Our industry’s number one analyst, Josh Bersin, recently recorded a podcast dedicated exclusively to skills taxonomies. And the message is landing - 93% of businesses see reskilling and upskilling as the number one priority for 2021.
There’s a growing undercurrent of thinking that a skills-based employment setup is not just economically more efficient, but also more socially equitable. Skills like data, digital, leadership and wellbeing are all becoming essential to the wider picture of organisational capability - and learner engagement.
However, despite all the noise around skills, there’s very little substance behind the discussion. And businesses are not even getting close to the reality of skills. But at Filtered, we work with skills all day every day, and we know what that reality looks like. We’ve condensed what we know into this piece, so read on to find out.
The world of skills is a mess, just as the world of learning content is a mess. There are hundreds of them in some companies’ skills frameworks, skills taxonomies, skills matrices etc. And there are many, many thousands of skills out there. They are badly defined and rarely grouped despite many of them being almost synonymous. User-generated skills clouds amplify the noise 10x. And if skills are a mess, we stand no chance of cleaning up our content.
The problem is that skills are, for the most part, unintentional. Making your skills intentional means going beyond a definition and fully articulating what your business means, by, say, resilience. The boiler-plate definition from the content provider is a start. But their definition of resilience is unlikely to be exactly what you intend for your workforce. Does it include getting a good night’s sleep? Should it? Does it include working well under pressure? Should it? Does it include time management? Again, should it?
The answers to these and other skills questions will be different for each company. They depend on many factors, including the industry, the cultural direction and company mission, what leaders at the company stand for and the motivations of its workforce. But they are questions that must be answered, and answered with nuance.
Other questions you need to answer to solve your skills issues: are we clear on how to define what a skill is, not just each individual skill? There’s a hotchpotch of associated concepts: competencies, capabilities, values, traits, behaviours. What is a skill vs a capability? Or a competency? Then there are hard skills and soft skills and power skills. Which of these are within the remit of L&D / HR / Talent, and which are not?
Every business will have different answers, so I’m not going to try and untangle all this here, but the graphic below may stimulate discussion.
The important thing is to pick a line and stick to it. Take, for example, a business practice I find particularly enjoyable and useful (and one Filtered has a long history of training): spreadsheeting. We hear clients say that Excel is a skill, definitively. And we hear clients saying the exact opposite, equally definitively. I don’t actually think it matters all that much in the broad, high-level industry discussions. But it does matter that everyone in your company agrees. Stakeholder buy-in at a senior level is crucial to achieving consensus.
That also helps break down skills silos. There’s a jumble of skills related to content in one system. Ditto job descriptions in another. Ditto learning objectives and KPIs. Ditto appraisals and performance management. The promise that skills can be that coveted ‘common currency’ all the way along the employee lifecycle is still a lofty aspiration rather than a hard reality.
We often miss skills. Some skills are just willfully or subconsciously overlooked. As Annie Lowrey pointed out last month in The Atlantic:
A great deal of skill is required to change the clothes of an immobilized senior who might not want to have her clothes changed, or to wrangle a class of toddlers, or to clean up an overgrown yard at breakneck pace, or to handle five tables of drunk guys who want their wings yesterday. The kind of patience and equanimity it takes to be a good care worker? Not a skill, apparently. The kind of fortitude it takes to be a fruit picker? Not a skill either.
These and countless other skills often remain effectively unknown to an individual’s manager, colleagues, and employer. The wastage here is lamentable and colossal. And it hamstrings the employer when identifying and developing talent.
Indeed, the role of the organisation is often unclear. People learn all the time, by themselves, for themselves, intentionally and unintentionally. That is one of the properties of a neotenic, lifelong learning species. Through social media alone we consume hours of content each day, much of which is informational and educational and almost all of which is genuinely consumer-grade. How are organisations supposed to help direct that learning at the skills they’ve identified as important?
All of these skills questions are difficult to answer. But I believe we can crack this tough nut with the right combination of pragmatic thinking and practice. And if we do, we’re affecting an agenda that everyone cares about and creating that rising tide that lifts all boats. To that end, a skills data blueprint...
We need to assemble the appropriate data in order to design our skills framework well and align business partners and stakeholders. Of course, not all the data we would like will be readily available. But in good, agile fashion we should start with what we have and iterate from there. Here are some key examples.
Top-down data sources
These data need to tell us what high profile initiatives are key for the company, what skills might support those, and what’s likely to change over time. They include:
Bottom-up data sources
We collate all relevant client or learner data that has a bearing on what’s valuable in their organisation. This includes:
Each of these sources need to be approached in their own terms. For example, Job descriptions, especially when they have written thoughtfully, used appropriately in interviews and followed up with throughout onboarding, can reveal much about the desired and required skills for the specific roles across your organisation. We use Content Intelligence to scan and draw out skills from these.
Alternatively, organic user search queries across talent and learning systems can also be informative. They tend to throw up a lot of expected results (agile, excel, leadership) but some of the less frequent, more nuanced terms can bring more actionable insights. Here’s a typical set of results for organic search on an LXP, SharePoint etc:
How useful is data like this? On the one hand, this is real, unprompted user behaviour. On the other, it’s from a small and possibly unrepresentative sample of the workforce population and a lot of the terms are very generic (eg ‘leadership’). In short: it’s of some use. So, we certainly need to combine it with other data. But here are some questions this kind of data might prompt:
Your organisation will have data like this scattered across its systems. It just takes some creative thinking and technical wrangling to unlock. But once you have a dataset replete with skills indicators, you can move on to...
Synthesise all your data and draw it together into a relevant, robust, flexible and extensible skills framework. In our experience, this tends to be 30-100 skills. The skills need to cover the business priorities today, as a minimum. They also need definitions, either accurate description in natural language or skills graphs (see the curiosity pyramid below). They need to be believed in by business partners (and ideally you’ll have a champion per skill high up in the organisation). They may include some levelling.
These components allow your skills development to be intentional and focused and, most importantly, to contribute to business goals. That is the true purpose of L&D.
Determine and define each skill. Take curiosity, for example. There’s a lot of interest in this right now as skills like learning to learn and learning agility gain popularity. But what does curiosity mean in the language of your business? There’s no objective right or wrong. The only misstep you can make here is to not define the skill or behaviour at all. Here’s one interpretation of curiosity.
Here, the terms at the top are more specifically pertinent to the skill under consideration, Curiosity. So for this client, asking the right questions and out-of-the box thinking are very much what Curiosity is about, and iteration, resilience, interruption science less so. If you have developed and articulated this understanding of the skill, your curators, both human and machine, have the basis to curate intentionally.
My co-Founder, Vinit Patel, wrote about how to design a skills framework to be transformative. At the top of that piece we shared some research into some of the most popular areas covered by such frameworks today:
By a distance, the most popular type of framework amongst our clients is some version of digital skills, and these are perhaps the most efficient to define in this manner.
Finally, make sure to consider which skills will be important for your organisation in the future, especially in relation to your industry’s trends. But my advice would be to not agonise over this; it’s more important to construct an extensible framework which can easily accommodate a new, unforeseen skill than to plan the next 10 years, which are so unpredictable.
You can now use those carefully-selected, high-value skills to determine both the learning content that you buy for your company and the specific learning assets that you put in front of individuals. In this section, I’m going to show some of the outputs and UI of our product to illustrate the approach and the art of the possible.
There are algorithms now which can take a nuanced understanding of your skills framework, including all of its levels and taxonomies, and apply that to thousands or millions of pieces of content and decide - as well as a human can - how relevant each is to building that skill. We call this Content Intelligence and one of the outputs looks like this:
Viewing this asset by asset gives you a curation accelerant and informs how your platform recommends content to develop skills your organisation cares about. We can use this data to inform the personalised recommendations in our smart LXP, for example. But the value of solving skills in this way doesn’t stop there.
Assimilating this data at a content library level allows you to see exactly how well each of your library providers caters to the specific skills you outlined in the skills framework process. It’s clear how valuable that can be as part of content procurement to inform content buying decisions and negotiations. Indeed, we’ve saved two different 100k- employee- plus organisations 30% on their content provider spend already. The potential is huge.
To give a specific example of our Content Intelligence in action - we ran our analysis on 6 of the major content libraries' learning assets to assess relevance to certain nuanced skills as well as overall quality. In this case, we looked at the power skill ‘leveraging data’ and assessed its relevance across the learning content that HBR, LinkedIn Learning, Coursera, getAbstract, Mind Tools and FutureLearn provide.
After we ran the analysis, we discovered that getAbstract, LinkedIn Learning and HBR are the most relevant libraries for businesses that are looking to develop or improve their employees’ skill to leverage data. While not all 6 major content libraries have an equal number of assets in total, if you're using or considering Coursera, Mind Tools or FutureLearn, you may be missing out on assets that rank more relevant for your nuanced skill requirement.
So, if your business is looking to focus on leveraging data as one of the top priority skills in your skills matrix, then investing in getAbstract followed by LinkedIn Learning and HBR which have the most relevant content coverage would provide the best ROI.
Of course, leveraging data is only one skill, and you wouldn’t make purchasing decisions based on one skill alone. The real power comes from the scalability of the approach. When you have the data on the 30-100 we mentioned before, you can make truly informed decisions on what content is working to build the skills and capabilities your team needs.
The skills you’ve put together are useful elsewhere too. The right (data-backed, and carefully synthesised), high-definition picture of skills does of course help with learning and development decisions and activities. It’s also interesting to those working in Talent, the C-Suite, PR - and really any hiring manager in the entire business. Because there’s at least a popular and growing opinion that skills is the best lens through which to think about workplace productivity and the full employee lifecycle.
Solving skills isn’t easy, but as I’ve outlined above, it is important. It’s something that myself and the talented Filtered team spend our days delivering for our wonderful, progressive clients. There are so many benefits to uncovering data, building a coherent skills framework and delivering content that builds those skills to your people. If you want help unlocking those benefits, we can help with any stage of the process, please just be in touch.