What Can Ants Teach L&D about Learning Evaluation?

But first, a little background…

We’ve all been there. You’re moving into a new place, which means you’ve got to sort through and pack all of your things.

You’re sure to discover something valuable or interesting—something you had intended to use or read—but had completely forgotten about till now.

Well, that’s sort of how this blog post about pheromone analytics came to be.

Several years ago, Andrew Downes attended the Learning Analytics and Knowledge (LAK) conference in Edinburgh, Scotland. LAK mainly focuses on the academic side of learning analytics (specifically the analysis of university students), so coming from a corporate learning analytics background, hearing different perspectives and ideas was really interesting.

There was one idea in particular, pheromone analytics, that caught his interest. So, Andrew did some follow-up research, documented it, and gave an internal presentation at Watershed. But that’s where we left the idea, sitting in a dusty corner of our shared drive for the last few years—until now!

Pheromone analytics isn’t a term we’ve heard before or since that LAK conference. And as far as we know, it’s only ever been applied in academic learning. But, we think it’s an interesting idea, and hope you and others in the corporate training world will like it, too.

What is pheromone analytics?

Pheromone analytics is a term coined in 2006 as part of research by B. Van den Berg, C. Tattersall, J. Janssen, and R. Koper. During this study, university students were given learning materials with the option to complete them in any order of their choosing.

Researchers recorded each path and then recommended the paths of those learners who passed the module (not those who failed or did not complete it). Over time, the paths with the most successful learners were given the strongest recommendations.

And that’s why it’s called pheromone analytics—because the process is modeled on the pheromone trails of ants.

When ants leave their colonies to look for food, they spread out in many directions. Like students completing a course, some ants are successful and some are not.

The successful ants that find food, take that food back to the colony and as they do so, leave pheromone trails they can follow back to the food. Other ants can follow this trail too, and as they find food they add to this trail, making the scent stronger and stronger.

And the same principle applies to successful students recommending and leading other students along the same paths.

And pheromone analytics outside academia?

The academic study referenced earlier was conducted in a tightly controlled environment. Learners worked through a predefined set of learning materials, and success was determined via an exam with pass and fail criteria.

While corporate training might have some learning programs that follow a similar structure, we suspect that pheromone analytics could be a lot more interesting if used to study learners out “in the wild” while exploring their own self-directed learning plans.

In the corporate environment, learners have access to huge libraries of content—not to mention the vastness of the internet, which allows learners to search and choose what they deem relevant. Similarly, real-world success isn’t often measured by exams, but by a wide range of metrics relating to a mix of goals and ambitions.

In this context, pheromone analytics will need to:

• look at data from a much broader range of learning experiences.
• examine different success metrics, such as specific business KPIs.
• go beyond looking at which order of a fixed list of learning experiences lead to success.
• explore the set of learning experiences completed by successful learners (since different learners will select different experiences).

Why pheromone analytics?

The goal of pheromone analytics is to identify successful people and what makes them successful so others can mimic that success.

If you can achieve this, and bring weaker performers closer to the level of your top performers, it could have a significant impact on business metrics.

But to make this happen, you need to understand how people have developed the competencies required to be successful—pheromone analytics is designed to do just that.

Here's our process

Here's how we think a simplified process for pheromone analytics in a corporate setting might look:

1. Identify a group of learners in a similar role.
2. Evaluate which learners are most successful based on business KPIs.
3. For those learners, determine which activities and interactions were most commonly experienced.
4. For those activities and interactions, find the sequences that most commonly lead to success.
5. Promote these activities and sequences to the entire group of learners.

So that’s pheromone analytics—taking lessons from ants to evaluate, improve, and reinforce the learning journeys people take.

Could it be the next “big thing” in learning analytics? We're not sure, but if it catches your attention and you’d like to discuss this further, feel free to leave a comment below or reach out to us directly!