In the past, predicting how employees’ careers were going to pan out was more of an art form than an exact science, but this is changing.
We are on the cusp of being able to gather a great deal of information about all the ways your employees engage with their learning content, such as what training they have taken, how deeply they have engaged with it, and how they perform on any included assessments.
This can then be joined with employees’ other HR information, including degrees and certifications, the results of any personality tests and other assessments, and of course their performance reviews.
What can you do with all this data?
By using business intelligence tools, it will soon be possible to predict how well a specific employee is likely to perform in a given role.
Better still, it will be possible to intervene and influence these expected outcomes and help every employee to fulfill their own potential within your organisation.
For example, by finding a correlation between new employees’ scores on proficiency tests and their on-the-job performance, and then tracking these metrics over time, you can start to build increasingly accurate models of what learning and on-the-job behaviours correlate with career success at your firm.
This makes it possible to identify potential future high fliers who may not initially stand out, and put in place steps to help them achieve their potential.
Once you know what changes you want to see, it’s far easier to work out how to measure them
The same approach works for many different areas, including sales, compliance or safety incidents, and performance KPIs.
Data analytics at a glance
Learning data analytics can be broken down into four stages.
1. The most basic is descriptive analytics, which explains what actually happened, for example, how many people watched training modules, how well they scored on an assessment, or how many industrial accidents they were involved with last year. (This answers the question ‘what happened?’)
2. Diagnostic analytics goes one step further and aims to link cause and effect: for example, finding out whether the lack of a specific training module or specific employee skills might have been responsible for a noted increase in industrial accidents. (This answers the question ‘why did it happen?’)
3. Next up is predictive analytics, which lets you peer into the future and predict what’s going to happen next. For example, in light of a set of new training videos you produced (and an employee’s high scores on a subsequent assessment), what is the chance that he will have an accident this year? (This answers the question ‘what is likely to happen in the future?’)
4. This can take us to the most valuable kind of analysis, prescriptive analytics. The goal here is to model what actions can be taken to change the predicted outcome.
For example, recommend a specific training your employees should undertake today to reduce the likelihood of an accident in the future. (This answers the question ‘what can I do to change the future?’)
Adding predictive learning analytics to your learning toolkit
If you’re ready to add predictive learning analytics to your organisation’s learning toolkit, here are a few pointers that will help.
Select the metrics you want to change
Whether it’s increasing sales quotas or improving safety ratings, start by figuring out what success should look like. Once you know what changes you want to see, it’s far easier to work out how to measure them.
Design learning programmes that deliver the data you need
When you’re designing your learning and development programme, make sure that you build in plenty of opportunities to measure how participants are engaging with and absorbing the content.
By combining this data with your results metrics, it’s possible to make meaningful predictions and then meaningful changes.
Consider your video strategy
Since video is such a rich data type, it’s a particularly effective source of interaction data. You can see who watched a video, when, how many times, and how much of the video they watched, and which parts were re-watched (potentially points of confusion).
Add interactive video quizzing into the mix too, and it becomes possible to delve even deeper and monitor comprehension and interest levels.
Look for software that supports xAPI
Choose software platforms that support the industry standard xAPI protocol. This is important because xAPI is the magic glue that lets you track anything the learner does —whether that’s through more innovative learning experiences (such as games, videos, or mobile apps), or through job tasks that put these learnings into practice.
Essentially xAPI allows you to bring data from your learning management system (LMS), apps, social platforms, classroom training, and real-world tasks together into one database (known as an LRS, or ‘learning record store’) to create comprehensive reports and analytics.
Aggregate your data in a learning record store
Start gathering as much data as you can as soon as you can and store it in a learning record store that supports xAPI.
From a data perspective, an LRS sits at the heart of your learning ecosystem and brings data together from all your learning systems, applications, and content.
Your LRS can also be set up to connect to operational systems, allowing you to add other data points, such as job performance metrics, to the pot too.
Even if you don’t plan on analysing the data in the LRS immediately, having a historical baseline to work from will make the results more useful faster.
Deploy a learning analytics platform
Use a learning analytics platform on top of your LRS to link your data to your metrics and extract meaningful insights and reports from your learning data.
Combined, these tools help you understanding of the learning happening within your company, and how that learning impacts the overall business.
Once you have this set up, you can start defining benchmarks and KPIs.
By programming these baselines into your learning analytics platform, you’ll be able to dig deeper and identify outliers, discover relationships, and better align with organisational goals.
Create a virtuous cycle
With these data gathering and analytics mechanisms in place, it’s easy to modify your process and learn from the results.
Put simply, stronger results generate more data that can be used to further refine your training efforts over time.
Predicting the future
Predictive learning analytics is set to become a key tool in the learning and development toolkit.
It will provide learning professionals with the insight needed to deliver more personalised learning to employees, helping to keep them fully engaged and ultimately to achieve their full potential within your business.
It should also mean that organisations find it easier to acquire the right people, assign them to the right jobs, and better develop their careers, thus retaining them for longer.
Looking for information about how to improve learning in your organisation? Read how AI can make e-learning easier and faster.
About Jeff Rubenstein
Jeff Rubenstein is the VP of Product Strategy - Learning + Collaboration for Kaltura, Inc. He has held senior roles in a number of educational and technology companies, including 2U and Wimba (prior to the Blackboard acquisition). He works with a number of other companies and standards bodies on learning interoperability standards, and how to create and measure engagement in rich media experiences.