Process Consultation

In the AI Story

The methodology emerged from Schein's MIT work in the 1960s with the National Training Laboratories and his subsequent consulting engagements. It was radical for its time: while dominant management consulting operated through expert diagnosis and prescriptive recommendation, Schein insisted that the client's own sense-making was the locus of genuine change. The consultant's job was to create conditions under which the client could see clearly, not to see on the client's behalf.

The relevance to AI adoption is direct. Current AI adoption typically operates through training programs (artifact level), strategic communications (espoused values level), and performance mandates (which increase anxiety without addressing its causes). None of these approaches touch the level at which basic underlying assumptions live — the level at which the transformation must occur.

Process consultation proceeds through specific practices: observing rather than prescribing, asking rather than telling, creating the helping relationship within which the client can tolerate examining assumptions she has not previously examined. These practices are slow. They do not fit the quarterly cadence of organizational performance measurement. They require a quality of attention that is alien to the metrics-driven culture of most contemporary organizations.

The methodology has been misunderstood as non-directive — as consulting without substance. Schein was explicit that this was wrong. Process consultation is highly substantive; its substance is different from that of expert consulting. The substance is in the quality of the questions asked, the accuracy of the observations shared, and the conditions created for the client's own insight. The consultant is not passive. The consultant is actively helping, just not actively telling.

Origin

Schein developed process consultation at MIT beginning in the 1960s, drawing on his training in clinical psychology and his experience with T-groups at the National Training Laboratories. The first edition of Process Consultation appeared in 1969, with substantially revised editions in 1988 and 1999. The methodology became the foundation for Schein's later work on humble inquiry and humble consulting.

Key Ideas

Help rather than tell. The consultant's job is to create conditions for the client's own insight, not to deliver the insight on the client's behalf.

The client's sense-making is the locus of change. Change that does not occur in the client's own understanding does not occur at all.

Observation and questioning replace diagnosis and prescription. The methodology's practices are specific and demanding, and they contradict the dominant model of expert consulting.

The helping relationship is the instrument. Without genuine relationship, the questions fail — they become procedure rather than inquiry.

AI adoption requires this methodology. The transition operates at the level of basic assumptions, and only process-consultation practices can surface and revise them.

Debates & Critiques

Expert-consulting traditions have long criticized process consultation as insufficiently directive, too slow, and poorly suited to situations requiring urgent decisions. Schein responded that the urgency is typically produced by the same assumption-level failures that process consultation is designed to address — that the choice between process consultation and expert consultation is often the choice between addressing root causes slowly and addressing symptoms quickly. The AI moment has intensified this debate: the pressure for rapid action is enormous, and the time for assumption-level work is exactly what rapid action foreclose.