Getting Started With Data-driven Learning Design
As a whole, the L&D industry hasn’t always been doing a terribly good job when it comes to designing learning. However, we have started to recognise that one-size-fits-all activities are probably not the way to go, and that we should design learning for the people doing the actual jobs, not for the company HR department. Fundamentally, designing better learning is about knowing your learners. In that aspect, the overall capabilities of the industry have developed tremendously over the past few years (with things like xAPI etc). However, as we start to accumulate more data and information, it’s important to know how to use it well. Thus, we decided to look at data-driven learning design, how to get started and the different types of data you can use in design decisions. We’ll divide this article into two, resembling an initial- and a subsequent round of design.
Understanding who you are designing for
At the start of any design process, you should always spend time understanding the problem and the “customers”. In corporate learning, this discovery is equally important, yet something that many organisations skip almost entirely. Here’s where data-driven learning design approaches already come in handy, albeit not perhaps in the way you expect.
Since it’s your people and employees you are designing for, you have an abundance of data available to you. However, this data is not necessarily siloed within the L&D’s systems or records. Rather, you might have to look for it in other places. For instance, demographic data might sit in an HR system. Assignment and task related data might sit in a performance management system. These kinds of data can help you create rough archetypes, or “personas” of your learners, i.e. who they are, what they do etc.
However, if we leave it there, we might still miss the mark. At the initial design stage, we should also explore how our learners can engage with the learning content at the workplace. As we don’t want to inconvenience them, it’s important to get to know the workflows and they ways we could instil learning into them. Now this a part of data-driven learning design that you don’t have an easy tool or a dashboard for. Rather, you have to get out there, start observing and exploring, and collect qualitative data. Different service design methods prove quite effective in this regard.
Understanding how learners engage with the content
Unfortunately, once you’ve put a learning activity together, your job doesn’t end there. Although the initial time spent on learning design does pay off, it’s still unlikely that everything works perfectly. Maybe there are pieces of content that the learners don’t engage with. Maybe they engage in ways different to what you initially thought. Whatever the actual usage and engagement behaviour is, it’s our job to find out.
To start out, tools like web analytics can provide handy insights into e.g. engagement times, devices used and geographical locations. Then, more specific tools for learning content analytics can tell us stories about how the content is being consumed. Finally, it’s tools like xAPI that enable us to practically follow the learners’ journeys through the material, tracking and seeing every interaction along the way.
Once we know what’s not working, we can fix it. Maybe we need to cater to different device sets than initially thought. Maybe the video we produced doesn’t actually engage the learners. Or perhaps the sequencing of learning activities seems to be wrong, as the data might show they jump between sections rather than following a linear path. Regardless of what it is, smart data-driven learning design enables us to get information, understand its magnitude, and make design decisions accordingly. Remarkable results are not produced in one iteration.
If we want to improve as an industry, L&D has to start working with data to be able to produce better outcomes. It’s easy to view data-driven learning design as something daunting and terrifying, but it’s really not. Sure, we need to adjust our mentality a bit. We need to become more comfortable with “betas” and iterations, and the fact that we may not always get it right the first time. But once we get past that, once we learn that, there should be a great future ahead. And if you’re not entirely comfortable with all this just yet, we are happy to hold your hand. Just contact us here.