The Symptom
When a documented procedure stops being followed reliably, the instinct is usually to look for a single cause and fix it. A related pattern, less often examined, is what happens when a departure from procedure has more than one plausible cause at once, and each cause only explains part of the pattern.
Consider an organisation where a supplier delivers late on roughly half of all orders, and the relationship nonetheless continues. On closer examination, the organisation’s own notice period to the supplier was short on roughly half of the late orders. On the other half, notice was not short — and those orders were still late.
What Leaders Usually Conclude
The default read is weak supplier management: a single, simple story in which the supplier is underperforming and the fix is to manage the relationship more firmly. It is a tidy explanation, and it happens to be half right.
Why It Happens
Both explanations — short internal notice, and independent supplier unreliability — hold for part of the pattern, and neither explains all of it. The relationship contains what is, in effect, two separate problems tangled together: a joint problem, where the organisation’s own short notice contributes to lateness, and an independent problem, where the supplier is late regardless of notice given. Because both problems produce the identical visible symptom — a late delivery — they are very easy to treat as a single cause with a single explanation, and very easy to under-diagnose as a result.
This same structure recurs whenever a standard operating procedure is not being followed for reasons that turn out to be mixed rather than singular. Some departures from a documented procedure happen because the procedure itself has become impractical — perhaps it was written for conditions that no longer apply. Others happen for reasons entirely unrelated to the procedure’s design — individual habit, inadequate training, time pressure. When both kinds of departure are occurring simultaneously, at overlapping points, the aggregate pattern of “the procedure isn’t being followed” masks the fact that it is actually two distinct problems layered on top of each other, each requiring a different kind of attention, neither of which will be resolved by addressing only the other. Erik Hollnagel, Robert Wears and Jeffrey Braithwaite make the underlying point in a 2015 white paper from the University of Southern Denmark, the University of Florida and Macquarie University: everyday work is routinely adapted from how it was formally imagined or prescribed.
The reason this goes undiagnosed so often is not a lack of data — the data to separate the two causes usually exists, as it did in the case above, once someone thinks to look for two causes rather than one. The more common failure is simply not looking, because a single tidy explanation is available and feels sufficient the moment it is found.
There is a real cost to leaving a mixed-cause problem misdiagnosed as a single-cause one. Whatever intervention follows a single-cause diagnosis — tighter enforcement, a stronger conversation, a formal warning — will, at best, address only the portion of the pattern the diagnosis was actually correct about. The other portion continues exactly as before, and because some improvement did occur, the intervention is often credited with more success than it actually achieved, making the remaining, undiagnosed cause even harder to notice on a second pass. A partially effective fix, mistaken for a fully effective one, can be more damaging over time than a fix that visibly failed outright, because it removes the pressure to keep investigating while leaving half the original problem untouched.
What It Looks Like in Practice
A procedure or agreement is not being followed consistently, and a single explanation has been adopted without checking whether it accounts for the whole pattern.
The available evidence, once actually examined, only explains part of the instances — with a separate, unexplained portion left over.
Two distinct causal patterns produce an identical surface symptom, making them easy to mistake for one.
A tidy, plausible-sounding explanation was accepted without being tested against the full range of the data available.
A further consequence deserves mention: mixed-cause problems of this kind tend to damage trust between the parties involved in a way single-cause problems do not. Because the supplier in this example is, in fact, partly at fault, a purely internal explanation (short notice) would be unfair to the organisation’s own process failures, while a purely external explanation would be unfair to the supplier for the portion of lateness genuinely outside their control. Adopting either single-cause story in isolation tends to produce a relationship conversation that neither side can fully agree with, which is its own quiet source of friction, independent of the original operational problem.
For how this pattern presents in other kinds of organisation, see other sectors.
Questions worth considering
- Where a procedure or agreement is not being followed consistently, has the current explanation actually been checked against every instance, or does it just sound plausible?
- Could there be two separate causes producing what looks like a single pattern?
- What data already exists that could separate one explanation from another, if anyone looked?
- When a problem has two plausible causes, what stops an organisation looking for both?
This perspective is part of Operations.
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.