Masterclass: Part III

Exploring intranet usability and redesign

In the final instalment of this masterclass on intranet usability and redesign, Stephen Musselwhite explores project ownership and usability improvements

Who is responsible for improving the usability of the intranet ecosystem?
The answer is simple – all of the teams involved with the intranet are responsible. The intranet manager should take charge for improving the usability of the intranet ecosystem – inputs and outputs – by ensuring everyone involved in the process is aligned around a philosophy of usability. Effort must be focused on shifting the attention placed upon usability away from the tactical, to the strategic, and to move from a reactive to a truly proactive consideration of usability.

Everyone has a responsibility for usability
In reality, few intranets will have dedicated resource for every constituent component. The central intranet team is likely to be a virtual team, with a variety of disparate skills and associated responsibilities spread across the organisation. This introduces challenges in the management of the inputs into the intranet, and the quality of the resulting outputs. The team that manages the hardware will doubtless have a portfolio of other platforms and servers to support, of which the intranet may not be viewed as the most important. The hardware support team must be involved in discussions and planning around the role of the intranet, because they play a critical role in ensuring the stability and availability of the system.
The developers must ensure that the templates they release for use by the content managers are usable and encourage consistently high-value contribution. The performance of the technology should entirely support a robust and comprehensive content management process. Small usability issues in the templates can lead to impaired content manager efficiency, and a greater likelihood that corners will be cut in the rush to get content to the end user.
Figure 1 considers how the responsibilities for strategy formulation, design and usability should be divided between the relevant teams.

Usability funnel – fast-tracking effective designs
Applying a variety of techniques to improve the usability of an intranet is essential. When used in the appropriate combination, a number of techniques used together can actually help to shortcut the design iteration process, and speed the development of an improved user experience.
A clear usability strategy that defines the process for creating usable designs (relevant, simple and consistent), coupled with the appropriate blend of heuristics and design patterns will result in the rapid provision of usable designs.
Once wireframes and prototypes have been constructed, it is time to conduct usability-testing on the new design to ensure it is an improvement on the existing intranet.
Split-testing is a particularly useful usability-evaluation technique to assess the impact of the improved design. A range of tasks is conducted on the existing design, and the new prototype (with the same metrics) is measured for each design. Split-testing is a useful technique to ensure usability improvements are benchmarked and regression-testing is considered as part of the redesign process.
The wireframe and prototype stage of the assessment of usability allows not just for the incremental improvement of usability, but also the innovative improvement of usability.

Innovative usability improvements
Innovation has a place in the redevelopment of any intranet. Innovative improvements to a usability problem should, of course, remain outcome-oriented and user-focused. Innovative improvements will only be possible if developers and designers are able to focus on addressing the need, and not simply improving a usability issue.
Consider the following scenario: the first wireframe (Figure 3) shows a proposed design for a standard content page. The wireframe has been based on prevailing best practice, and is an attempt to improve the clarity of the display of content based on the incremental improvement of the existing intranet page.
By focusing, instead, on the problem (users struggling to find content that relates to certain activities and the individual contact responsible for that content), and not on the usability issue (the page does not contain a clear introduction-section), innovative improvements are possible. By providing a right-margin that provides a link to search, and exposes keywords, related content, as well as author and content manager details, the designers have addressed a number of usability issues concurrently.
Once a prototype has been finalised, the template-build can begin, same as the task of improving the quality of content can begin.

Improving content and information architecture – designing for information scent
The redesign of an intranet should place a significant emphasis on the improvement of the quality of the content. The components of good content usability are introduced here. In Designing for the scent of information, Spool et al introduce the concept of trigger words, to explain how users typically start their search for information:
“We’ve noticed that people looking for information all exhibit similar patterns. They first scan for trigger words – words or phrases they associate with the content they’re seeking – in an attempt to pick up the scent.”

Information architecture
Move from generic to specific (minimise the number of high level categories). Use clearly descriptive categories.

Placement of information
Users of intranet sites will pay more attention to content at the top left of the page (assuming they are viewing content displayed for reading left to right). Place the most used information and functionality in a visible and consistent position, above the ‘fold’ of the page.

Page length
Users will scroll if they believe there is a benefit in doing so. Consistency helps considerably in enabling the user to predict the value of scrolling a page. The consistent placement of page ownership information and associated metadata at the base of the page allows users to quickly identify the owner of the content.

Link-text
Link text should clearly describe the destination of the link to ensure a user’s confidence in locating the information is maintained. Naming applications with obscure names destroys information scent and frustrates the user. As Spool et al point out, “clarity always trumps fun”.

Link length
Because link titles must reflect the title of their destination, link length will be dictated by the length of the title of the page they direct the user to.
During their research, Spool et al have “repeatedly found that longer links and pages make for better scent”. The research concludes that “links seven to twelve words long are more likely to take users where they want to go”.
That is not to say, of course, that shorter links cannot be effective, but it does again illustrate the importance of clearly descriptive links. Longer links are more likely to contain those all important trigger words, so longer links tend to be more effective.

Search or browse?
While some users may have a tendency to prefer browsing or searching for information, it is likely that the technique they choose will be heavily dependent upon the information architecture of the site. Intranet users typically use a combination of browsing and searching to fulfil their information needs.
The way users navigate around an intranet will be influenced by a range of factors including the following:

Their familiarity with the site;
Their familiarity with the subject matter;
Their assumptions about how easy the information will be to find; and,
Their perception about the size of the information space.

Focusing on the efficiency of the design and management process
Inefficiency in the management of an intranet can be introduced at any stage in the process. Consider the two following approaches in Figures 4 and 5 to the creation of intranet sub-sites. For example, in the first approach, requests for a new sub-site is met with the design of a new template to accommodate the unique demands of the request. In the second approach, the unique template requires bespoke training material and support in addition to unique usability testing. Following the launch of the new site, there will be an inevitable requirement for bespoke maintenance and support.
The process shown in Figure 6 demonstrates the efficiency and effectiveness of pursuing a consistent approach to template development and deployment.

Migrating content
The migration of content from an existing intranet to its replacement can be a significant challenge. If the content is well structured and clearly separated from its presentation and the new version of the intranet is built using the same content management technology, the option to automatically migrate content will likely be the preferred approach.
Where content quality is variable (perhaps as a result of deficiencies in the content-management process and technology, or insufficient content manager training and support), it may be necessary to manually migrate the content.
Although the automated approach to migrating content into a new intranet is clearly preferable, this approach will only be possible with well managed content and information architecture.
Neither approach is without its challenges. The process of manual migration need not be as complex and time-consuming as it may appear, when the exercise is carefully planned and focused around the most used content. The first stage in a manual migration of content should be the careful analysis of metrics and measurements to reveal the most used content on the site and the various sub-sites. This exercise will reveal the priority areas for migration. Intranet content is only valuable at the point of use, therefore content that is never used offers no value. Intranets need not be vast repositories of information, they should be repositories of used information. Focus should be applied, therefore, on ensuring the intranet only contains the information employees need.
Lean intranets are more efficient to manage, and effective at meeting user needs than vast intranets.

Launching the redesigned intranet
Before a redesigned intranet is launched, a decision must be made as to how the launch will be positioned and promoted. The methods available are numerous, but will essentially result in a ‘soft’ launch – where changes are not promoted, and a ‘hard’ launch – where users are notified of the change prior to its arrival.

The ‘soft’ launch
The soft approach to launching an intranet is typically taken if the changes are minimal, and where previous changes have not been successful. If the changes are largely cosmetic in nature (rebranding or changes to the interface to support a change in internal visual identity, for example), the decision may be made to introduce the launch with no significant prior promotion to the organisation. If previous intranet redesigns have failed to arrive on time, or to general approval, it is sometimes safer to proceed cautiously with regard to the promotion of the new intranet, to avoid missing deadlines that have been promised to the organisation at large.

The ‘hard’ launch
When significant change to the intranet is planned (the introduction of personalisation options, for example) and users may need some prompting and assistance in optimising their use of new features, a ‘hard’ launch will be required. Hard launches are best suited to launches that have a very high probability of launching on time after an appropriate period of user-testing to ensure usability problems have been removed. The objectives of a hard launch will be:

A reduction in support (dealing with calls relating to new functionality and interface) required from the technical service desk; and
Create an interest in the intranet redesign and its core features, and to encourage users to make use of those features (which, of course, they will only do if these features are relevant to their needs – but creating awareness of new features in advance assists in ensuring adoption is more rapid).

The danger of the hard launch approach is that, if a widely-publicised deadline is missed, the credibility of the intranet team may be damaged.

The (perpetual) ‘beta’
An alternative approach to launching new functional modules, or even whole websites that are growing in popularity across the web, is the beta release. This approach essentially involves the introduction of new features for use in the live environment, without officially releasing those features as production features. This is largely an exercise in positioning, of course, but positioning the redesign as a beta has a number of advantages. First, the approach is quite engaging and user-focused, if managed with care. The active soliciting of feedback from users gives them the opportunity to be involved in improving the design. Any effort to engage the user is worth pursuing.
Second, the expectations of users are managed quite discretely. The beta label is associated with the consultative introduction of advanced Web 2.0 features on the web, so users will infer that the development is cutting edge. Users are aware that there may be issues with the features, and changes to improve issues will be implemented responsively.
The main problem with this approach is the risk that the beta status of the release will remain in place almost indefinitely. Too much of this will challenge the credibility of the intranet, so it is better to identify individual components of the intranet as ‘beta’ than the entire intranet.
A beta release should not be considered as an approach to usability-testing, and introducing new functionality as a ‘beta’ release is, of course, no excuse for failing to conduct usability testing. A ‘beta’ release should only be launched after exhaustive usability-testing, and should ideally be separate from the existing site (with the older and more familiar approach available to users). One approach to launching an entirely new intranet is to run it for a short period, in conjunction with the existing site, to allow for the comparison of layout and features by the user. This approach can effectively be used to trial a single sub-site section

After the launch
The first few weeks immediately following the launch of the intranet are crucial in establishing the success of the new or redesigned intranet. Following a substantial redesign that has affected the underlying content-management processes and not just the design of the site, additional support resource will be required not just for the end user, but also for the content management community.

Early-life support
The provision of additional resource to support the implementation of many major systems-based projects, is frequently referred to as early-life support. Providing early-life support for an intranet involves informing technical support or helpdesk about the changes well in advance, and involving them in the usability-evaluation process. The technical support team will be the first line of support for users if they struggle using any element of the new intranet. The technical support team could also be in a position to support content managers and content authors.

Gathering feedback
A clear mechanism for sending feedback to the intranet team must be available. Intranets will generate unsolicited feedback throughout their lifecycle, but a peak of unsolicited user-feedback should be expected immediately, following any substantial change to the intranet. All feedback relating to usability problems should be acted upon quickly.

In summary
This series has explored the reasons behind a range of common intranet usability problems, and strongly advocates a holistic approach to addressing them. It has reinforced the view that intranet redesign will be more successful if the motivations for redesign are driven and supported by the strategy, senior sponsors and governance structure. If the redesign objectives are not aligned to the intranet strategy, the redesign will struggle to achieve anything, other than temporarily improving the aesthetics of the intranet. Failing to tackle the fundamental causes of the usability issues ensures those problems will return.
The importance of the separation of roles and responsibilities has been discussed and approaches to improve the efficiency of the introduction of usable designs have been advanced.
It has also underlined the importance of the quality of content available through the intranet, and the dimensions against which content quality should be judged.

This article is an excerpt from Ark Group’s Intranet Usability and Redesign report, by Stephen Musselwhite. For more information contact Robyn Macé at rmace@ark-group.com