It’s not a robot. It’s the employee of the future. (Illustration: Andrés Moncayo)

.

Employee Training Isn’t What It Used To Be

And thanks to big data, that’s a good thing.

In 2009, Pep Boys, the nationwide auto parts and service chain realized that their traditional ways of educating their employees about theft—through posters, classes, and meetings—weren’t really working. They turned to a new Canadian-based startup called Axonify to try a different approach, where the information was stripped down to the most critical concepts and presented more like mobile games: quick sessions that employees could complete on their phone in just three minutes each day. Using the system was voluntary, with the incentive of earning points that could be redeemed for rewards.

The program didn’t take long to prove its worth. Unlike many corporate learning systems, not only did employees use the system, but doing so generated measurable business results: Pep Boys saw their losses due to theft at their more than 700 stores drop by $20 million in the first year alone, because their employees were better able to identify suspicious behavior and report it properly. Before the experiment, “they took for granted that employees knew what to do,” says Axonify CEO Carol Leaman, but it turned out that they needed to actually learn theft prevention tactics, not just be exposed to it.

The human resources industry is in the midst of a huge shift in how it thinks about employee training and learning. “A lot of other areas of business have already been transformed through technology, but HR, as is often the case, hasn’t had the same level of investment until rather recently,” says Jon Ingham, a UK-based consultant in human capital management. The HR software market is now estimated at $15 billion, but not all of that money is being put to good use. According to analyst Josh Bersin, despite the fact that learning management systems are the fastest growing segment (currently worth about $2.5 billion), up to 30 percent of the corporate training material that companies develop is wasted.

The very idea that training should be measured by what employees actually learn is a conceptual breakthrough in and of itself. In the 1990s, traditional classroom training started to give way to “learning management systems,” which helped companies better scale their training efforts, because instruction could be centralized and distributed on-demand via their corporate intranet. But the data and reports they generated were primitive. “At that time, it was very much about who attended the courses,” says Jonathan Ferrar, vice president of IBM’s Smarter Workforce, “but that’s of almost no value. What companies really want to know is whether employees actually learn and retain the information, and whether it’s the right information for improving business performance.”

Advances in big data analysis and machine learning now allow IBM to isolate variables and discover which are responsible for significant learning insights. “Five years ago, that type of analysis would take statisticians and data scientists days or weeks,” says Ferrar, “but now it can be done in minutes or hours.” He notes that when companies have an accurate assessment of employee knowledge, they can actually save money. “Rather than wasting employee time by making everyone sit through an hour-long compliance training each year, for example, companies should first find out who actually needs the training, and who already knows the regulatory standards.”

In Axonify’s platform, assessment and training are directly tied together. Because many employees use Axonify regularly, the platform is able to constantly track employee knowledge and intelligently provide the information needed to close an employee’s individual knowledge gap, says Leaman. The app also leverages learning research to optimize retention by repeating the questions in specific time intervals. Even after an employee “graduates” out of a specific topic, the questions will still be revisited about seven months later to help lock in the knowledge.

IBM uses behavior data a bit differently, to deliver useful training materials to employees when they actually need it. For example, when a new IBM employee schedules their first meeting with other employees, the assistant detects that it’s their first time, and proactively presents material about how to conduct a meeting. “We’re closing the gap between new and experienced employees, and accelerating that transition,” says Kramer Reeves, IBM’s director of messaging and collaboration solutions.

.

Then: 1990's and Earlier

Traditional Classroom Training

How did it work? Exactly how you’d expect it to work. In person lectures gathered employees and trained them collectively in organized sessions.

What did it measure? Little more than attendance and, if there were tests and quizzes, individual performance scores.

Learning Management Systems

How did it work? It brought the classroom experience to the computer screen and removed the need for in-person lectures or sessions. Training could now be done individually at the employee’s convenience.

What did it measure? LMS’s were largely limited to measuring completion of the training and, if there were tests and quizzes, individual performance scores.

Now: 2000's

The Big Data-Driven LMS

How does it work? With new tools in big data analysis and machine learning, you can identify insights of what works and what doesn’t in your training tools in minutes—as opposed to days in the past.

What does it measure? Big data can definitively show how well your training works—making the process more efficient and cutting down on unnecessary training.

The Smart LMS

How does it work? Training’s been unbundled and different tools teaching different skills can be deployed a la carte when relevant challenges are encountered.

What does it measure? The Smart LMS can measure how often different skills in the position are needed and how necessary training is for the various skills.

The Social LMS

How does it work? The social web has broken down walls that once resulted in employees being trained in a vacuum. Instead of having a single system that teaches all employees the same things, new employees can learn from experienced ones.

What does it measure? By bringing together the training needs of new employees with the experience of more tenured ones, employers can better close the knowledge gap between them.

Why does all this matter?

U.S. organizations spent $171.5 billion on employee training and development in 2010 and $156.2 billion in 2012.

Share This
 

But to really get insight about what employees know and how they’re learning, analytics systems will need to take into account more than just HR-provided training material. “The things that happen in a learning management system are less than ten percent of the activities that real people pursue when they want to learn something,” says Tim Martin, a co-founder of Rustici Software. “If you want to learn something, you don’t go to an LMS, whether you have access to it or not—you usually go to Google or a co-worker.”

Martin is one of the creators of the Tin Can API, a new standard for communicating and storing information about employee learning events. Tin Can is the modern successor to SCORM, a specification that was originally created to standardize content across different learning management systems. The only things that SCORM could measure and track were those where a single user was logged into a learning management system, taking a prescribed piece of training in an active browser session. Tin Can, on the other hand allows companies and employees to record more common learning events: attending a session at a conference, say, or researching and writing a company blog post. “Companies are starting to recognize how employees actually learn and allowing them to do it the way they wish to, rather than forcing them into a draconian system,” Martin says.

Reeves says that this type of outside integration is part of a larger trend in IT departments. More and more CEOs are demanding technology solutions that support external collaboration, according to IBM surveys. Across industries, companies are shifting from controlled, closed environments to more open environments. It’s no longer feasible to expect a single program or tool to do everything—instead, employees expect multiple applications to work well together in a useful way.

One example of useful linking is the way IBM has integrated social collaboration tools into their talent management and learning systems. Social interaction has long been missing from virtual classroom instruction, and after all, learning is “very much a social activity,” says Jacques Pavlenyi, IBM’s program manager for social collaboration software marketing. IBM has found that employees learn and retain more when they’re working socially.

As job-related learning becomes more user-friendly and comprehensive, it also empowers employees to improve their own performance. Leaman says that in surveys of why employees voluntarily use Axonify, she was surprised to see that the most common reason wasn’t the rewards offered, but “because it helps me do my job better.” When people have knowledge, she says, they feel more empowered, more confident in taking action, and “are actually much better employees.”

Ten years ago, says Ingham, HR technology was mostly meant to be used by the HR department, whereas now companies are more focused on employees themselves as the primary users. In the future, Ingham would like companies to use technology not to control employees, but to enable and liberate them to increase their own performance. “The opportunity is not to use analytics to control but to give employees meaningful data about the way they’re operating within an organization so that they themselves can do things to improve their working lives and their performance,” he says.