Story of an Intranet Redesign - Part Two: The Users

As we left off in Part One we had the structure and IA of the site bought of on by our various stakeholders and review boards.  At this point we decided that we needed more input from our users before we moved any further ahead.  Our first step was to gather some more basic information about how people used the current Intranet and try and get some details on what our users liked, disliked, needed and wanted.
 
We started by re-investigation of our logs and asking ourselves tough questions about what was important to our users as a whole as opposed to what was important to a particular segment (i.e.. department) of our users.  We got ClickTracks to try and figure out how it was people were finding different areas of the site, and how our navigation was being used.  Most of this data reinforced what we already knew.  People we’re having a hard time getting around the site and with a few exceptions there were huge gaps where information just wasn’t being accessed.
 
It was fairly hard to judge which parts of the sites were hard to locate vs. parts of the site that were simply of little use to most of our users.  Some were obvious, but if we were to try and pair the site down and eliminate, or rather, move back certain unused parts of the site we needed more information.  Realizing that we began to formulate plans to touch base with our users.
 
We started off with a survey that was posted on the current Intranet homepage as well as sent out in global emails.  It was in essence a call to action to help us make the new Intranet better and more useful by submitting feedback.  The first few questions were to find out how many of our users fit into our two main audience areas, based on the card sorting work we had done, administrative or clinical.  Then we asked some general questions about how important the Intranet was to their job, what their main uses for the Intranet were, how often they visited and what their overall experience was.  We also asked some specific questions about what they liked and disliked, what they’d like to see added or removed and what they saw as the biggest issues with the site.  At the end we asked for volunteers to participate in focus groups and usability studies.  There were 500 responses and 80 volunteers — a lot more that I’d expected.  It was obvious that our users are a highly engaged bunch and they really cared about what were were trying to accomplish.
 
We analyzed those survey responses as a whole and as two groups, clinical and administrative.  We then took those results and spent a long time mulling them over.  It seemed that most were ok with the site as a whole but that certain aspects needed improvement.  There was nothing too surprising, our users were telling us that they were having problems finding things and that they wanted more information and easier ways to search and browse the site.  We knew that at the least we needed to improve the phone directory search and the overall navigation, implement a search engine and figure out some new ways to get more quality information out to our users more often. 
 
The survey turned out to be a huge success and really helped us make the right decisions moving forward.  As a side note, we sent out another survey a bit later on in the process that focused on the information/communication piece of this. We asked questions about how our users got information about hospital news and events and how useful that news was, as well as some specific questions about our Intranet news channels.  I’ll most likely get back to that in Part Three when I talk about our channel development and implementation of Movable Type.  A little teaser there for y’all.
 
Back to the matter at hand.  Ok, so once we had all this great data about what our users wanted and thought of the site, we took the second big survey win, the 80 volunteers and began to work up some ideas about how to use them.  We’re a small team, working on a tight budget with little time.  We needed to get into the nitty gritty of designing and implementing this stuff very quickly.  We decided the best way to maximize our time and get some of the hands on data we needed was to go out to the users and talk to them one-on-one.
 
At this point we had begun to work up some wire frame ideas and some basic design mockups.  We threw together a functional prototype with a fairly loose look and feel that we could use as a sounding board for both our users and the various stakeholders around the hospital.  We decided that we could use this prototype as a basis for some basic user testing.
 
I came up with a quick script that included some general feedback and interview type questions (view or download an Word doc example), coupled with some task based actions to help us test our wire frame and the findability of our IA.  As a team we made appointments with our volunteers and spent a few days at their desks conducting interviews and (gorilla) usability tests. As you can imagine, there were some real eye openers during these sessions, and if I could do it over I’d have done more of them and done them earlier.  Having said that, there wasn’t anything we couldn’t handle, and the design and IA we still in a rough enough stage that we could easily change them.
 
We all got back together as a team and went over all the usability data we had and compared that to the functional prototype.  We made a list of issues, from things that were very important and needed to be taken care of to things that were mere suggestions and user comments.  We tossed around the idea of working up some personas based on this data, but based on the workload decided that we would have to wait for that if we did it at all.
 
Having spent some quality time with the users, and feeling like we had a good idea of what we needed to do to make the site work best for them we started to work on a final design.  I’ll not get into this much except to say we went for something very straight forward, as we needed something that would fit the look of the majority of the existing pages.  We used a rather plain, “Amazonish” tab structure along the top for our home page and main sections, with a left hand nav we dubbed “look inside” to drill down into the content.  Each section has a color associated with it and a few of the main information channels an icon.
 
The front page is a fairly text heavy three column layout, with the “look inside” on the left along with a search area, a weekly “feature” story with a photograph top center, links to announcements, press releases and features below that and events listings and special employee offers along the right.  Here is a screenshot of the final design to give you an idea.
 
Each of the main content sections, which, at this point we are calling channels, has the same layout with the exception of the center column which contains news, announcements and features for that particular channel.  For instance, if a department brings on a new director they could place an announcement on the “Department Channel”.  This allowed us not only to provide more quality information to our users but lighten the load on the numerous small updates content owners were wanting to make.
 
The design evolved a bit as time went and and we made various implementation decisions, but we kept true to the basic prototype throughout.  So, at this point we’d done our due diligence, talked to our users, gathered the data, made some decisions and decided on the design.  It’s time to get down to business and get the site built.  But more about that in Part Three.