Data and analytics skills are essential for the future of learning and development. With the digital skills crisis presenting a defining challenge for the global economy, knowing how to analyse and make decisions based on trends in learning data is key to future-proofing your workforce.
However, despite the importance of these skills for L&D, data analytics is still among one of the five weakest skills reported in the 2022 LPI L&D Dashboard: “When L&D lacks the skills needed to define, collect and assess the data that proves learning efficacy and impact, it not only lacks the insight required to inform learning strategy but also the influence to improve organisational decision-making.”
In addressing the skills gaps of other employees, is L&D failing to develop its own skillset for the future? If so, how can we take steps to improve our own capabilities in data analytics?
1. Identify the data you have against the data you need
You can’t start collecting and analysing data until you’ve identified what data it is you need to collect (and whether it’s available to you). The data you need will depend on what you’re trying to prove. For example, are some of your customer support staff having problems with communicating updates effectively to customers? Delivering training around communications would help to reduce the amount of helpdesk ticket requests if customers are equipped with the knowledge they need from the outset.
Many L&D professionals are still relying on learner feedback forms (‘happy sheets’) as evidence of success. However, if the goal is to evidence an improvement in performance, it doesn’t matter how many enthusiastic learner responses you get – the data in this case is irrelevant to the context. It doesn’t actually evidence a positive change in performance as a result of the learning, instead merely showing that learners liked it.
Perhaps you’re using a learning technology platform such as a learning management system (LMS). Today’s newer, more advanced learning systems go beyond completion rates and are able to record data that builds a bigger picture of learner experience. With the in-built xAPI capability of a learning experience platform (LXP), you can identify which resources and topics are most accessed, the devices used to access learning materials, the answers and attempts for individual questions and how much progress individual learners have made towards personal learning goals.
The key is to ensure that the data you’re able to collect, whether qualitative or quantitative, feeds back to what you’re trying to prove.
2. Talk to other departments
What if the data you need isn’t available? Or perhaps you have access to data but you’re not sure where to start with analysing it?
You need to start collaborating with other departments to create a robust, systematic approach to data collection, using metrics already being collected.
Marketing - Your marketing team tracks a significant amount of data to determine what strategies are working and how this influences behaviour. The more you talk to your marketing department, the more you’ll notice the parallels between marketing and L&D. Your marketing team will be able to talk to you about search terms, device usage, when people are most active and what learning content is generating the most engagement. This data will be readily available in today’s more advanced learning systems – Thinqi, for example, can integrate with Google Analytics for deeper insights into the content being searched for, the number of people accessing this content and how long they’re engaging with it so you can report on that data.
HR - The 2022 Linkedin Workplace Learning Report notes that learning leaders are breaking down traditional silos to collaborate on a more holistic vision for HR. If you’re collecting data around skills gaps, staff development and succession planning, your HR system will be able to provide this. If you’re using a smarter learning system like Thinqi, this can seamlessly integrate with your HR system to pull through organisation structure and performance data, then assign the correct roles and permissions for their training.
- Business leaders - What do you want your data to prove to the wider business? Research has found that for three out of four organisations surveyed, aligning learning strategy with business goals is a top priority. Learning programmes should always start with business goals in mind, linked to relevant metrics and KPIs. Talk to your business leaders to learn how prioritising groups of people for training interventions will impact the business mission or bottom line. Understanding the business impact will help you determine what data you need to justify your training priorities and prove your hypothesis.
3. Practise data storytelling
According to Chief Economist at Google, Dr Hal R. Varian, “The ability to take data – to be able to understand it, to process it, to extract value from it, to visualise it, to communicate it – that’s going to be a hugely important skill in the next decades.”
You must be able to use insights and evidence-based information that tell meaningful stories to your stakeholders. Data storytelling uses a combination of three key elements:
To make your story relevant, it needs to link back to your initial goals. For example, perhaps you need to demonstrate how training has reduced attrition rates or improved customer satisfaction scores. Consider four types of data analytics to tell your story:
- Descriptive analytics - What has happened?
- Diagnostic analytics - Why has it happened?
- Predictive analytics - What is likely to happen?
- Prescriptive analytics - What action should we take next?
These four categories enable you to paint the full picture of how the training has impacted the initial goals. Too often L&D falls into the trap of simply describing data (descriptive analytics) without supporting it with the context afforded by various other types.
Jim Stikeleather made a valid point in Harvard Business Review when he said that “visualisation in its educational or confirmational role is really a dynamic form of persuasion.” Think about what story your audience is interested in, how best to present your data and how your narrative can persuade key stakeholders to accept what you are trying to prove.
4. Be mindful of causation vs correlation
When looking at your data, it can be tempting to attribute causation to any data sets that happen to correlate. However, it’s worth bearing in mind that just because two data sets have moved in the same direction over the same period of time, this does not automatically imply causation.
You need to consider a variety of extraneous variables that may be affecting your data. Simply making the assumption that correlation means two data sets are linked could lead to ineffective decision-making and skew the direction of future strategies. Reduce your variables and consider a/b testing, changing just one thing at a time to determine the impact. This allows you to fine-tune your strategy and gradually optimise for the best results. The more you can isolate and manipulate the independent variable to determine its effect on the dependent variable, the more accurate your results will be.
If you do then discover a relationship between two variables, then you can safely make predictions about one in relation to another.
5. Never stop learning
Reskilling and upskilling are the top priorities in this year’s Global Sentiment Survey. Change has accelerated since the start of the pandemic, resulting in a greater demand for upskilling and reskilling to adapt to new ways of working.
However, the 2022 Linkedin Learning report revealed that L&D’s rapid rise has spurred new pressure to deliver results and with their to-do lists exploding, time for personal learning has fallen to the bottom. Compared with other active learners on Linkedin, L&D learners spent 23% less time learning in 2021.
Cross-functional collaboration is a great way to learn from others who are already proficient with data analysis. Likewise, greater collaboration promotes social learning opportunities, understanding of stakeholder requirements and development of communication skills to prove results. In summary… According to Towards Maturity, 51% of L&D professionals say they cannot use data effectively due to a lack of in-house data skills on their team.
However, by following the five simple steps outlined in this blog post you can start building the skills you need to get confident with data analysis today. We also have a handy free expert guide to really take your data analysis skills to the next level – perfect for upskilling in your next coffee break.
As important as it is to upskill our own people, it's still vital that L&D never stops learning.