In this talk, Andy Campbell challenges the default habit in learning and development of jumping straight to building a training course whenever a stakeholder says “we need training.” He argues that designers should first diagnose why people can’t perform their jobs—identifying barriers such as missing information, confusing processes, broken tools, unclear expectations, or environmental constraints. By understanding the real performance problem, designers can choose the most effective intervention, which might be a quick walkthrough, just‑in‑time help, practice with feedback, manager reinforcement, or even fixing the underlying process, rather than automatically creating a course. Campbell stresses that learning is messy, ongoing, and embedded in work, not a single event that ends with a quiz and a green check‑mark in an LMS. He advocates treating learning as a system of support that spans before, during, and after performance, and encourages designers to ask probing, “annoying” questions to uncover the true needs before deciding on any solution. The talk ends with a call to shift the designer’s role from course‑builder to performance‑enabler, using the right tool—whether that’s a course or something else—to help people actually do their jobs.
Chapters
Here's something we do in learning and development that drives me a little nuts. Somebody comes to us and says, we need training, and what do we do? We start designing the training almost immediately. We start thinking, okay, what should the course look like? How long should it be? Is this e-learning? Do we need a video? Maybe we should add a scenario. Maybe some interactions, and before you know it, we're 20 minutes into talking about the course and nobody has asked, why can't people do this right now? That's the question I care about. I'll admit it, I didn't always think this way. Earlier in my career, if you asked me to build training, I was excited to build training because building stuff is fun. I still love building stuff. Give me a weird design problem and I will happily disappear down that rabbit hole. But that's also the trap, because you can build a really good solution to the wrong problem. And I've become much more skeptical when somebody tells me, we need training. Maybe, but let's figure that out first, because sometimes people already know what to do. They just can't find the information when they actually need it. Sometimes the process is confusing. Sometimes expectations are crap. Sometimes the stupid tool sucks. Sometimes people know exactly what they're supposed to do, but the environment makes doing it way harder than it needs to be. And I can build you the prettiest 30-minute e-learning course you've ever seen, and it's not going to fix any of that. That's probably one of the biggest challenges in how I think about this work now. I don't think my job is to make courses. My job is to help people perform. And those sound similar. They're not. Because if my job is to make a course, I've already decided what the solution is before I've really understood the problem. But if my job is to help somebody perform, now I've got options. Maybe they need practice. Maybe they need feedback. Maybe they need A 5-minute walkthrough. Maybe they need something they can pull up while they're actually doing the work. Maybe their manager needs to reinforce something. Or maybe, and this one might be my favorite, we need to fix the crappy process instead of teaching people how to survive it. That happens way more than I think we like to admit. And look, I'm not anti-course. I build courses. Sometimes a course is exactly what we need, but I want that course to earn its place in the solution. I don't want to build one just because somebody filled out a form somewhere and typed, we need training. There's another reason I've started thinking this way. Think about how you actually learn something complicated, not how our learning models say you're supposed to learn it. How you actually learn it. You hear about something, you try it, you screw it up, you look something up, you ask somebody, you try again, you get a little better. Then 3 weeks later, something weird happens, and you think, crap, how did I do this again? So you look it up, you figure it out. And you keep going. That's learning. It's messy. It happens over time. And it happens while we're doing the work. But then we turn around and design learning like this: take a course, pass a quiz, click complete, congratulations, you've learned the thing. Except that little green checkmark in the LMS doesn't mean somebody can actually do the job. It means they completed something. Those are not the same thing. And that's why I've started thinking about learning less like an event and more like a system. If I'm designing something now, I want to know what happens before this person has to perform? What happens when they actually do it? What happens when they get stuck? Where can they practice? Who gives them feedback? Why can't they reference later? Or what can they reference later? What happens 3 weeks from now when they forget half of this crap? And how do all of these pieces fit together? To me, that's design. That's a much more interesting problem than deciding whether the next button belongs in the bottom right-hand corner. And it changes our role. Because now we're not just sitting around waiting for somebody to hand us a course request. We get to investigate. We get to ask annoying questions. We get to say, okay, you're asking for training. Help me understand what's happening. What are people doing today? What should they be doing? What's getting in the way? What happens if we do nothing? And sometimes after asking those questions, I'll probably still say, yep, we need training. Cool, let's totally build the crap out of it. But sometimes we're going to discover that the thing everybody was performing to spend 6 weeks building wasn't actually going to solve the problem. And that, that is the kind of designer I want to become. Nice. Not somebody who's really good at making courses, somebody who's really good at figuring out what people need to order in order to perform. So the next time somebody comes to you and says, we need training, try not to open an authoring tool. Don't start sketching screens. Don't start worrying about the next button. Just ask, what's stopping people from doing this today? Start there. Then figure out what deserves to be built. Sometimes it'll be a course, sometimes it'll be something completely different, and sometimes the best learning experience you can design won't look like training at all.
