranet, which supports 300 staff and 15,000 students, was designed and developed by a single person -- Adam Liptrot -- who might be the true hero of this year's design competition.

Good intranets are found all over the world. The thirty winners for the years 2001-2003 are distributed as follows (for multinational organizations, we count the headquarters location as the country):

  • United States: 57%
  • Canada: 10%
  • Europe: 27%, of which
    • Germany: 3%
    • Spain: 3%
    • Sweden: 7%
    • Switzerland: 7%
    • U.K.: 7%
  • Australia: 7%

We've had several close runners-up from Asia and Latin America, but no winners from those regions yet.

Mergers Can Produce Good Intranets

Three of the ten winning intranets this year were the outcome of corporate mergers. Mergers are typically very stressful for the employees of both the acquired and acquiring organizations, so it might seem initially puzzling that good intranets could result. Four explanations come to mind:

  • After a merger, it's obvious that something needs to be done to create a new, unified intranet. Thus, resources are often made available for a more or less clean-slate design that has the mandate to go beyond "business as usual" for the two previous intranets.
  • A merging intranet can draw on the best features and ideas of both pre-merger intranets . In contrast, the average company intranet team only has its own design as inspiration. (The need to expose intranet designers to additional good designs is the main reason we publish the Intranet Design Annuals.)
  • The newly merged organization will be in flux and have less established fiefs and ways that things must be done to satisfy entrenched political interests. Many of the most severe intranet usability problems stem from the intranet team's inability to overrule the narrow interests of individual departments.
  • The CEO of the merged company typically establishes an explicit one-company program to create a single corporate culture and encourage all employees to commit to the new organization. This executive directive provides a mandate for the intranet team in its attempts to create an integrated and consistent user experience.

We obviously don't recommend that you go through the upheavals of a corporate merger simply to get a better intranet, but it's worth trying to find ways to get similar benefits within your existing organization.

Logistically speaking, mergers can require a fairly rushed "Day One" design that can be rolled out the minute the merger takes effect, bringing employees up to speed on the new company's identity and policies. It's often good to have a two-stage strategy that explicitly recognizes the Day One intranet as an interim solution and leaves more time for the long-term intranet's user-centered design.

Workflow Support

Intranets are changing from being document repositories to being work support tools. Many of this year's winners had changed their information architecture (IA) from one determined by how documents are produced (usually the company's department structure) to one determined by users' tasks. But the work-support trend goes further than IA.

The client relationship management features on Landor's intranet are a good example of workflow support. The system tracks current projects and provides access to rich case history records and visual assets from past brand development. The system also tracks client correspondence, who works on what, and integrates with sales force applications to track potential future projects and wins and losses throughout the sales process.

Monitoring tools abound, letting intranet users keep track of important events at a glance. Examples range from Design Matters' green-yellow-red coding of project stages, to the Mayo Clinic's indicator of various facilities' bed availability, to the Coast Guard's indicator of current force protection conditions. The trend is to have the information come to the user as the task requires it, rather than have it sit somewhere waiting for users to track it down when they need it.

This trend is easy to describe, but hard to implement. To truly bring people what they need, when they need it, requires a deep understanding of their work. To deliver, intranet teams must analyze tasks much more extensively than usual. In the past, IA was the main concern of intranet design: it's not easy to develop a taxonomy for the millions of pages on a big intranet. While IA will remain an issue for intranet projects, the emphasis is likely to change to the more traditional human factors concerns of understanding work behavior, both on the individual level and on the workgroup and collaborative level. This change in design focus will drive an associated change in usability activities, with more use of methods like field studies.

Several projects explicitly set out to reduce the amount of email and supplant it in certain areas, such as attachment circulation. Email is a highly informal collaboration technology. Being unstructured has some advantages, but email usually turns into a big mess where nobody can find anything, especially later in a project. In contrast, many of this year's winning intranets have begun integrating explicit collaboration tools. Discussion groups are becoming common because they combine a degree of informality with permanence, searchability, and a bit of structure. Intranet discussion groups have a further advantage: they can be offered in the context of project areas or other targeted intranet sections.

Self-Service Content Management

Most of the winning intranets have recognized the importance of providing employees with simple tools for adding and maintaining content. Intranets live by content currency and contain large amounts of specialized information that originates in widely diverse departments and teams. The more people can take care of their own content creation, the more the content will be up-to-date.

Several winners built special wizards to support complex content management tasks, such as creating a new project's support area. In general, I'm not a great believer in wizards as a user interface mechanism because they force a very restricted interaction style on users and reduce the interface's flexibility and power. For complex tasks that are performed infrequently, however, wizards have their place. One winner originally had a wizard that took up to an hour to complete. After recognizing that this was too tedious, the designer redesigned a significantly shorter wizard. If you use a wizard, it's best to provide features that let you modify the results later and tweak advanced areas.

Most winning intranets chose a more application-oriented approach to day-to-day content management, such as adding a news item or letting employees update their directory listing. Wizards would be overkill for such tasks.

Companies varied in the approval processes associated with the widespread publishing that self-service technology supports. Some companies required explicit approval by designated managers or review boards. Others restricted the most important capabilities to special "power users" who could then assign more localized update rights to other users in their group.

Enforced or Guided Consistency

Many winners ensured a unified look and feel and a consistent user interface by using a content management system (CMS) that generated all the pages from templates. Others created detailed user interface standards that departments must follow in designing their intranet areas.

Consistency is still a challenge for a big intranet, but year after year, we do see a trend toward more unified designs.

Technology Platform

The only conclusion here is that there is no conclusion. Each of the winners used different platforms to build their intranets, and most teams felt a need to heavily customize or write their own software add-ons. Intranet technology is clearly not mature yet.

Half the winners used some variant of Microsoft technology as a major platform component. This is the first year we've seen such widespread use of Microsoft, which traditionally was not considered robust enough for enterprise-wide solutions. Simultaneous with this growing Microsoft influence, however, we saw an opposing trend: some winning projects abandoned proprietary servers and platforms in favor of open-source tools that provide added flexibility, lower cost, and perceived long-term stability relative to commercial products.

Several projects chose unified technology to support both their intranet and extranet, sometimes even adding the external website to the unified solution. Unification lets users update information once and have it reflected in multiple systems. Designers typically add filtering to ensure that internal information stays on the intranet, and only client-related information is published on a client's extranet site.

As intranets move from being a nicety (where users can look up vacation policy) to necessities (that run the business), reliability and uptime requirements increase. If you're preparing to bid a multi-multi-million dollar bridge contract, for example, you don't want to lose your access to the library of photos and engineering docs from past bridge-building projects. Furthermore, enhanced usability almost always leads to substantially increased intranet use. Thus, any project to improve an intranet should include a serious technical effort to strengthen the underlying hardware and software, both of which should be tested under intensive loads.

Design Process

On average, the winning projects spent twelve months on their intranet redesign. This is substantially faster than the two-year projects that dominated our earlier design annuals. Hopefully, this quicker turn-around indicates that intranets are becoming less of an ordeal to design.

Excluding projects resulting from mergers, the new intranet designs were launched an average of 35 months after the previous intranet designs. Thus, we basically see three years between major redesign projects, with the redesign itself taking one year. Obviously, the intranet teams are not idle during the intervening two years. All the winners used a continuous improvement strategy, with minor enhancements launched from time to time and a continued emphasis on converting hold-out department subsites to the standard corporate design.

The design projects' main usability methods were as follows:

User testing of new design prototypes 50%
Surveys (typically by email) 40%
User testing of the old intranet 30%
Heuristic evaluation (expert review) 30%
Card sorting 30%
Field studies 20%

The numbers add up to more than 100% because many teams combined multiple usability methods -- a strategy I recommend.

The teams used several other methods to engage users in the design process. One winner employed participatory design and determined the final design by letting all employees vote for their favorite from among a set of alternatives. Another winner placed a prototype design on a staging server and invited twenty volunteers to try it out and provide feedback. Note that such feedback is more credible when people actually use a functional design, rather than simply view screens and comment on them.

Several winners took special care to include users from international locations in their usability activities. This was typically done through some kind of remote technology such as WebEx; only one team conducted actual site visits to offices in multiple countries. International considerations range from making sure that pages download fast enough in Amadeus' Nigerian office to selecting navigation labels. For example, the term "white pages" didn't make sense as an employee directory description for many of ChevronTexaco's employees outside the US.

Improvement Metrics

There continues to be a paucity of detailed usability metrics for intranets. Most teams focus on doing a good job, not on justifying their existence. Overall, intranet use typically increased substantially after the winning redesigns, with growth rates between 47% and 1,500%.

More narrowly defined usage metrics increased dramatically as a result of usability improvements. For example, use of the Mayo Clinic's search engine grew by 700% after the team added better summaries to the results page. The Mayo intranet also made sending a pager message 20 seconds faster. Given how often paging is used at Mayo, this seemingly small savings added up to the equivalent of a full-time staff member in one location alone. Landor Associates decreased the time to collect and distribute sales wins, losses, and pursuit status from 30 days to 5 days, for a task performance gain of 500%.

Even though I'd like to see more ROI metrics, there is no doubt that good intranet design -- as evidenced by this year's winners -- can offer employees a much-improved user experience. Most intranets around the world are operating far beneath their potential. I thus believe that the intranet age is still to come in terms of seeing how work life will change once most companies invest in elegant and usable intranet designs.

Share this article: Twitter | LinkedIn | Google+ | Email