Looking at those four options of the poll that I created yesterday, I now notice they're all about diagnosis. And they matter. But I'm curious about what happens before the drop-off. Because usually the person isn't gone the day they decide to leave. They're already half-gone. The design made it easy to drop off is the one that interests me because it's the only one about what you built. The other three are about them. And yeah, people are busy and unmotivated, and we can't change that immediately. But what if we start by focusing on what we can change in that very moment? We can change the friction. What if the first thing that signals drop-off isn't a reason at all, but a moment where the structure stops pulling them forward? Not because the content is bad. But because nothing is asking them to come back. Which of these resonates most with what you're seeing in your world?

When someone in your organization stops engaging with a learning program, what's usually the first assumption people make about why? I ask because the assumption shapes the fix. And most fixes I see are aimed at A and B, while C and D are doing most of the actual work. What's your experience?

We believe we're giving our children everything. Safety. Stability. The best schools. Experiences we never had ourselves. A safety net that catches them before they ever fall. And that's precisely the problem. The human organism isn't optimised for comfort. It's optimised for adaptation. It grows through friction: through failure, through the moment when no one steps in and you have to find the answer yourself. Nassim Taleb calls this antifragility: the property of systems that don't just survive disruption, but become stronger because of it. Bones that are never loaded become brittle. Immune systems that are never challenged overreact. And people who are never allowed to truly fail learn the wrong lesson – that the world will always catch them. We mean well. That's not the question. The question is: what happens to a generation raised in an environment so carefully managed that real challenge has to be artificially reintroduced? I think we'll find out over the next 20 years.

It’s the ones that reduce the most cognitive load to get you started and then switch to 'structured uncertainty' to keep you staying. That’s why pure designers tend not to like behavioral designers. We might not win the most design awards, but we move the metrics that actually matter. For today, I would like to motivate you to think about the concept of 'structured uncertainty': Many of the systems we design in organizations try to eliminate uncertainty. We create detailed plans, predictable processes, step-by-step instructions, tightly defined KPIs, and learning programs where every step is known in advance. The intention is good. Predictability feels safe. Control feels professional. But something strange happens when everything becomes too predictable: Attention disappears. When the brain already knows what will happen next, it quietly switches to autopilot. We stop truly engaging with the situation. We go through the motions. Work becomes execution instead of exploration. That is why many environments that look perfectly optimized on paper slowly drain energy from the people inside them. Adoption collapses! Structured uncertainty works differently. Instead of removing uncertainty completely, you introduce just enough of it to keep the brain alert and curious. The path is not fully known, but it is still navigable. The challenge is real, but not overwhelming. Think about good games, good research, good entrepreneurship, good hobbies or even good conversations. None of them are fully predictable. Yet they are not chaotic either. They operate in a zone where the next step requires attention, judgment, and a bit of courage. This is where engagement lives. Not in perfect clarity, and not in total randomness, but in environments where the future is partially open and our actions genuinely matter. The uncomfortable question for many organizations is this: Have we designed our systems to produce performance, or have we designed them mainly to eliminate uncertainty for managers?