Designing for Cultural Change

Why This Guide Exists

You know something could be better. Maybe it's how your module is assessed. Maybe it's how your department uses technology. Maybe it's something as fundamental as confronting what your students are actually learning versus what the curriculum says they're learning.

You've probably tried to change it. You've raised it in meetings, sent the email, maybe even presented at a teaching conference. And nothing happened. Or worse, everyone agreed it was a great idea and then carried on exactly as before. Because change is hard.

This guide is about what you can do to make change happen.

It draws on research from network science, behavioural science, and systems theory to explain why institutions resist change, why most change efforts fail, and what the evidence says about how to give your ideas the best possible start. It won't promise an instant revolution, but it will help you design an approach that grows and becomes more impactful over time.

1. Why the System Fights Back

In 1884, the French chemist Henri Louis Le Chatelier discovered something peculiar about chemical reactions: they resist interference. If you change the conditions of a system in equilibrium (add heat, increase pressure) the system adjusts to counteract that change and restore balance. It's not malicious. It's just what systems do.

This principle extends well beyond chemistry. Complex human systems — universities, healthcare organisations, governments — behave in remarkably similar ways. They develop their own equilibrium over time, and when you try to change them, they push back. Not because the people within them are obstructionist (though we all know those individuals who are!), but because the system itself has settled into patterns that resist disruption.

John Gall, a paediatrician and systems theorist, spent decades observing why human systems so often coexist happily with the problems they were designed to solve. His conclusion, articulated in his book Systemantics (later republished as The Systems Bible), is uncomfortable but important: systems tend to oppose their own proper functioning. They grow. They encroach. They develop their own logic that may have nothing to do with their original purpose. And critically, they kick back when you try to change them.

Gall's most useful insight, however, is a positive one. Known as Gall's Law, it explains how successful complex systems actually come into being:

Think about that in the context of your institution. Every successful practice you've seen, the module that students rave about, the team that collaborates well, the assessment approach that produces genuine learning, almost certainly started as something small and simple that someone tried, refined, and grew over time. It didn't emerge from a strategic plan.

The implication is profound: you cannot redesign a broken complex system from the inside. Adding more complexity to fix complexity makes things worse. What works is building something small and functional alongside the existing system and letting it prove itself.

Ed Bradon, writing in Works in Progress magazine, traces this pattern across domains from the first electric grid (a handful of lamps connected to a water wheel in Godalming, England, in 1881) to Estonia's digital government, which began with electronic ID for signing documents and filing tax returns before expanding over two decades to cover voting, prescriptions, policing, and nearly every public service. At no point did someone design the finished product from scratch. It evolved from simple systems that worked.

The lesson for anyone trying to change educational practice is clear: don't try to reform the whole curriculum. Don't redesign the assessment strategy for an entire programme. Start with something small that works, in your own module or with your own team, and let it grow.

2. Why Broadcasting Doesn't Work

Most institutions approach change the way you'd spread news: broadcast it. Send an all-staff email. Put it in the newsletter. Get a senior leader to endorse it at a town hall meeting. The assumption is that if people know about the change, they'll adopt it.

Damon Centola's research in network science reveals why this assumption is wrong. Centola draws a crucial distinction between spreading information and spreading behaviour change. These are fundamentally different processes that require fundamentally different approaches.

Information spreads like a virus through weak ties and broad networks. One person tells another, who tells another. A single exposure is often enough. This is why news travels fast, why rumours spread, and why viral or influencer marketing occasionally works. It's the model that most change management approaches are built on, often implicitly.

But behaviour change doesn't work like that. Changing how someone actually does their job — how they teach, how they assess, how they use technology — requires something much more demanding than awareness. It requires social reinforcement. People need to see others like them adopting the change. They need multiple exposures from multiple trusted sources. They need to feel that the new behaviour is becoming normal within their peer group, not just endorsed from above.

Centola's research identifies several important findings that challenge conventional thinking about change:

The people who spread viruses are bad at spreading cultural change

In epidemiology, highly connected individuals (so-called "super-spreaders") are crucial for transmission. But for behaviour change, these central influencers are largely ineffective. They have too many weak connections and not enough deep ones. They follow trends rather than creating them.

One-way transmission doesn't work

A single instruction from a manager, a single presentation at a conference, a single endorsement from a Vice-Chancellor — none of these are sufficient. Behaviour change requires peer-to-peer influence through multiple reinforcing interactions within connected social groups.

Awareness outpacing adoption is dangerous

When everyone has heard about the change, but they see that almost no one they know has adopted it, you've created the worst possible conditions. People now know about the idea and have already decided, consciously or not, that it's not for them. The idea has been inoculated against. It's much harder to revive than something that was never broadcast in the first place.

3. What Actually Works

If broadcasting from the centre doesn't work, what does? Centola's research, combined with Gall's Law, points to a set of principles that are remarkably consistent across different contexts.

Start on the Periphery

The most effective change doesn't begin at the centre of an organisation's network, with the Dean, the Head of School, or the most prominent professor. It begins at the periphery, with small, connected groups, who have enough autonomy to try something different and enough connection to each other to reinforce the new practice.

In the 1960s, South Korea faced a population boom that would have overwhelmed the country within twenty years. Other countries like Pakistan, facing similar challenges, used broadcast approaches — national advertising campaigns, celebrity endorsements, etc. — with limited success. South Korea instead targeted female social groups in villages. They didn't prescribe a specific solution; they communicated the importance of family planning and let communities choose their own approach. Within twenty years, they had achieved their demographic targets, and interestingly, different regions adopted different methods. The change was locally owned, peer-reinforced, and sustained.

Build Wide Bridges

Centola's research emphasises the importance of "wide bridges" — connections between different social clusters that involve multiple overlapping relationships, rather than a single point of contact. A narrow bridge is one person who happens to know someone in another department. A wide bridge is when several members of one team have genuine working relationships with several members of another.

Wide bridges matter because behaviour change requires reinforcement from multiple sources. If only one person in your department has tried the new assessment approach, their colleagues can dismiss it as an individual quirk. If three or four people across two departments are doing it, it starts to feel like a real trend.

This means that the most valuable people for spreading change aren't the most well-connected individuals, they're the people who sit between groups. The lecturer who teaches across two programmes. The learning designer who works with multiple schools. The colleague who's involved in both a research cluster and a teaching committee. These are your bridge-builders.

Protect Time and Space

Change takes time to take hold. Centola's research shows that sustained social norms emerge when communities have protected time and space to develop them. Rushing adoption, such as mandating a new practice before people have had time to experiment with it, observe others doing it, and make it their own, undermines the very conditions that make change stick.

This is why mandated change has a short lifespan. During the Covid-19 pandemic, mask-wearing was widely adopted under mandate but rapidly abandoned when mandates were lifted. The behaviour was never socially reinforced to the point where it became self-sustaining.

4. Designing Your Approach

The following principles aren't a step-by-step recipe (that would be its own form of the systems thinking hubris that Gall warns against!), they're a set of design considerations to help you give your idea the best possible conditions to take root.

Find Your First Allies

Look for two or three colleagues who share your frustration or curiosity. Not necessarily the most senior or the most influential but the most genuinely interested. Ideally, they should be connected to different groups within your institution so that when they adopt the change, it becomes visible across multiple clusters rather than contained within one.

Start with Something That Works

Apply Gall's Law. Don't propose a programme-wide redesign. Build the smallest possible version of your idea and make it work in your own context first. If you want to change how assessment works, change how one assessment works in one module. If you want to introduce a new technology, use it yourself for a term before suggesting it to anyone else. The working prototype is worth more than any proposal document.

Make Benefits Visible Without Mandating Adoption

Share what you're doing in the spaces where your colleagues naturally congregate (staff rooms, corridor conversations, existing meeting agendas) but don't push for adoption. Let people observe the change from a distance before participating. Centola's research shows that people are community-minded by evolution, but they need to see social proof before they commit to new behaviours. Around 25% adoption is the tipping point: until roughly a quarter of a community has adopted a new practice, it hasn't reached the threshold for self-sustaining change.

Frame as Enhancement, Not Accommodation

This is about narrative. Nobody wants to hear that what they're currently doing is wrong. Frame your change as something that makes their existing practice better, not something that replaces it. The persuasion hierarchy developed by advertising strategist Dave Trott is useful here: first get their attention (Impact), then make the message meaningful to them (Message), then make it frictionless for them to act (Persuasion). If the message sounds like criticism, it will be rejected regardless of how valid it is.

Use the EAST Framework

The EAST framework, developed by the UK's Behavioural Insights Team, provides a simple checklist for designing interventions that align with how people actually make decisions. When introducing a new practice, ask:

Is it Easy?

Have you removed friction? Can someone try this with minimal effort and disruption to their existing workflow? If adopting your idea requires learning new software, attending training, filling in forms, or getting approval, you've made it too hard. Strip away everything non-essential.

Is it Attractive?

Does it look good? Does it feel like something worth doing? This isn't about marketing spin, it's about making the benefits concrete and visible. Show the student feedback. Share the time saved. Demonstrate the outcome.

Is it Social?

Can people see others doing it? Are there opportunities to discuss, collaborate, or compare experiences? Change that happens in isolation stays in isolation. Create the conditions for peer-to-peer sharing, even if it's just a regular coffee conversation between three people.

Is it Timely?

Are you introducing this when people are most receptive? The start of a new academic year, the redesign of a module, the arrival of a new cohort — these are natural moments of openness. Mid-semester, when everyone is drowning in marking? Not so much. Timing is everything.

5. When to Push and When to Wait

Centola's research reveals that change doesn't follow the smooth bell curve of Rogers' Diffusion of Innovation model that you'll find in most management textbooks. The reality is closer to a steep S-curve with a tipping point at around 25% adoption.

What this means in practice is that change happens slowly, then suddenly. The early phase, when you and your first allies are experimenting, refining, and sharing, will feel frustratingly slow. It's supposed to. This is the phase where the foundations are being laid: peer connections are forming, social proof is accumulating, and the practice is being adapted to fit the local context. If you try to accelerate this phase by broadcasting or mandating, you risk letting awareness outpace adoption and inoculating people against the idea.

It's also worth reconsidering the people who don't adopt your change immediately. The Diffusion of Innovation model calls them "laggards," which implies stubbornness or backwardness. Centola's work suggests something more sympathetic: many of these people actually lack confidence. They're scared of being left behind, not opposed to progress. They need to see that the change is safe, normal, and supported before they'll commit. Treating them with empathy rather than impatience isn't just kinder, it's more effective.

Once you've reached the tipping point — roughly 25% adoption — the dynamics shift. Now the influencers and senior leaders who were irrelevant in the early stages become useful for amplification. Now is the time for the all-staff email, the conference presentation, the strategy document. Not to create change, but to recognise and codify change that has already happened. The narrative shifts from "here's something new you should try" to "here's what we're already doing."

6. References & Further Reading

• Centola, D. (2021). Change: How to Make Big Things Happen. Little, Brown Spark. — The essential text on how social change actually spreads through networks, and why conventional approaches fail.
• Gall, J. (2002). The Systems Bible: The Beginner's Guide to Systems Large and Small. General Systemantics Press. — Originally published in 1975 as Systemantics. Witty, brief, and genuinely useful on why systems behave the way they do.
• Bradon, E. (2025). 'Magical Systems Thinking.' Works in Progress. Available at: worksinprogress.co — An excellent argument against top-down systems design, with compelling examples of Gall's Law in practice.
• The Behavioural Insights Team. (2014). EAST: Four Simple Ways to Apply Behavioural Insights. Available at: bi.team — The original EAST framework publication.
• Trott, D. (2015). One Plus One Equals Three: A Masterclass in Creative Thinking. Macmillan. — On the persuasion hierarchy and creative problem solving.

Acknowledgements

This guide was co-developed with Morna, a custom AI agent built on Claude and designed using the Creative Compass by David Calum Millar.

With special thanks to Dr Ross McKenzie, who pointed to the article by Ed Bradon that sparked this guide: Magical Systems Thinking (Works in Progress, 2025).