Winnicott developed the concept across three decades of clinical work with mothers and infants at Paddington Green. His insight was that the cultural ideal of the perfectly attuned mother was producing its own pathology — mothers paralyzed by fear of failing their children, children deprived of the developmental benefits that come from adequately-managed frustration. The 'good enough' reframe was meant to liberate both from an impossible standard.
The parallel to contemporary organizational discourse is precise. The AI transition has produced its own impossible standards: the demand that every organization perfectly support every employee through an unprecedented transformation, the expectation that every worker will adapt smoothly if only the right change-management protocol is deployed, the fantasy that with sufficient preparation the disruption will be experienced as growth rather than loss. These standards are as unhelpful to organizations and workers as the perfect-mother ideal was to parents and children.
The good-enough organization is specifically possible to achieve. It requires none of the impossible elements — perfect foresight, unlimited resources, complete equity. It requires only: recognition that disruption is real, acknowledgment that grief is appropriate, provision of sufficient time and support that the attachment system can register safety, and consistent communication through actions rather than words that the organization-worker relationship will survive the transformation. Most organizations are failing at this not because the standard is too high but because they are pursuing a different, impossible standard that distracts them from the achievable one.
The analogous concept for AI tools — the good-enough tool — may be equally important and equally absent from current development. Tools designed to be perfectly responsive, always available, never frustrating, may be producing the same pathology that perfect-mother ideals produced in children: users whose working models are not developing the capacity to tolerate the ambiguity and imperfection of reality because the tool has protected them from that experience. The good-enough AI tool might be the one that fails visibly, asks for clarification, acknowledges its limits, and refuses to perform omniscience.
Winnicott introduced the concept in his 1953 paper 'Transitional Objects and Transitional Phenomena' and elaborated it across subsequent writings, particularly Playing and Reality (1971) and The Maturational Processes and the Facilitating Environment (1965).
The concept has been extended into organizational psychology through the Tavistock tradition and into clinical work with a wide range of populations through the British Independent tradition of psychoanalysis.
Imperfection is developmental. The good-enough caregiver's manageable failures produce growth that perfect responsiveness would prevent.
Reliable not perfect. What matters is the reliability of adequate response, not the achievement of perfect response.
Action over rhetoric. The good-enough caregiver communicates security through consistent behavior rather than through reassurance.
Scales to organizations. The framework applies directly to organizational life: the good-enough employer is achievable; the perfect employer is fantasy that distracts from the achievable standard.
May apply to tools. Good-enough AI tools — those that acknowledge limits, fail visibly, and require judgment from their users — may produce healthier relationships than tools designed for perfect responsiveness.