Ask Fuse Product Director Rhys Giles how people actually learn, and the first thing he’ll tell you is that content aggregation is definitely and decidedly not the answer to engaged learning. Many enterprise level companies have invested in learning experience platforms (LXPs) that advertise themselves as putting learning in the hands of the learners rather than administrators, and which rely completely on algorithms and content aggregate engines to recommend learning content.
Make no mistake: there’s a looming skills gap in enterprise that analysts like Gartner have always been quick to highlight. In its 2018 UK Shifting Skills Survey, Gartner communicated with over 7,000 employees who were asked to self-assess their level of proficiency in in-demand skills. Of these, 70% said they hadn’t mastered the skills they needed for their jobs at present, and 80% said they lacked both the skills they needed both for their current role and their future career.
In this uplifting Q&A with Rhys, we got some great answers to why LXPs focused on content aggregation aren’t solving the skills gap, as well as his thoughts on how to empower people to go deeper and connect with the knowledge and expertise they need to improve their skills and perform beyond expectations.
Q: Rhys, why isn’t the current LXP approach to skills working?
Rhys: Before we get on to the technology itself, or indeed some of the problems LXPs are having with engagement in learning, it’s important to look at the history that led up to what is currently going on with LXPs’ approach to skills.
The LXP category was invented because in the past, many companies felt that LMS platforms focused too much on the management of learning to the detriment of experience. With a poor experience, you don’t get engaged learning.
LXPs rushed to solve the problem, promising to reduce the management of learning by putting the experience first, in a consumer-grade fashion similar to Netflix. By focusing on content aggregation and discovery, LXPs would connect content libraries together and recommend content to learners based on their perceived skills gaps. Through automated curation, L&D could scale the amount of content it offered, theoretically meaning that every job role could be covered.
Here's the ‘but,’ and the key issue: by matching large amounts of content to generic skill categories and serving feeds, LXPs had sacrificed something extremely important - relevancy. Context is key to a great recommendation, and by this I mean understanding who the learner is, and understanding their situation and their motivation. What is their job role? What business unit are they in? Which industry is their company part of? And beyond that, there is further context, such as, is the learner looking for an answer to a question, or a full course?
For example, let’s say I have a perceived gap in communications, and I have to present to the board. A feed of content with communications from three different vendors is unlikely to be very helpful, and it’s likely you’ll get choice fatigue - just like with Netflix!
If your only data point is skill, then content is always going to be generic and not truly relevant to the learner. We need to go much further if we want to satisfy the learner’s need - we need to go beyond generic macro skills.
Q: You mentioned engagement in learning, and this seems important to explore before we jump into where the technology is not meeting the demands of the skills gap. Does learning have an engagement problem?
Rhys: At the most fundamental level, learning is about being engaged, and where many companies come up short is that they are just pushing automated technology platforms, and they haven’t been created strategically, based on a proven model that focuses on how people actually learn.
At Fuse, our whole business model is built around creating engaged learners, and that’s why when you look at our customer case study page, you don’t see stories about the companies that simply plugged Fuse in, walked away and hoped for the best. You see stories about Avon, Panasonic and Hilti, all of which dramatically increased their learning engagement in a short period of time, because of their understanding of what it takes to engage learners, and because of how the Fuse platform supports this.
Our customers know that there are many core elements that go into building and maintaining engaged learners, but you need to start with a willingness to engage: people need to feel they have a choice, and they need to want to partake in engaged learning. Next, you need leadership that will lead by example, and who will help create a culture of engaged learning, where people feel comfortable to learn continuously.
A fundamental part of engaged learning is the strategic nature of the content itself - but this is a whole separate discussion!
Q: We will get to the strategy and specific nature of engaged learning content in a minute, but first, let’s get back to the operational technology bit I asked about in the first question: why aren’t LXPs, which rely on algorithms and content aggregation engines to recommend learning content, helping to close the skills gap?
Rhys: Firstly, I’m not saying that recommendation engines don’t have their place: they do. But this place is alongside a very personalised learning model that is developed and mapped to meet the skills gap. To be truly relevant and engaging, recommendations need to be contextual, and they need to help us to understand more about the user beyond just their required and perceived skills.
Unlike Fuse, LXPs (except when they work with an LMS) are just course aggregators, pulling in courses from third party providers, which is largely generic content and not related to roles and responsibilities. Often, the content is from content libraries, and then a recommendation engine goes over the top of that. What you end up getting is just a stream of content that can be very hit or miss and is unlikely to solve a need learners have in the flow of work.
What if, for example, as a learner you have a meeting with an industry analyst in the morning and want to best know how to position yourself? This is an example of a given skills gap that exists in the here and now.
The other problem with LXP’s learning recommendation engines is that they aren’t designed with learning in mind, and rather, they are designed on the Netflix media model of consumption, where people get recommendations based on their viewing history. From a learning context, this doesn’t work, because people end up getting recommended more of what they’ve consumed, rather than what they actually need (now or in the future), which could be entirely different content all together.
To be truly relevant, we need to deliver content in the context of the industry. Generic content on marketing or sales doesn't cut it.
Q: What’s the solution?
Rhys: For learning to be really successful, we can’t just be looking back at a learner’s history. We have to be looking forward, at the situation a learner is in at the moment, and what they need to achieve. It is about understanding through data the world that they live in, and what tasks and responsibilities they have at that moment in time, aligned with their role, responsibility, community, and the company they’re in.
At that point, you can start to take a huge amount of content and break it down, as well as input from other data points: in Fuse, we can look at up to 330 data points.
We can also build a picture as to how learners engage with content on the platform: what are they looking at, what are they not looking at? Social is another set of data points. We’re able to crowdsource the popularity of content, and to see whether people are sharing or ‘liking’ content and interacting with one another about it.
Communities are an important part of the social element of Fuse. They allow us to take content, and match that content with audiences through the context of a community, which can be aligned to skills and/or job roles, or even hierarchies. Not all employees will see all content, because they won’t be part of all communities. We use communities to look at what is popular amongst similar types of users - for example, what is trending in Marketing or Retail. The best part of crowdsourcing popularity is that communities makes it more targeted
Contrast this to a system where you have access to everything: LinkedIn Learning alone has 280,000 videos. If you combine that with a further 100,000 learning resources a company may have accumulated over the years and you just put a simple recommendation engine over it, what’s the chances of getting a piece of content that’s relevant to you? Slim, at best I would say.
At the end of the day, all of this is used to create a snapshot of one moment in time (mind the Whitney Houston quote) to create a more specific, relevant and engaged learning experience. Given that the amount of data and knowledge in the world has likely quadrupled as we’ve been speaking, the idea of getting as specific and relevant as possible seems like a top priority for enterprise learning and development, especially if they are to focus on the skills gap so that employees can keep pace and perform well in today’s accelerated business landscape.