Learning in the Flow of Work: How to Design for Impact

Learning in the flow of work is having a moment, again.

Nearly every L&D publication lists it as a trend for 2026. The common proclamation is something along the lines of, "we need to bring the learning to the people," "merging work and development."

Which, duh.

Work has always included learning. Employees, whether newly hired or nearing retirement, can learn every day on the job. From their interactions, from taking risks, and from making mistakes. This isn't a novel concept, it's how development works.

What's changing is L&D's ability and capacity to design and deliver learning in the flow of work (LIFOW, hereafter). When LIFOW became popularized in 2018 by Josh Bersin, there were still significant barriers to making it a reality. Things like the scalability of personalized content, not to mention the challenge of getting it into the right workflows.

LIFOW can only transform our learning strategies if we can show the content is working. Otherwise we're just contributing to the growing readiness debt. So, let's talk about how to make sure you don't fall into that trap.

If you're in a rush, watch this quick overview.

Learning has always happened at work

Decades of learning science show that knowledge is situated, meaning it is a product of the context and culture in which it is developed and applied. Put simply, where you learn something influences how you learn something.

For workplace learning, that means work and learning have always been intertwined, with workplace learning existing on a continuum of planned, structured experiences to spontaneous, unstructured experiences.

What's shifting is our concept of the workplace. While in some industries the workplace has remained a constant, in others it is now a coffee shop or a couch. Or a cell phone on an evening commute.

In 2018, Josh Bersin conducted research on workplace learning, finding that employees have an average of 24 minutes per week for formal learning, about 1% of a working week. So, he introduced LIFOW as a paradigm shift in how organizations design for learning.

I'd argue that the LMS is responsible for a lot of the disconnect between learning and work, introducing separation between what people need to learn and its context, and focusing evaluation on completion rates instead of application. But that's a rant for another day.

Learning that employees want 

I'm a naturally curious person. I love learning. (Learner is my number one theme in CliftonStrengths, so clearly I'm in the right profession). What that means is I'm motivated to seek out knowledge for my own edification. LIFOW makes it easier for people like me to find well-designed content.

But it also makes it easier for people who are only looking for information when they need more support. That could be:

• A sales rep is about to jump onto a call with a prospect. They know the pricing model recently changed, but they want a refresher on the nuances. Fortunately, there's a 30-second video sitting in Salesforce about the changes and a script for how to address objections.
• A new hire is starting their second week. They can't remember the steps for a workflow, so they ask an internal AI agent that searches the onboarding learning content and surfaces the interactive demonstration.
• A manager is preparing for a meeting where they have to deliver constructive feedback. They remember learning a framework in their development program (S-B…something), so they go into their manager's Slack channel. An infographic of SBI is a pinned resource.

In each of these scenarios the learning need is prompted by the work, be it a knowledge gap, task complexity, or a time-sensitive incident.

When (not) to design LIFOW

I read something earlier this year that proclaimed courses, and by extension, 'traditional' or 'structured' L&D is dead. It was LIFOW or bust. (I'll get t-shirts made for us.)

I disagree.

LIFOW does not replace formal learning experiences, especially the ones designed around discussion and problem-solving. Instead, it serves as the connective tissue, helping employees apply the skills they've learned, get feedback on that application, reinforce the knowledge, and get just-in-time support.

There are two factors to consider when determining whether or not to use LIFOW:

1. Do I have to verify the outcome to a reporting body? If you are responsible for reporting on the outcomes of a training to legal, regulatory, or safety authorities, then I would strongly recommend against LIFOW. Compliance training is the most compelling use case for keeping an LMS for most enterprises. There's a reason why.
2. Does the learning require sequential scaffolding? Skills take time to build. If a learning experience needs scaffolded progression, where foundational knowledge must be in place before progressing to a more advanced experience, then LIFOW could fragment the learning. You can't reinforce something that hasn't been conceptually grasped yet.

Now, in both cases, you can still use LIFOW as a support mechanism. If there's a safety incident, you can post a quick reminder video on an intranet's homepage to remind everyone of best practices. Once someone has progressed through a structured learning experience, they can get reminders to practice, perhaps with an AI coach for reinforcement.

Here are a few examples of how that could work.

Remember how earlier I said our workplace is evolving, but not for everyone? For farmers and pastoralists in East Africa, their workplace continues to be the land. And LIFOW works for them too.

Thanks to organizations like Justdiggit, they no longer have to wait for a trainer to travel to them to deliver a session on how to dig semi-circular bunds to capture rainwater and reduce erosion. They can be on their land, on their phone, watching a video, pausing and rewinding as needed, while they follow the process.

A caveat: in working environments where roles are highly prescriptive and have low autonomy, it can be more challenging to design LIFOW. There's less incentive to seek out learning, and therefore more learning may need to be pushed to the employee.

LIFOW needs a measurement plan

Four years after his seminal study on LIFOW, Josh Bersin conducted follow up research. As of 2022, most organizations considered LIFOW the gold standard for workplace learning, and yet only 12% felt they do it effectively.

That gap is a measurement failure.

AI is making it easier for learning designers to design and deliver more personalized content at scale. But as a profession, we haven't been able to realize the same gains in how we evaluate that content.

How are we measuring LIFOW? Not just conceptually, but tactically.

Given the potential volume of content that may be delivered in the flow of work, I recommend identifying what "good enough" (more on that here) impact measurement looks like for each piece of content. That means one metric that you can reliably and predictably collect that shows the impact.

With this information, you should be able to confidently say something like, "With this LIFOW content, we expected behavior X to change; we see evidence it did or didn't; metric Y moved in the expected direction or didn't. Here's what we'll keep doing, stop doing, or adjust."

I'll use an example from a sales enablement team we work with. A new rewards program had been rolled out at their organization, and yet sales reps were not promoting the program on calls. This was surfaced in analysis of call transcripts. So, they created a microlearning to remind reps about the program: what it was, what was in it for them. After sharing this content, they saw a six-figure increase in rewards program revenue and a 61% increase in mentions of the program on calls.

The behavior they wanted to change was more selling of the rewards program, and they could show that through call transcript analysis, with a clear baseline picture before and after over a three month period.