Giant intranet redesign anyone? Don’t run away just yet, hopefully this retrospective of the process we went through will provide some useful takeaways.
Discovering the task and problems to solve
Getting to grips with the size of the task and prioritising organisational goals was the first step, collating surveys from a range of stakeholders with precollected user frustrations. Simply there was a huge amount of content (in multiple languages) that had grown over the years that had to be better discovered within a secure platform.
My vision is to help others develop their content areas in ways that best serve the users.
Top issues which became key requests:
- Clear navigation, categorisation and search
- Consistent site behaviour
- Flexibility in structuring and featuring content
- Be able to feature content easily
- Able to set content expiry
- Avoid broken links
Defining the process
We decided the best way of working as a team would be using agile methodology with weekly meetings and task setting. Each month we would demo progress for stakeholders outside the team. This allowed us to be flexible in planning, but still deliver chunks in set milestones. Within this framework I could lead the design process at each point.
It was also decided to move the site over from Drupal to Liferay for the functionality we needed and improved security. There was a lot of initial knowledge sharing and learning about the platform required, helping us plan for development.
Getting a wealth of user input
Where the success of the intranet will live and die on colleagues using it, get their input. To avoid the risk of building things that colleagues don’t know the benefits of, we engaged with them from the beginning of (and throughout) the project to learn needs and desires and how those can be served.
We set up an initial focus group session with the aim to find out how different types of colleagues use the site and how they think when it comes to the content. Starting with a questionnaire we began to see how similar, or different, the needs of the group were.
- Those working deskless and in remote places need curated info or quick access to current topics
- People new to the site need to know what is on the site, again with curation
- Those from headquarters discover and share info for others
We selected a broad range of popular content articles and asked the group to discuss and group them as they saw fit using the card sorting method. Through the exercise they managed to create sensible content groups, some easier than others. We prioritised the importance of each group which overall gave a great impression of how they saw the site content.
We then conducted a blind card-recall to test where they would find certain content articles, which resulted very well from the previous work.
A key learning was that the focus group didn’t reference current naming conventions suggesting this needed to be re-thought from their perspective.
A site reaction test of a broad range of content sites gave a clear idea of how dense the amount of information should be and how they expected the site to look.
Aside the focus group we also created an optimal cart sort online asking participants to group 30 example content cards to test and gain a wider perspective. This resulted in patterns emerging when grouping content and being able to standardise naming the groups.
The grand audit
Early on in the process it was key to comb through all the content; not a task for the faint hearted. We set about undertaking a content audit. Luckily Drupal could output a number of key attributes, but mostly this was fun documenting in Excel.
We captured several key attributes for each content piece. Aiming to find issues which helped us make decisions on the content, these were the kind of questions we wanted to know:
- Is it a popular page, where does it rank?
- If it has links to other pages, are they broken?
- Does it have a meaningful description attached?
- Is it in the wrong place?
- Does it have an obvious topic?
- Is it inaccurate? Is there any key information that’s missing?
- Is the content duplicated anywhere else?
- Is it useful? Does it add value? Has it naturally expired?
- What type of content is it (eg. landing page, blog, etc)?
- What language is it written in?
- Does it need language/punctuation improvements?
- Does the tone and style meet the guidelines (something we had started to define)?
- Could the content be supported by an image or illustration? If yes, what?
Some key outcomes from the audit were reducing the amount of content greatly, weeding out duplicate, expired and content that could be combined. We were able to see region specific specific content and see what was most valued.
Defining a tone
We wanted to define a language and tone to bring coherence to the site. Championing site values included a set style of writing and consistent language and punctuation rules. One to mention by way of example would be consistent date formatting, especially important for an international platform.
Helping people find what they are looking for
Next we attempted to define the conceptual organisation of content through information architecture. This involved looking in depth at the taxonomy of the site and taking on board what we’d learnt to group content as it would be on the site.
We needed to test this out again on the different colleague groups, and set up tree jack exercises online to test out our suggest groupings. This helped confirm some things and refined our hierarchy of topics.
Designing the system
Whilst designing to meet the needs discussed it also paid to work to Liferay’s strengths. We were able to design the site site in a modular way (Liferay calls these asset portlets) with a flexible structure. We could then make use of page templates to reuse and create consistency.
Quickly design patterns emerged and were defined, some select examples:
- Landing pages, needed to house a topic or chunk of content on the site. This particular template had to be split into two variations; one to list topics with their description and two to list the articles contained. This due to the shear amount of content in some high level topics and content overload.
- Related modules to surface content organically
- Playlists of related media that previously had been accessible by clicking through several pages
- Sharing content, both within and outside the site. When content could be shared outside the intranet it was clearly labelled so.
- Faceted search, allowing great control over lists of articles and general search.
Each pattern needed to be able to adapt for viewing on smaller screens in a responsive way, which helped define the design. For the first time, people would be able to view content on their devices, a key requirement for those working remotely.
A living style guide
An important document for the team was the style guide created, both as a development aid and a reference for the whole team. The guide detailed all the patterns we used to create the site, grid and page division (based on a customised bootstrap framework), typography, buttons, colour, images, icons and key interactions.
Review and launch
In the early part of the project design tasks were very much in parallel with the wider team and towards launch we conducted design reviews of the development. This involved more testing with those who would be using the site, testing both mock-up prototypes and later the working site. Following this we were able to review in detail and prioritise any loose ends or usability issues prior to launching.
As part of the launch several help pages and videos were created to help people get to grips with the new site. This helped with a smooth transition and make people feel comfortable.
How did we do?
It was key to stick to solving the key problems faced at the start of the project, and everything was answered in that regard. The huge task of organising all the content has provided colleagues with several ways to find the content they are looking for, aided by featuring most requested topics and content clearly. The voices of the different colleague user groups were present throughout the process. As a designer, this was critical to finding the direction.
Now that’s a site I’m going to be able to use!
This was no mean task, the team working for over a year to full launch. A lot of things were learnt along the way for all, through a design-centred approach, and there are still some things we’re excited about introducing on the platform as continued improvements.
Client unnamed due to sensitive nature of the organisation.