What it takes to bring science into a conversation that didn't ask for it

In short: Most evidence-informed practice assumes you diagnose first, then prescribe. In commercial L&D, the brief is usually written before you walk in. This article covers three moves that create space for evidence anyway: slowing down the brief, renegotiating what counts as success, and finding the smallest structural change that actually fits the constraints. It's not clean. But it's where the real work happens.

The condition most evidence-informed practice discussions assume is that you start with a problem and then choose your response. Diagnose first, prescribe second. That is how the science is supposed to enter the room.

In my work, the response is usually chosen before I arrive. Customers buy a learning management system, and then ask how to make their e-learning better. The brief has already been written. The technology has already been procured. The stakeholders have already aligned on what success looks like, which is usually expressed in completion rates and learner satisfaction scores. The science was not invited to that meeting.

This is a different starting position than most evidence-informed practitioners describe, and I suspect it's one more of us actually work in than we admit. So I want to write about what it takes to bring evidence in when the commercial context didn't ask for it, and what that has taught me about the practice itself.

The first move is diagnostic, not prescriptive

Before I respond to a brief, I try to slow it down. The customer has decided they want a better course. The question I want to introduce is whether a better course is what will actually move the performance they care about.

That is straightforward in the performance consulting literature — Mager and Pipe wrote about it in 1970, Gilbert in 1978, Brinkerhoff built his Success Case Method on it — but it is not how procurement conversations work. By the time I am in the room, the brief assumes the answer.

So the first piece of evidence-informed work is not telling them what the research says. It is asking the question that the research would ask: what behaviour, in what context, by which people, under what conditions? Most of the time this question has not been asked yet. Asking it doesn't feel like science. It feels like good consulting. But it's the move that creates the space for everything else.

The second move is reframing the unit of measurement

Once the conversation is open, the next constraint is what counts as success. If the brief says completion rates, the evidence will lose. The literature on transfer of learning is clear that what people complete is not the same as what they apply, and what they apply is not the same as what changes performance over time. Eraut's research on informal workplace learning (2004) is one entry point. Brinkerhoff's work on high impact learning is another.

But none of this lands if the customer is still measuring completion. So before I can introduce any evidence about what works, I have to renegotiate what we are measuring. I don't always succeed. When I do, the rest of the conversation becomes possible. When I don't, I work within the metric they'll accept and try to add a second one quietly.

The third move is choosing the smallest leverage point

Even after the diagnosis is open and the metric is honest, the constraints are real. The customer has six weeks. The budget is fixed. The managers haven't been briefed. The legacy infrastructure doesn't support what the research would suggest.

So I look for one structural change that punches above its weight. Sometimes that's a manager conversation guide so the line manager actually has the conversation that drives transfer. Sometimes it's a performance support tool that sits in the workflow rather than in the course — Gottfredson and Mosher's work on the five moments of need is useful here. Sometimes it's reframing what the course is for, so it stops trying to do what it cannot.

These are not the cleanest evidence-informed interventions. They are the ones that fit through the constraints. The science gives me the direction. The constraints draw the route.

Judging whether it works, without the outcome data

This is where I am most uncertain, and most willing to be honest about it. I don't have my own outcome data on most of these interventions yet. I am drawing on what the research says, applying professional judgement, and reading what comes back through the customer.

Because that is the position I work from. I work with the customer, and the customer works with their learners. The behavioural signals from the workplace reach me indirectly, if at all. So what I read are signals about the customer themselves. Are they coming back with a different question, or the same one in different language? Are they reporting completion numbers, or are they starting to talk about what people are doing differently? Are they describing what the course is achieving, or what their managers are doing alongside it?

These are softer signals than I would like. They are not controlled studies. They are the readings of someone working one step removed from the workplace where the actual behaviour change has to happen. I think this is honest evidence-informed practice in a commercial setting, even though it sometimes feels like an admission rather than a method.

The work isn't applying evidence. It's making space for it.

The framing I came in with was that evidence-informed practice is what you do once you have a problem and a research base. After several years in this work, I think that's the second half of the job. The first half is the part nobody trained me for: creating the conditions in which evidence can enter the conversation at all.

That is what I find myself spending most of my time on. Not citing studies. Asking earlier questions. Renegotiating what counts as success. Choosing leverage points that fit through the gap between what the science suggests and what the organisation can absorb. None of this is dramatic. Most of it is invisible. But it is, I think, where evidence-informed practice actually lives in commercial L&D, and where most of the difference gets made.

Frequently asked questions

What is evidence-informed practice in L&D? Evidence-informed practice means drawing on learning science research — on how people learn, retain, and transfer knowledge — to shape how training programmes are designed and evaluated. In practice, it means asking whether the intervention you're designing is actually likely to change behaviour, not just complete a course.

Why is evidence-informed practice hard to apply in commercial settings? In most commercial L&D contexts, the solution has already been decided before a specialist is involved. The technology is procured, the brief is written, and success is defined in completion rates. That leaves little room for the diagnostic questions that evidence-informed practice depends on. The challenge is creating that space within existing constraints, not waiting for ideal conditions.

What is transfer of learning, and why does it matter? Transfer of learning refers to whether what someone learns in a training context actually changes how they behave at work. Completing a course and applying its content are not the same thing. Research by Eraut (2004) and Brinkerhoff shows that most learning investment is lost between the course and the workplace, making transfer one of the most important and most undertracked outcomes in L&D.

What are the five moments of need? The five moments of need is a framework developed by Gottfredson and Mosher that identifies when people need performance support: when learning something new, when learning more, when applying what they've learned, when things go wrong, and when something changes. It's used to design support that sits in the workflow rather than in a course.

How do you measure the impact of learning without outcome data? In commercial L&D, direct outcome data is rarely available. Margot Sprenkels describes reading indirect signals instead: whether customers return with different questions, whether they start reporting on behaviour change rather than completion, and whether managers are beginning to play an active role alongside the training. These are softer signals, but they're often the most honest measure available.

References

• Brinkerhoff, R.O. (2006). Telling Training's Story: Evaluation Made Simple, Credible, and Effective. Berrett-Koehler.

• Eraut, M. (2004). Informal learning in the workplace. Studies in Continuing Education, 26(2), 247–273.

• Gilbert, T.F. (1978). Human Competence: Engineering Worthy Performance. McGraw-Hill.

• Gottfredson, C. & Mosher, B. (2011). Innovative Performance Support: Strategies and Practices for Learning in the Workflow. McGraw-Hill.

• Mager, R.F. & Pipe, P. (1997). Analyzing Performance Problems (3rd ed.). Center for Effective Performance.