Most transformations succeed… and still fail. Technology milestones are met, systems are deployed as planned, indicators are green. And yet real effectiveness and user adoption sometimes stall.
Technical debt or ill-suited tools usually take the blame. But technology is only the structure of a project; its success depends on its ability to absorb the unpredictability of the real world and of human behavior.
The gap between system performance and customer reality
In programs handling several million transactions a month, systems typically reach high levels of robustness and availability. But a transformation only meets its moment of truth when it reaches the end user — when nominal scenarios collide with thousands of atypical situations every hour.
Even a system that is 99.9% reliable leaves blind spots that add up to thousands of degraded cases. A customer does not perceive a statistical success rate, but a direct failure.
At this scale, the problem is no longer technology. It is how reality is interpreted — and the blind spot of many governance models: steering through global indicators without measuring the accumulation of micro-frictions.
People as the pivot of exceptions in an automated world
The more automated a system becomes, the more critical the human role is. Automation handles the nominal flow; people remain the only safeguard against the exception. In open-flow mobility, exceptions are not marginal — they are structural.
During massive shifts to e-commerce in times of crisis, service continuity relied less on infrastructure than on the ability to mobilize and support hundreds of employees to maintain a bond of trust.
What indicators don’t show
KPIs create an impression of control, but often hide what matters most: misunderstood situations, interrupted journeys, customers lost in the system.
These grey areas act with a delay. A customer facing a degraded experience does not leave immediately: they use less, hesitate to come back, quietly switch to an alternative. KPIs stay green while erosion is already under way — until it reaches loyalty, customer value and, ultimately, business performance.
Toward a living governance model: beyond the spreadsheet
The question becomes organizational: who owns the end-to-end experience, and who arbitrates between system performance and the simplicity users actually experience? Without a clear answer, each team optimizes its own scope at the expense of the overall experience.
Data must be reconnected to reality: confronting indicators with lived situations every day, and processing thousands of concrete feedback items a year so as to stop steering blind.
Conclusion: transformation as a living system
A large-scale transformation is not a fixed project but a living system. Technology makes it work. People make it hold. And that is often where everything is decided.