Complexity Isn't the Enemy; Ignoring It Is

I’ve just spent two days at World of Learning, and I came away with big thoughts, small thoughts, challenging thoughts, and a fair few frustrations. I’ll share more over the coming days, but one conversation kept surfacing in different guises, with different people, in different sessions, and I want to start there.

Everything is more complicated than most of us are willing to admit.

That was the thread running through almost every meaningful exchange I had across both days. Whether it was about AI adoption, culture change, skills strategy, or workforce planning, the same conclusion kept presenting itself: the problems organisations are grappling with right now are deeply interconnected, layered on top of other still-active changes, and resistant to the neat, singular solutions that so much of the industry would love to offer. This isn’t about the volume or velocity of change, which we’ve been discussing for years now, but about how multi-faceted each change is when it lands inside an organisation that’s already mid-way through several others.

This creates an interesting tension with some of the offerings in the expo hall. I should say, though, that there was an inspiring number of organisations willing to engage with the complexity of the problems they were trying to solve, openly acknowledging that their tool, technology, or service couldn’t possibly encompass the entire solution. One day, that might be everyone’s stance; not today, but it’s progress.

What does this mean for us? A few things worth sitting with.

First, it reinforces the case for systems thinking. In complex situations, it’s more important than ever that we understand what connects to what, and what those connections connect to in turn. That means the organisation, the leadership, the people, the policies, the processes, and the customers. It was encouraging to hear systems thinking getting a few mentions across sessions and in corridor conversations, because that’s what this calls for: a willingness to look at the interconnected web of systems that form an organisation rather than assuming we can change one thing without affecting everything around it.

Second, it strengthens the argument for experiment-led approaches. When we’re operating in highly complex, highly changeable environments, we won’t always have all the information we’d like before making a decision. We need to get comfortable making decisions within uncertainty, and that means making much smaller plays. As I’ve discussed at length in previous articles on experiment-led design, uncertain environments demand that we resist the urge to attempt one enormous fix and instead make small, deliberate changes, observe what happens, and iterate. This is the old principle of TNTs (tiny, noticeable things with a big impact), where we can use them to affect and cascade change across systems, reducing risk rather than compounding it.

Third, and perhaps most importantly, it highlights the need for us to know our organisations deeply and continuously. We can’t afford to wait until someone asks us to create an intervention before we start looking at performance data, business figures, attrition, revenue, or whatever matters most in your context. We need to be across the data, all of it, all of the time, so that when a problem arises, we already understand everything around it and which levers we can pull.

Note: When I say “all of it all the time,” I am well aware of being excessively absolute. This should be our goal. We should be aiming to be aware of everything we can possibly know from every part of the organisation, all of the time. We will fail 100% of the time doing this, but by making this the goal, we’ll hopefully get close and at least not be blindsided by performance challenges and change requests that come our way.