TL;DR

A training program built to teach doulas and midwives to recognize dangerous postpartum complications looked complete on paper. But it wasn’t leading to conversations that got clients to seek medical care. The redesign that followed didn’t start with better content. It started with a question: whether learning and development exists to build what it’s asked for, or to challenge the ask itself, and what changes in practice once that answer shifts.

A nonprofit organization was trying to bring down the postpartum death rate among women in the U.S. It had identified one group of people who could help achieve this: doulas and midwives, who already had the client relationships that made it possible to act once a client showed a warning sign of a life-threatening complication. The nonprofit's first attempt to prepare them for that moment followed a familiar shape. It taught the warning signs, and used the medical vocabulary those signs required. In other words, it ran the way clinical training usually runs.

Heather Burright, Founder of Skill Masters Market and host of the Learning for Good podcast, was brought in after that first pilot, and her verdict on it is blunt: “There was an expert in the front of the room who told you everything you need to know. Now, go do it.”

Nobody left that room having learned something false. They also, mostly, didn't change what they did afterward. Recognizing a warning sign and acting on it inside a real conversation, one complicated by mistrust of the healthcare system, or something as practical as not having a ride to the hospital, are two different skills. The training had only built one of them, and fixing that gap became Heather's job.

Earning the Question

None of that could start, though, until Heather had a stakeholder willing to answer honestly what was really driving the request in the first place. That willingness is not automatic, and it is not fast.

“Those of us who really try to operate more strategically, we lead with relationship versus solutions. I also want to be very outcomes-focused.” Leading with relationship, in Heather's account, means resisting the instinct to walk into a stakeholder conversation already selling a fix. It means asking what is keeping someone up at night before offering anything at all.

That habit compounds over time. “The more we can build relationships with the other leaders in the organization and understand what their needs are and what outcomes they are looking for, the more likely they are to come to us with a problem versus a solution.” A stakeholder who trusts the relationship brings a problem. One who doesn't brings a request that has already been decided.

Credibility, in her experience, does not come from a single well-run project. It accumulates by “being approachable, being kind, being respectful, being curious, asking questions, trying to understand things from the other person's perspective and not pushing your own agenda into the conversation.” While how someone prepares for a meeting and follows up afterward might not look strategic, it determines whether, months later, a stakeholder is willing to let an L&D practitioner slow a request down and ask what is actually wrong.

Heather reaches for a saying to describe how fragile that credibility is once it exists: “Trust is built in drops and lost in buckets.” One rushed conversation, one moment of pushing a solution before understanding the problem, can undo months of everyday interactions.

None of this guarantees the question actually gets asked, or answered. Heather is direct about the alternative some stakeholders choose regardless of how much trust exists: “It could just be, 'I'm not interested in that. We need the training. Just go do the training.'” A change agent who runs into that has a choice: comply, or hold a line that risks the relationship the whole approach depends on.

Even when that trust is fully earned, it only makes the diagnostic conversation possible. It doesn't make it certain.

The Question Nobody Asked Before the First Pilot

In the doulas and midwives project, that earned access is what let Heather treat the request as a starting point rather than a finished brief. Instead of accepting “we need a training” at face value, she went looking for what was actually driving it, and for why the request arrived looking the way it did in the first place.

Heather has a name for how most training gets commissioned, and it isn't flattering. By the time a real problem reaches the person who will actually build a solution, it has usually passed through several other people first, and each one restates it a little more simply than the last.

“Someone looked at a team or across the organization or at an individual, and they saw a problem and they thought, 'Okay, we have a problem. This is a problem.' But what they actually communicate is 'we need a training.'”

She calls this the telephone game. Nobody along that chain is lying. The distortion happens because training is familiar, and a familiar tool is what people reach for under pressure, whether or not it fits the actual problem.

Her answer to that isn't a better piece of content. It's a different question, asked earlier. “I think we have to uncover what it is that's driving the request. So what is the pain point? What is the problem? Once we understand that, we can flip the script a little bit and say, 'Well, what does success look like?'”

Applied to the doulas and midwives project, that question turned into months of listening instead of months of building. Heather spent that time with the nonprofit's stakeholders, trying to understand what the doulas' and midwives' actual days looked like: what a typical client interaction involved, and what usually happened once someone noticed something was wrong. “Through that process, I was able to uncover what it was that they actually needed the doulas and midwives to do and the environment in which they would do it.”

What that process surfaced was narrower than the first pilot had assumed. Doulas and midwives were never going to become diagnosticians, and didn't need to. They needed to notice a specific set of signs, and then do something about it inside a relationship they already had. That distinction, between what someone needs to know and what they need to do next, is where the redesign actually started.

From What They Should Know to What They Had to Do

Recognizing a warning sign is information. Acting on it, inside a relationship, against real resistance, is something else entirely. Heather's redesign of the doula and midwife training treated those as two different design problems, not one.

The change itself, she argues, “typically involves an action. It involves a behavior. So what is it people need to do? Knowledge doesn't equal behavior change.”

The first pilot had stopped at the first half. It taught the signs and assumed the rest would follow. Heather's redesign started from the second half instead, and worked backward to whatever knowledge that behavior actually required.

That behavior broke into two layers. The first was noticing. The second one was doing something about it in the moment, with a client who might be dealing with transportation problems, no childcare, or a general mistrust of the healthcare system. “We need them to listen to their clients. We need them to notice things that their clients are presenting with. And we need them to then take action and use that relationship that they're already building so that their client will seek medical care.”

The target of the training was never really the doula or midwife's own behavior in isolation. It was their client's behavior, whether she went to seek care, and the doula or midwife's actions were only useful insofar as they moved that outcome. A curriculum built around “what do they need to know about warning signs” has no reason to notice that distinction. A curriculum built around “what does success look like” does, because success here was never defined as knowledge transfer, but as a mother getting to a doctor in time.

That target only holds if an organization's leadership is willing to accept it as the definition of success before a program even starts. In the case of the nonprofit, it would have been simpler to promise its leadership a finished training and a completion report instead. Accepting a measure that depends on whether a doula or midwife has a conversation with a client months later is a harder commitment to make, because nothing about it can be counted the week the program ends.

Built for the Room They're In

Once the target of the redesign was defined, namely action inside a real conversation rather than recall of a fact sheet, delivering it still had to fit inside the doulas' and midwives' actual days. This is the part that gets skipped most often, and Heather names the reason directly: “Behavior change is complex. Some of it is motivation. Some of it is habit formation. But sometimes it's the system, the environment that you're in.”

“The closer we can create the experience to the learners' actual environment the better. I should not create an e-learning if the person doesn't even have a desk.” That single constraint ruled out an entire category of design before it started.

Doulas and midwives spend their days in conversation, not at desks or in front of screens, so the training itself had to be conversational. “For this particular group we wanted to make sure it was really conversational.” That extended down to individual questions asked during the training itself. A closed prompt like “Have you faced a situation like this before?” invites a yes or no answer. An open one, “What did you do when you faced a situation like this? How did you respond?”, asks someone to reconstruct their own experience and connect it to the behavior being taught.

None of this shows up in a curriculum review. A slide deck and a conversational, relationship-based design can look equally thorough on paper. The mismatch only surfaces later, in whether anyone actually changes what they do, and by then it reads as the training failing.

The redesign didn't change what the training was trying to accomplish. It changed whether the people sitting in it could actually carry what they learned back into a conversation with a client, rather than leaving it behind in the room.

What Came Back at the Final Session

Whether that actually happened, whether anyone carried it into a real conversation, isn't something Heather usually gets to find out. “As a consultant, as an outsider, I don't always get to see the outcomes of the work that I do. I'm not always privy to the actual data or feedback.” This project was an exception.

The redesigned training ran as a second pilot, structured differently from the first: sessions spaced closer together, time built in to actually practice the new behaviors, and a final session built more like a community of practice than a classroom. It was at that final session that one of the participants came back with a report.

She had been working with a client who was showing warning signs of a dangerous postpartum complication. She talked to the client about seeking medical care, helped her work through whatever real or perceived barriers stood in the way, and the client sought care. Medical professionals later told the her that the client coming in when she did probably saved her life. Waiting might have ended differently.

Heather says, “not every training is going to save a life, but it's pretty cool when it does.” Most of what changes because of a redesign like this can't be measured this cleanly. In this example it could, because a woman decided to seek care instead of waiting.

The Subtle Shift That Changes Everything Else

Heather traces all of it back to a single reframe. “I have started saying in Learning & Development 'We aren't just learning designers.' I think if we focus on the change first, we're actually change agents. And that subtle shift in how we view ourselves makes a big difference in how we show up, what questions we ask. If we can change how we view ourselves within the organization, we'll change everything else.”

Heather's own filter for a program that hasn't started yet comes down to two questions: “What outcomes should this create and how are we going to measure it?” Neither is about the training itself. Both are about the same thing the postpartum project was actually judged on: whether a client's behavior changed, not whether content was delivered.

Nothing about the doulas and midwives project required new software, a bigger budget, or more sophisticated instructional design. It required a practitioner willing to spend time earning the right to ask a harder question, willing to redesign around a behavior instead of a body of knowledge, and willing to build training aligned with the reality of the target group’s working environment. Every one of those decisions followed from the same identity, chosen before the project began.

That has a direct implication for whoever sits above L&D. Approving a training request quickly looks responsive, but sending it back to ask what the actual problem is looks slower. Some stakeholders perceive it as friction, even when it is the more strategic move. If success is still measured by whether training was delivered, rather than whether anyone's behavior changed, there is no real incentive for a practitioner to make that trade. The identity shift Heather describes does not survive an organization that keeps rewarding the old one.

What changed for one nonprofit, one training, and one mother who decided to seek care instead of waiting, started with a question that is not often asked: what is actually driving this, and what would success even look like. Whether that becomes routine or stays rare depends less on L&D's willingness to ask it, and more on whether anyone above them is willing to let the question be asked at all.

Practical Takeaways: Turning L&D Into a Change Agent Function

The single shift that makes L&D strategic is not a better skill set. It is treating "we need a training" as a question to diagnose rather than a request to execute, and designing for what people need to do differently rather than what they need to know. What follows is what that shift requires from HR leadership: what to ask before approving a request, what to expect from the L&D relationship, and what to hold the function accountable for once a program is running.

1. What does it actually mean for an L&D leader to operate as a change agent instead of a learning designer?

The shift begins with what a practitioner treats as the default question. Learning designers start with what people need to know; change agents start with what people need to do differently and what has to shift for that to happen. Heather Burright (Founder, Skill Masters Market) frames this as an identity change, not a new skill. It affects which questions get asked at intake, what gets built, and how the function is evaluated afterward. Leaders investing only in tools or design capability without addressing this identity gap will not get the outcome they expect.

2. What does it look like, in practice, to diagnose a training request before agreeing to build one?

A training request is rarely the real problem; diagnosis means finding what is actually driving it before agreeing to build anything. Heather Burright (Founder, Skill Masters Market) does this by uncovering the underlying pain point, then defining what success would look like if it were solved. In one nonprofit program preparing doulas and midwives to recognize postpartum warning signs, this meant months spent understanding their daily client interactions before any content was built. Skipping this step means the resulting program answers a symptom rather than the real problem, what the first version of that program did before it was redesigned.

3. What is the tell that a training request has already lost the real business problem behind it?

By the time a real business problem reaches the person who will build a solution, it has typically passed through several other people, and each one has simplified it further. What began as an actual issue arrives as a generic request for training, with the original problem already lost. Heather Burright (Founder, Skill Masters Market) treats a flat, generic request as itself a signal, not a finished brief. HR leadership should read the same signal: if a request has no specific problem attached, diagnosis, not execution, is the next step.

4. What happens when a program is designed to transfer knowledge but never designs for the action behind it?

A program can transfer accurate knowledge and still fail to change anything if it stops at what people should know instead of what they need to do differently. Heather Burright (Founder, Skill Masters Market) built her method on that exact distinction: teaching the correct facts does not, by itself, create new behavior. In one nonprofit program that trained doulas and midwives to recognize postpartum warning signs, the first version taught the correct information, but little changed in practice until it was redesigned around the action itself.

5. What has to be true of the L&D relationship before a stakeholder will hand over the real problem instead of a pre-decided solution?

A stakeholder will only hand over the real problem behind a request once enough trust exists to make that honesty safe. Heather Burright (Founder, Skill Masters Market) builds this by leading with relationship rather than a pre-packaged solution, understanding what a stakeholder's own outcomes are before offering anything at all. That credibility accumulates slowly through ordinary interactions and can be undone quickly by one rushed exchange or one pushed solution. Without it, a stakeholder keeps issuing pre-decided requests, and diagnosis never becomes possible.

6. How does designing for a learner's actual working environment change what a training program looks like?

Designing around a learner's actual working environment can rule out an entire delivery format before anything else gets decided. Heather Burright (Founder, Skill Masters Market) chose not to build e-learning for doulas and midwives because their work happens in conversation, not at a desk, and built the training as a conversational format instead, down to using open rather than closed questions. The decision follows from where the learner works, not from a default template. Programs built on a standard format regardless of the learner's setting are a common reason training goes unused.

7. What happens when a stakeholder insists on the original training request rather than allowing diagnosis?

Diagnosis does not guarantee a stakeholder will give up the original ask. Heather Burright (Founder, Skill Masters Market) names this as a real outcome, not a hypothetical: a stakeholder can hold to the training as originally framed even after the pain point has been raised, and the practitioner then has to choose between complying or holding a line that puts the relationship at some risk. For HR leadership, the more useful decision is made earlier, by deciding in advance whether to defend that friction when a stakeholder pushes back, rather than reacting once it happens.

8. What should HR leadership hold L&D accountable for, and what should it check before any program starts?

If success is measured only by whether a session was delivered, the accountability shift Heather Burright (Founder, Skill Masters Market) is describing has no incentive to survive, since the organization keeps rewarding the old model. Heather's own filter for any new program is two questions asked before it starts: what outcome should this create, and how will that be measured. Neither is about the training itself. Applied consistently, both give leadership something concrete to check at approval instead of relying on L&D to self-police the shift.

Closing Reflection

The shift from learning designer to change agent rarely requires new tools, more budget, or better instructional design. It requires a practitioner willing to ask a harder question before building anything, and leadership willing to protect the time that question takes. Whether that becomes routine in your organization depends less on L&D's willingness to ask it, and more on whether leadership rewards the asking.

To hear how the full conversation played out, listen to Heather’s podcast episode.

If you’d like to be informed about more articles like these and community updates, subscribe to our newsletter.

To keep the conversation going with other HR leaders wrestling with the same shift, follow Level Up HR on LinkedIn.