Business leaders love pendulum metaphors. What goes up comes down. Markets swing from greed to fear. Organisations oscillate between centralisation and autonomy.
It's intuitive. It's comforting. And for simple systems, it mostly worked.
But look at the venture capital cycle since 2020: ZIRP-era growth obsession → sustainable growth after rate hikes → back to aggressive scaling driven by AI. Pendulum, right?
Wrong. Each "swing" moved the baseline. The growth obsession of 2024 feels different from 2021's - more desperate, more extractive. The brief focus on "sustainable growth" didn't reset anything; it just taught everyone to optimise for different metrics.
This isn't oscillation. It's what complexity scientists call a ratchet effect - systems that appear cyclical but actually compound stress with each turn, moving the baseline further each time.
Then there's technical debt. Teams promise to "pay it down next quarter" but keep adding shortcuts faster than they can clean up. Eventually the rate of debt accumulation outpaces the team's ability to manage it, and the codebase tips into a state where simple changes become impossibly slow. That's a rate-induced tipping point - when the speed of change overwhelms a system's adaptation capacity.
I encountered these concepts during my complexity science masters (15 years ago - time flies). They're part of a broader toolkit for understanding systems that don't behave like simple machines. As ecosystems get more complex, we need mental models that match that complexity.
Understanding the New Dynamics
Some systems ratchet, accumulating irreversible shifts. Others tip, breaking suddenly when change outruns adaptation. The skill of leadership is telling which dynamic you're facing.
The Ratchet Effect in Tech
We see the ratchet effect everywhere. Once a company adopts a specific cloud platform, each new service adds to the cumulative cost and effort required to migrate elsewhere. This "ratchets up" the switching costs with every new turn of the development cycle.
Another powerful example is the "gravitational pull" of a monolithic business platform. When a company adopts a platform (e.g., for CRM or ERP) and begins to dump data into it without a strong, disciplined integration strategy, each new service, custom field, or domain built on it compounds the influence. The platform starts cannibalising other tech stacks and eventually dictates the company's IT strategy, rather than serving it. The initial choice of a single platform for a single purpose morphs into an irreversible system that controls the organisation's technological destiny.
In org culture, the "always on" mentality can become so embedded that a leader can't just declare it over. The baseline of expected commitment has moved permanently.
The key for leaders isn't to fight the ratchet but to understand it. What seemingly small decisions are compounding to create irreversible complexity?
When Systems Tip
The rate-induced tipping point is also highly visible. You can track this in your own engineering organisation using a simple framework. When deployment frequency and change failure rate start moving in the wrong direction, i.e., deployment frequency decreases while change failure rate increases, coupled with an increasing feature backlog, it's a sign that your system is close to a tipping point. The rate of change is overwhelming your team’s ability to stabilise the system.
Rapid growth can trigger a cultural tipping point as well. A small, trusting startup culture can tip into a state of siloed, guarded communication when the sheer rate of new hires overwhelms the informal bonds that held the team together. The old cultural norms can't adapt fast enough.
The New Role of Leadership
Linear thinking worked when systems were simpler and changes were slower. But when you're operating in highly networked, fast-changing environments - whether that's modern software development, distributed organisations, or AI-accelerated markets - you need frameworks that account for compound effects, feedback loops, and irreversible change.
The pendulum is an outdated metaphor. Complexity science offers dozens of frameworks for understanding systems that don't behave like simple machines - from network effects and emergence to phase transitions and adaptive cycles. Ratchets and tipping points are only two of many mental models that explain the world we’re actually in. As our organisations and markets become more complex, we should be mining this field for better mental models. What other mental models might help us see the irreversible shifts happening in your domain? What patterns are you seeing that the traditional business frameworks can't quite capture?