The Signal Was Never the Problem: Diagnosing the channel before blaming the noise

This book is about the channel, not the signal. When a feedback loop misbehaves, the instinct is to filter harder, gather more readings, smooth the noise out of the numbers. I once watched a fabrication job do exactly that. A team spent months tightening a quality metric, adding data every week, getting a cleaner and cleaner reading. The feedback path had been wired to the wrong measure from the start. The cleaner the signal got, the more confidently it reported on a variable nobody actually cared about. That case is a composite drawn from my own work, and I am stating that plainly.

A closed loop can only ever be as right as the variable its feedback path measures. No amount of filtering fixes a loop connected to the wrong thing. Garbage in stays garbage out no matter how good the processing in between. So the question I keep returning to is not how noisy your results are. It is whether the wiring behind them was ever pointed at the right node.

The hard part is that a mis-wired channel hardens. Conway's law says an organization ships a design shaped like its own reporting structure, so a broken channel does not stay a technical problem. It becomes the product. Fix the structure of the loop, or keep polishing an output that was answering the wrong question all along.

For readers working through feedback loops, signal versus noise, control theory, network topology, root cause analysis, and Conway's law, this is a short book about diagnosing the channel before blaming the signal.

This English edition was written from an Indonesian operator's context, with examples and figures drawn from running small businesses in West Java; the diagnostics apply broadly to small and mid-sized firms in other emerging markets and to many developed-market settings as well.