Why Efficient Businesses Break First
Efficiency has never had to defend itself. For most of the last century, it was assumed to be good, in the same unreflective way punctuality is assumed to be a virtue. Streamline the process. Remove the waste. Standardise the exception until it stops being one. The businesses that did this fastest became the ones every other business was told to study.
Nobody asked what all that streamlining was removing along the way.
An efficient system is built to behave exactly as designed, every time, without surprise. That is precisely what makes it dangerous the moment the environment stops cooperating. Markets no longer move in the tidy cycles that made careful optimisation such a reliable strategy. Customer expectations shift mid-quarter. Artificial intelligence erases competitive advantage faster than planning calendars can register. Geopolitical shocks, regulatory upheaval, and technological disruption used to occur sequentially; now they overlap.
Most organisations are still running the old playbook: optimise for stability, and treat any deviation from plan as a problem to correct rather than information to use.
A cross-trained employee who could, in principle, cover someone else's job looks like redundancy on a headcount spreadsheet, right up until the week someone leaves and the work doesn't stop. A budget line for an idea with no guaranteed return looks indefensible in a quarterly review, right up until the year the guaranteed-return ideas stop returning. Most finance functions are trained to treat both as waste, because both fail the only test efficiency knows how to run: does this get used every time?
There's a reason that test misses the point. In 1956, the cybernetician W. Ross Ashby described something he called requisite variety: a system can only absorb as much disturbance as it has variety of response to meet it with. A thermostat with two settings can only ever manage two kinds of weather. A business that spends a decade eliminating every deviation, every redundant process, every "unnecessary" alternative approach has done the same thing to itself without quite meaning to. It has traded away the very variety it would need to survive whatever it hadn't planned for.
None of this is an argument against discipline, or for wasting money on purpose. It is an argument that some of what looks like waste is actually the organisation's stored variety, and that the businesses who protect a small, deliberate amount of redundancy and unproven experimentation from the scrutiny that rightly governs everything else have understood something the spreadsheet can't measure: that part of the business was never meant to be efficient. It was built to be ready.
Financial performance tells one story about a business. Customer behaviour tells a different one, often at the same moment. Employee sentiment, competitive movement and cultural shift each add a further angle, and none of them agrees with the others often enough to be comfortable. Most reporting structures resolve that discomfort by picking one measure and getting increasingly precise about it, mistaking precision for truth.
Early twentieth-century painters working in Andalucian tradition ran into the same problem from a different direction, and refused the same shortcut. They were not trying to make art more difficult. They were testing whether a single fixed viewpoint could ever describe a subject completely, and concluded it could not: a face seen only from the front is a face with half its information missing. Painting it from several angles at once, and holding all of them inside the same frame, was not a stylistic flourish. It was an argument about how much a single vantage point leaves out.
Long-lived organisations are rarely the ones that got obsessively precise about a single number. What they share instead is a trained tolerance for holding several contradictory readings at once and acting anyway, the same discipline the Cubists were practising a century before anyone applied it to a quarterly board pack.
Throughout business history, disruption was like weather—something happened, leadership responded, and operations returned to normal. That model is breaking, and not for the reason most leadership decks suggest. Amy Edmondson and Michaela Kerrissey, writing in Harvard Business Review, draw a useful distinction between a crisis, the kind with a beginning and an end, and what they call a sustained crisis: a period of ongoing difficulty with no clear return to normal on the horizon, because normal is no longer where the business is heading back to.
That distinction shifts the goal. If uncertainty is permanent, resilience isn't about quick recovery but about being prepared beforehand. This moves the focus to an earlier, less dramatic stage than a crisis, like a routine Tuesday, when minor improvements happen without questioning their cost.
Every process designed for efficiency teaches what success looks like, whether intended or not. Rewarding speed and predictability repeatedly makes people stop questioning assumptions and see experimentation as risky. No one chooses to be less adaptable, but organisations become experts at repeating what worked last time, mistaking repetition for strength.
Getting overtaken rarely comes down to a lack of discipline. Most of the businesses it happens to have discipline in abundance. What they're missing is a second lens: the willingness to keep asking whether the thing they perfected is still the thing worth perfecting.
McKinsey's twenty-year research shows change is eight times more likely when leaders visibly behave differently, people are trained in new skills, and systems reward it, compared to just making an announcement. None of that happens from a stage. It happens the Tuesday afternoon someone's experiment fails, and their manager either asks what they learned or moves them onto something safer without saying why. That's the real vote on whether curiosity is welcome, and it gets cast in rooms no memo ever reaches.
When repeated enough, a vote becomes culture, which influences how an organisation reacts when faced with new market questions. Efficiency determines what is precisely repeated at scale without change. Resilience decides what gets left deliberately unsettled, because certainty there would be the costlier mistake. A business needs both operating at once, from different parts of the same organisation, and building that is a harder design problem than any process map has ever solved.
The more useful strategic question for the next decade is not how to optimise what already exists. It is whether today's strengths are becoming tomorrow's blind spot, one well-rewarded, perfectly efficient habit at a time.
