Predictive analytics is the field of data analysis focused on producing data-backed predictive. There are plenty of tools and resources available for constructing predictive models. All you need to do is feed all of your learning metrics into them and you’ll know what the future holds, right?.
Obviously, if it were that simple, everyone would be doing it. And truth be told, most people aren’t doing much more than that – feeding data into simplistic engines that blindly spit out a trendline. But that’s not really giving anyone a deeper understanding of how to plan for and anticipate the needs of their Training team and their organization. So how can L&D leaders use predictive analytics to leverage their training metrics for the most impact?
The answer lies in breaking out of an unfortunate and reactive mindset that many Training teams find themselves forced into. Let’s consider why that mindset develops, and what to do about it – but first, let’s consider what the possibilities are if Training teams can be more future-oriented.
What Your Learning Metrics Could Achieve
How would you answer the following question?
“The Board is concerned about the labor market’s impact on our operations over the next year. Does the Training team have any data that might inform our talent management strategy?”
Quite a few teams might say something like “Yes, the labor market is a major challenge. Give me a week, and I can put together a report on what we’re seeing from the L&D side.” But this response mostly indicates that you don’t have anything specific to say right now. Imagine if you could confidently respond with something like this:
“We’ve been keeping track of the market and our own operations, and we’ve identified some problem areas. We think we’ll be short about a dozen qualified welders by Q3 next year, because the market is very tight and our in-house welder training program has some serious resource constraints. We’ve partnered with Production to put together a business case for funding additional instructors and equipment for the welder training program – we should be able to avoid serious overages on next year’s Q3 and Q4 production targets if we invest now. We’re also optimizing schedules for the next year to get the most out of the program as it is. Give me a few minutes and I can send you everything we have so far.”
Getting that level of detail might not always be possible. But it’s never possible if you aren’t planning ahead constantly. The L&D metrics you collect every day, combined with other data sources (that we’ll discuss below) and a future-facing mindset, are a powerful tool for proving Training’s ROI. Predictive analytics are where the rubber meets the road.
So, What’s Holding Predictive Analytics Back In L&D?
There are two problems facing most Training teams trying to leverage learning metrics into predictive analytics. First, managing operations is often too unwieldy to add sophisticated future predictions into the mix. Secondly, Training teams often find themselves siloed off from outside, non-training data. Both of those factors need to be addressed to make the most of predictive analytics.
Tech Stack Management
In Administrate’s experience, Training teams use anywhere from 6-12 different software systems in their learning tech stacks. Large tech stacks can quickly become cumbersome – especially when doing advanced data work.
Different learning metrics might naturally accumulate in different, disconnected systems. The LMS collects grades, the LXP records what content learners have seen, the HRiS records their certifications, etc. But what if you need to compare how different employees performed on two different versions of the same content?
Well, you can look forward to manually tracking down and pulling the data from all of those systems, and constructing a spreadsheet to bring it all together. Then you can look forward to manually updating all of those systems if you change of the data on that spreadsheet.
Very quickly, wrestling a disconnected tech stack starts to take up a lot of time. Many Training teams don’t realize how much capacity they’re losing to fighting their own software. And that lost capacity forces many teams into a purely reactive mindset. There’s no resources or time to handle issues that aren’t immediate. So the team is always on the back foot, responding to requests and issues as they arise, instead of being proactive and preventative.
Until teams can get better control of their training data, and simplify the process and handling and managing it, advanced predictive analytics are likely more aspirational than practical.
The Need For Non-Learning Metrics
The fundamental descriptive analytics of a training program, from learner grades to instructor assessments, are always going to be the foundation for reporting on a training operation. But the kind of powerful predictive analytics that Training teams should be utilizing can’t be done with just Training’s own internal data.
Think back to our original example. A Training team can’t project demand for skilled welders in a year just by looking at their own data. Data from Production, HR, and from the labor market itself will all be necessary. Ultimately, it’s Training’s job to provide other departments with the skilled professionals that they need to complete their tasks in the current and future business environment. So it stands to reason that Training is going to need data from all of those sources.
But at too many organizations, the Training team just doesn’t have that access. There are a variety of factors at the level of the Training team and at the level of the whole organization that limit L&D professionals’ access to outside data. Identifying and addressing them is critical to ensuring that you have all the metrics – not just the learning metrics – that you need to fully anticipate future changes and respond with data-driven insight.
Ultimately, predictive analytics will be an extremely powerful tool to the Training team that can get them right. But they’re only one part of creating a better overall strategy for using and leveraging your training data.
Administrate is committed to ensuring that Training teams are empowered to make the best decisions, with the best data, that they possibly can. That’s why we put together our guide, Unleash Training Intelligence for Data-Driven Decision Making. It details how to develop a mindset for utilizing data for decision-making in the L&D context.
