The observation
Most organisations measure outcomes: the number that resulted. Fewer measure the conditions that reliably produce those outcomes, which is a different and harder thing to track. Research by Nelson Repenning and John Sterman at MIT Sloan School of Management, published in California Management Review, used system dynamics modelling and more than a dozen case studies to examine why process-improvement efforts so often fail to hold. Their finding was that organisations under pressure default to “working harder” — applying more attention, more effort, more senior oversight — rather than “working smarter,” which means changing the underlying conditions that produce the result. Working harder produces a real, visible, often impressive improvement in the outcome measure. It does not, on its own, change anything about the system that will operate once the extra attention moves elsewhere, which is precisely when the gain tends to fade.
This distinction — between an outcome measure and a measure of the conditions producing it — is one of the more consequential ones in how organisations track their own performance, and one of the least commonly made explicit. An outcome measure answers “how did we do?” A leading measure, closer to the conditions Repenning and Sterman describe, answers a different question: “is what we are doing sustainable without the current level of attention?” Most performance dashboards are built almost entirely from the first kind. The second kind is harder to define, harder to standardise, and consequently rarer — which means most organisations have excellent visibility into results and comparatively poor visibility into whether those results are durable.
Why it happens
The bias toward outcome measures over condition measures is not an oversight; it is a natural consequence of what is easy to count. An outcome — a number of units produced, a percentage on time, a cost ratio — is usually a single figure, collected automatically, comparable period to period. The conditions that sustain an outcome — whether the underlying process still requires senior intervention to function, whether the people doing the work have the capacity the work actually requires, whether a fix depended on one person’s attention rather than being built into how the work operates — are harder to reduce to a single number, and often require judgement rather than a count.
Repenning and Sterman’s research adds a second, sharper insight: this is not merely a measurement gap, it is a reinforcing loop. When a process-improvement effort succeeds through added attention, the visible result is celebrated, which is entirely reasonable — the number did improve. But the celebration typically closes the loop before anyone asks what made the improvement work. Attention, once redirected to the next pressing priority — as it eventually always is, because attention is finite and organisations have many competing demands on it — leaves the underlying conditions exactly as they were before the effort began. The measure does not fail gradually. It returns to its previous level roughly as quickly as the attention departed, which often makes the reversal look sudden and surprising even though, in structural terms, nothing unpredictable has happened.
A related misconception worth naming directly: leaders frequently interpret a faded gain as evidence that the improvement effort itself was superficial or poorly executed. The research suggests a more specific and less flattering explanation. The improvement was often entirely real. What was missing was a mechanism for sustaining it once the person or team who drove it moved on to something else — which is a different failure than the effort having been hollow in the first place, and it calls for a different kind of scrutiny than simply trying harder next time.
What it looks like in practice
Consider an organisation where a focused effort improves a key operational measure by roughly a third — a genuinely strong result by any standard. Nine months after the effort formally ends, the measure has returned to close to its previous level. The first and most natural explanation offered is that the improvement was superficial, or that the effort was wound down too early.
A closer look at what actually drove the improvement complicates that story. The gain was real, and it was driven substantially by a specific change: a senior manager began reviewing cases personally, on a weekly basis, throughout the effort. That review caught issues before they compounded, kept the work to a consistent standard, and created a visible incentive for everyone downstream to maintain that standard while it was being watched. When the manager’s attention moved on to the next priority, as it eventually did, the underlying conditions that had produced the improvement — the review cadence, the visible scrutiny, the consistent standard-setting — moved with it. Nothing about the work itself had changed. The gain had been produced by attention layered on top of the work, not by any change to how the work operated, and attention, unlike a process change, does not scale or persist by default.
This pattern is common enough to be recognisable across very different functions and industries: a metric jumps during a period of unusual senior focus, holds for as long as that focus continues, and reverts once the focus is redirected — often interpreted, incorrectly, as the initiative “not sticking,” when the more precise description is that the initiative was never built to operate without the attention that launched it.
Looking for your own sector? The Other Sectors page points to the closest match.
Questions worth considering
- Was this improvement built into how the work actually operates, or is it currently being held up by someone’s ongoing attention?
- If the person currently driving this result stopped paying attention to it tomorrow, what would the measure look like in three months?
- Which of the organisation’s current dashboards measure outcomes only, and which — if any — measure the conditions that make those outcomes durable?
More on this topic: Performance.
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.