Interactive visualization of stochastic noise in organizations. See why adding 'just one more' process creates exponential chaos.
The gap between what you plan and what actually happens isn't because you don't plan well enough, or because your teams aren't aligned. It's because you're ignoring stochastic noise. It's because over-imposing layered goals, top 10 prio-lists, structures, processes, rules, roles, and adding dependencies exponentially grows your complexity to the point where it's virtually impossible for you to build great products, deliver on time, or maintain low operational costs.
Stochastic noise is the inherent, irreducible randomness embedded in systems with interdependent actors and fluctuating inputs. It's what makes prediction fragile, and variance unavoidable. Complex and over-constrained systems have more noise. This noise is the source of unwanted variation. The more noise you have, the more unwanted variation you get.
While you're getting delays, the wrong problems being solved, fragmentation, duplication, that's the effect of the noise. What the noise might look like first hand is an incident interrupting the team, an API becoming unavailable, a team member becoming sick, new information making old decisions seem odd, people being tied up in artificial processes or events with no bearing on the work, two teams having completely different data sources in direct conflict, a person not being available to make a decision, and so on.
Simply put: Your outcome is a function of your current state, the actions you take, and noise. Most leaders respond to unwanted variation by adding more processes, more meetings, more approval gates. But this actually makes things exponentially worse. Because every attempt to constrain offers additional error points in your system.
This is why the CEO of Ford praised Tesla in 2023/2024, because Tesla had not only conducted an aggressive disintermediation, they also optimized for throughput, flow efficiency, and to the point where their cars, work processes, and factory setups were all configured with this in mind. This at a time when Ford, and every other car producer had intermediated their process, and now found it extremely difficult to make changes, improve quality, or make small pivots. The amount of error points in their value chain were exponentially more.
The work of Dave Snowden, Esther Derby, Simon Wardley, Steven Bungay, Don Reinertsen is effective, in different ways, of cutting through the noise, filtering for desirable behaviors, and avoiding adding more noise.
This interactive model below shows you why "just one more" process creates multiplicative effects that push teams past their productive capacity. Play around with the configuration to see how noise adds, multiplies, and grows exponentially.
Note: This simulator is an interactive illustration meant to visualize the noise that we can't see, but that's there. This is not meant to be a decision making framework. It's not a PHD thesis!
Mathematical foundations and real-world applications of stochastic systems in organizations.
Viktor Cessan • Enterprise Agile and Product Coach