The Rockwell Automation Technical Communications Team presented their SharePoint redesign project at the Cleveland User Experience Professionals Association (Cleveland-UXPA) meeting on September 26th, 2013.
AmitUser experience and Information Architecture is practiced in two different parts of the organization here at Rockwell Automation. As part of the Technical Communications team, we architect, develop and deliver technical content that describes our products and solutions. As the Usability and IA team, we focus on usability, findability and organization of the content that make up our deliverables, regardless of medium (print, web or both).The CVB UX team focuses on the usability and user experience of our flagship software products. Although we are located in different parts of the organization, our core goals and objectives are similar as in we are working towards making our products and related technical information easy to use.
Amit: In most cases, designers and users are not part of the same population. Users typically do not design the tools they work with, and designers typically do not use the tools they design outside of a beta testing environment.These principles work for more complex and extensive domains like banking, finance, ecommerce, healthcare etc.In other domains like open source software – Linux, Apache, Perl etc. the designers are the same as users. Designed by geeks for geeksWe were designing something that only the TechComm group will ever use. None of the other groups within Rockwell that use SharePoint are expected to know and use our site.Our designers are core members of the larger audience
Amit: Basic User-centered design methodology helped us break down our project into 4 stages, shown here.
Sharon:Since the initial launch of the TechCommSharepoint site back in 2008, we had been getting an ever-increasing amount of feedback letting the administrators know that information was difficult to find and organize.There were a number of reasons for this, including:Initial site was designed by a team of 5 without taking a survey of wider TechComm populationSite grew and added information and componentsAdditional TechComm teams from other regions began to use the site and were used to organizing information differentlyManagers started using the site for tracking and reporting functions that were not initially planned forOur new plan, now that we had far more experience using SharePoint, was to make sure that we redesigned the site for the way it was actually being used, instead of the way we had theoretically imagined the TechComm community would use it.As part of our redesign project, we were offered the consultation of usability people within Rockwell, who helped us formulate a methodology for determining user behavior and some data analysis techniquesSince the initial launch of the TechCommSharepoint site back in 2008, we had been getting an ever-increasing amount of feedback letting the administrators know that information was difficult to find and organize.There were a number of reasons for this, including:Initial site was designed by a team of 5 without taking a survey of wider TechComm population – no previous SP XP to draw onUser base and needs evolvedSite grew and added information and componentsAdditional TechComm teams from other regions began to use the site and were used to organizing information differentlyManagers started using the site for tracking and reporting functions that were not initially planned for
Amit: Since the design team was made up of SharePoint users, we didn’t need to do user research to the extent that an external design team would have. We actually knew all of the domestic and many of the international users personally, and we all work together in the same job functions. The design team had members from each US-based TechComm group in it, to ensure that any group-based idiosyncrasies were accounted for.
Sharon:To make sure we started off addressing the issues of greatest import to the TechComm teams, we created a survey and sent it to the users to see what they wanted and how often they used the site. Over 66% of users used the site more than once a week.
Sharon:Based on a preliminary cardsort we did, the four most important items on our site are shown here. The word cloud illustrates the other responses we got. The responses, when compared to the previous slide, show that these top 4 items in both cases were identical. We felt that this was a positive response, because we then knew what our most critical items were in terms of findablity and usability.
Sharon:According to our survey, the four most important items on our site are shown here. The word cloud illustrates the other responses we got. The responses, when compared to the previous slide, show that these top 4 items in both cases were identical. We felt that this was a positive response, because we then knew what our most critical items were in terms of findablity and usability.
Sharon:We conducted external user research in order to define the who, what, where, why characteristics of our Tech Writers. Here’s what we foundFamily thinks writers carve hieroglyphs and symbols on stone tabletsSociety thinks writers create documents with clippie from Microsoft Office circa Windows 95Recruiters think writers are people who know software packages and that’s what makes us writersSoftware developers or Engineers think writers slack off on the job most of the dayTech writers know that in order to do our work, we must make sense of complex concepts and link them in a way that makes them easy to understandOn average, we actually spend about 20% of our time writing and 80% of our time doing everything else. This everything else could include attending meetings, document design and layout, content management, project management, and lab work and testingA well designed SharePoint will help us do some of the our more mundane activities more effectively ultimately providing more time for direct project-related work.
Sharon:Because the designers are also users, we did not need to do extensive user research and make detailed personas. But what we did do was compile lists of important job-related tasks for each group and look for ways we could accommodate them.
Sharon:Information Developers, our heaviest user demographic, use the SharePoint site as a project tracking and management resource, as well as a home base for collaborative projects with other Developers.
Sharon:Managers, however, use the site primarily for project management, resource tracking and planning.
Sharon:Because we are not usability experts, we consulted with ours to determine the best way of surveying our users about whether the new site design was effective.Based on past experience with cardsorting methods and software, Amit suggested we use OptimalSort, which, in addition to being easy for our survey group to use, offered partial automation of the data analysis.Because it was online, it worked well for our audience, which is spread out over multiple states and three countries.We had 35 people complete the card sort exercise, which was approximately 85% of our total user base.
Sharon:OptimalSort returned far more data from our exercise than we had expected. We could use spreadsheet downloads, dendograms, similarity groupings, percentage correlations, and there was enough there to keep us busy for years while we figured out the best way to look at the data.
Sharon:In the end, we decided to use the PCA groupings that OptimalSort created using an internal algorithm. This gave us a common denominator to start from and we determined that if we could get the most important types of data sorted, we would then be able to slot the outliers in to sensible categories.
Sharon:For our final information architecture, we used as simple spreadsheet so we could verifythe categories and associated content.We used the data that we had collected from the cardsort and the 80/20 rule to develop our preliminary groupings.21/35 participant responses were closely aligned with this baseline, based on card sort data
John: In one sense, the design stage was fairly easy, because SP provides a standard set of out-of-the-box components that provide the underlying structure of the site. SharePoint can be customized, but we’ve been advised by our IT department to use standard features for the sake of making any future upgrades and/or migrations easier and cleaner.As this slide depicts, at the basic site level there is a standard page header, a dual purpose pane for browse mode or work edit mode. When in browse mode breadcrumbs are displayed in this area. When in Edit mode, a tools ribbon is displayed. There is also a navigation pane at the left side and a main page. These parts of the site structure are imposed.One feature that we did not use on the original site was the tabs, which appear just below the browse/ribbon. In most cases at Rockwell Automation, SharePoint has a main business site, with many subsites below it. It those scenarios, the tabs are used to access the subsites. But, since our site has been and is a stand-alone site, we learned that we could utilize the tabs to display individual pages on the site, which further helped us to build our architecture, which I will explain on the next slide.
John: Here is our final design. We added links to the #1 , #2 and #4 most accessed/used pieces of content in the Nav. Pane, which is referred to as the Quick Launch.One item that may seem out of place, as far as most accessed content, is the Collaboration Site list (#2), which appears on the lower, right side of the Home page. But, we opted to leave the list where it was originally, because users were familiar with the location (nothing changed).The tabs in this view show the nine main categories of grouped content that were developed in our IA blueprint. Each of these categories reflects a common high level writer or manager task.
John: This is one of the tabs that focuses on the Publish task.
John: Prior to finalizing our design, we validated the site concept with our management. We were able to built a prototype site at a new URL (obtained by our IT department) so we could keep the existing site up while we worked on the new site.Once we had enough of the prototype site built, we administered a self-guided usability test, with Amit’s guidance.The testing focused on 13 scenarios or tasks and included 10 users – 8 frequent site users and 2 infrequent site users.We did opt to provide a subset of the scenarios/tasks to our infrequent users, because some of the tasks were not realistic for their role.We asked each user to locate the information on the site, track the number of clicks it took to get to the correct information and to rate how easy they felt it was to find that information.
John: (Amit, do you think this slide is necessary? I think it could be dropped from the deck, if you agree.)
John: 58 responses received the highest rating score of 10 - equated to 1 to 2 “clicks”Only 4 responses received the lowest rating score of 1 - came from the need for “many clicks” or “gave up” (expense report)Out of 105 scored responses, the average rating was 8.6Achieved usability goal 86%!Reviewed low scores and feel that with familiarity, and some site tweaks, these responses can be addressed
John: (Amit, if you do not think this data is relevant to usability, I would be fine with dropping it from the deck.)
John: As I mentioned earlier we were able to implement our new design while maintaining our existing site. Users needed to continue to manage ongoing project tracking tasks without duplicating data or being locked out of the site for extended periods of time.Since we were creating a new site, we did not want to recreate or duplicate content, if possible. Therefore, we completed a site inventory of all of our content so that we would know if it was successfully migrated to the new site before we went live.We had to work closely with IT in order to schedule and begin a migration of the content. At the same time we had to prepare our users for the migration by providing them time lines and any site black-out dates and the go live date.
John: The key design elements that we implemented include:A less cluttered Home page that offered items of general department interestUtilizing the Quick Launch (nav. Pane) to provide links to the most sought after contentOrganizing task focused content on the tabs. The tabs work nicely in that they reflect the ribbon in Microsoft products and are a common organizational structure for all users.
John: Our biggest challenge was receiving a very untimely notification from IT that our site, actually, our sites, would be migrated to a new version of SharePoint - 2010.I say untimely because we had completed our usability testing and were already building our new site, and, content migration had already begun.The SP 2010 migration would include a two day site content freeze and one day of site verification by the site administrators.We stopped what we were doing and awaited the migration, with the plan to continue once it was done.To make things worse, we had communicated all of these confusing details to our users and then the SP 2010 migration was delayed!So, we jumped seized the day and actually determined that we could rebuild our existing site, performing much of the work on the new structure behind the scenes and rolled out the re-designed site before the IT migration.By doing so, we benefited from not having to migrate any content to the new site or performing all the associated tasks, we maintained our current URL and saved a least of month or two of delay on rolling out the new site.
SharePoint Site Redesign : Information Architecture and User-centered Design at Work