Some outcomes that look like successes are failures in disguise — and the institutional learning they prevent is as consequential as the failure they appear to avert.
Successful Outcomes That Mask Failures
Some institutional outcomes that are recorded as successes represent the fortuitous convergence of good luck with bad process rather than the reliable production of good outcomes through sound practice. The project that was delivered on time despite fundamental failures in planning, coordination, and risk management — because a series of fortunate events compensated for the failures — records as a success. The safety-critical operation that produced no harm despite systematic violations of safety protocols — because none of the violations happened to trigger the failure mode they were designed to prevent — records as a success. In both cases, the success is real: the intended outcome was achieved. The failure is also real: the process that was supposed to reliably produce the intended outcome was not the process that actually produced it.
The harm of the failure masked by success is the learning it prevents. The institution that records the successful outcome as a validation of its processes when those processes were in fact bypassed or failed has reinforced its confidence in processes that are less reliable than the outcome suggests. The next iteration will again deploy the failed processes, without the fortunate events that compensated for their failures on the prior occasion, and will produce the outcome the processes actually generate rather than the outcome that luck produced last time.
Identifying Success-Masked Failures
Identifying the failure masked by success requires analysis of process rather than outcome — examining how the intended result was achieved rather than simply recording that it was achieved. This process analysis is not naturally produced by outcome-focused performance management systems, which typically stop their investigation at the successful outcome without asking how it was produced. Producing it requires deliberate post-event analysis that specifically examines the gap between the intended process and the actual process, even when — especially when — the outcome was satisfactory.
The failure masked by success is the most dangerous kind, because it prevents the learning that would address it while reinforcing the confidence that makes the learning seem unnecessary. The institution that only investigates failures will never learn from the successes that were actually failures in hiding.
Discussion