Design System for UTP's Online Courses

Some context

The UTP course virtualization (turning an offline, in person course into a remote course – I'm not talking about running OnlineCourseOS on some program in my laptop to run unsupported learning programs) team is mainly made up of Audiovisual Communicators (CAs, because this was technically written in Spanish first and in my day to day life I refer to them as such and changing the abbreviation just felt... wrong) and Learning Experience Designers (LXDs) who work in pairs, lead by team coordinators. Each pair is assigned 1-2 courses they must work on. That was a lot of parenthesis! Well anyways, this team I work with is in charge of designing the structure and learning material of online courses - most of which have been adapted from pre-existing online courses. This, as you might imagine, is kind of a complicated process that involves a lot of raw material that we then have to turn into something that is structured enough for students to keep pace with mostly on their own, but also flexible enough for the students to be able to use all the material at all times. It's a narrow line to walk, but it's fun.

Because of all this, course production is an intense process. We use (kind of, you know how it is) an agile methodology custom made for education (SAM, in case you're curious). And as I mentioned, one of the big outputs from the process itself is lots and lots of multimedia files – the learning resources. In order to get students to want to use them and study (with) them, we take care to make them eye-catching and engaging. Also, if we do our jobs right, students will likely be looking at these documents several times during a semester and we may as well make it an enjoyable experience for them. To do this right, we like using all sorts of creative visual and audiovisual elements that can capture the attention span of students.

However, the recent increase in demand for online courses and enormous proliferation of high quality, high detail learning resources, has forced us to increase the quality of these resources as well. This increased the workload the team had to handle and required a greater time investment on our part.

What I wanted

The main goal I had for this idea was to streamline the creation of learning resources that still encouraged reviewing, reading and studying with them. To this day I think this sounds super simple but boy is it hard. I had some...secondary requirements as well.

1. To allow for some creative flexibility for our team and give them some room to customize the material with their personal style.
2. To ensure that the final products created are still consistent with the design system established for the new UTP LMS, because the guidelines used to create resources prior to this moment adjusted to the visual style of now unused Canvas LMS.

What I did

First, I talked to some of the CAs on the team about what their main pain-points were while creating course material. For extra context, at this point, we asked each design duet to design a distinct visual style for each course they created before production really kicked in. Of the 5 team members I spoke with, the main issue they mentioned was precisely that: designing a visual style that can follow existing guidelines and still be unique and interesting, without compromising quality or design time. Yes in retrospect this seems extremely obvious, but back then this process was so messy we had to untangle many stages and requirements to even make that work.

Additionally, due to previous conversations I'd had with my team leaders, it had also become clear that we wanted to start standardizing a lot of workflows in the coming years, since our workload was going to grow very quickly and the pace of work we maintained was not gonna be able to keep up with that new ask.

Because of these two things, it became clear that the right solution for this problem was to create a design system of sorts – with new generalized workflows to facilitate the use of its components - which would actually be elements ranging from text boxes the team could use in multiple formats to entire resource templates.

Finally, I decided that the new system should also be accompanied by a full training plan for the team, since we would have to roll out the new design system quickly to meet our desired deadlines and not interrupt production.

The Design System

Building the system

As I mentioned, I chose to make the design style consistent with the pre-existing visual style created for UTP+Class, UTP's new LMS. Luckily, they have a fully fledged design system already! Ah, but, because that team works like normal designers tend to do and my team is a weird mishmash that creates deliverables in all kinds of different file formats and programs, their design system was not gonna produce the exact outputs we wanted. My team, by comparison to theirs, creates very few sofware or web-native products for our courses. Instead, we create:

This means that the design system must not only include elements that can be adapted to all of these resources, but that it must also consider that all components and elements must work or at least, be adaptable to, various pieces of software.

To create it, I reviewed pre-existing material and noted what types of visual components I saw most, and categorized them for each type of asset. The final list looked something like this:

Some resources had too much variability to warrant the creation of specific components for them. Several of these resources also take up a lot of work hours for the team, so instead of itemizing internal elements, I decided to simply create whole templates for those resources.

So finally, this turned into a to-do list of elements to design:

Before building the templates themselves, I also decided to stick to some of the basic design elements used in the rest of the university's online platforms. So I only used colours, typography styles and general elements from the existing design system.

Workflows

Along with the templates, I implemented new, cleaner and more efficient workflows for how to use them. I made this decision because the skills for using all the necessary software to use all the elements of the system were not equal in the team and it would be easier to train the junior memebers if we had a clear competency standard to aim for. Therefore, each template and element was accompanied with video tutorials, best practices and instructions on how to use them more efficiently. All of this was added to the team's Wiki.

Also, since flexibility was a big concern, a lot of the elements in the design system exist in different formats. This way, our team could create a beautiful reading using both Adobe InDesign and Microsoft Word, and the result should be of equally high quality.

Delivering the system to the team

To deliver the design system to the team, the Wiki was completely updated to show all the changes and I planned a short cycle of training sessions to learn the new system step-by-step. All of this was done only a couple of weeks before the start of the new production process.

Q.A

You know how earlier I mentioned some of the members of the production team? Yeah, I forgot to mention the Q.A guys. They exist, and they keep a checklist they use to review every piece that comes out of the production team. As the design guidelines changed radically, I worked with them to adjust the criteria they use to measure asset quality. Due to the standardization of the visual style through the design system, this actually really simplified the previous QA process and reduced the items and the ambiguity on the quality checklist without really compromising the end result. The previous checklist had 32 items, and the new one has 25. The review process, therefore, was streamlined along with the work. Also, the results and stats from the Q.A team started being a lot more useful since we could use it to pinpoint problems the team had adapting to new guidelines.

Finally... results

The results of the implementation of the design system have been very positive. The team can generate resources 5-28% faster, which is a huge variation but it sort of makes sense considering production time is heavily dependant on the format. The biggest time savings ocurred in the production of readings and infographics. Team satisfaction has also been high - but some internal suggestions have emerged, along the lines of:

• We should be determining how much variability is appropriate within a template and how do we ensure that the QA process does not flag it as an issue.
• Some of the templates should be tweaked to organize content to improve student experience.
• We should consider more alternative formats for some templates and components, to speed up the production of particular resources in some of our less used programs. Personally, I would get rid of those programs but they do have some specific niche uses for us.

These and other requests or suggestions have already been taken into account and will be implemented as modifications or additions to the system eventually.