Organisations often declare success at visible milestones: system launched, structure announced, training delivered, communication executed. Momentum peaks at go-live.
But integration begins where launch ends. Completion is event-based. Integration is behavioural. And behavioural stabilisation requires structural consistency.
Why go-live feels definitive
Launch milestones are measurable. They signal progress, demonstrate movement, and validate investment. Executives need markers of achievement. Delivery teams need closure points. So go-live becomes the symbolic finish line.
But people do not reorganise their behaviour on symbolic timelines. They reorganise when incentives, authority, and expectation align consistently over time.1
Integration requires three stabilising conditions: clarity of decision rights, alignment of incentives, and consistency of reinforcement. If any of these remain ambiguous, behaviour will oscillate. People comply when visible, revert under pressure, and improvise locally when constraints conflict.2
The system appears partially adopted. But partial adoption is not integration.
Why instability resurfaces post-launch
Post-launch instability often surprises leadership. Adoption dips, variance increases, and old patterns reappear. This is interpreted as fatigue or insufficient reinforcement — and sometimes it is.
But often, launch exposed structural contradictions that were temporarily suppressed during implementation. Under delivery pressure, teams aligned around deadlines. After go-live, they return to operating under normal incentives and authority patterns. If those patterns contradict the change, behaviour regresses.3
If integration instability is treated as engagement variance, more communication is deployed, more oversight is introduced, and additional reporting layers are added. These interventions increase load. Load amplifies variance. Variance triggers further intervention.
Without architectural recalibration, the loop strengthens. The organisation concludes the change “didn’t stick.” In reality, the structure never fully supported it.
What disciplined integration looks like
Disciplined integration asks: are incentives reinforcing the new behaviour? Are performance metrics aligned with stated priorities? Have decision rights shifted where necessary? Is legacy structure still rewarding prior patterns?
Integration is not about sustaining enthusiasm. It is about aligning design. When structure and behaviour reinforce each other, integration stabilises naturally. When they do not, reinforcement must be continuous and exhausting.4
Declaring completion too early shifts attention away from architectural consistency. Launch becomes the proof point. But stabilisation requires continued coherence across sponsorship, operating model, measurement, and practitioner stance.
This is one way of understanding how integration depends on the layers explored throughout this series — from diagnostic integrity to governance clarity.
-
Feldman, M. S., & Pentland, B. T. (2003). “Reconceptualizing Organizational Routines as a Source of Flexibility and Change.” Administrative Science Quarterly, 48(1), 94–118. https://doi.org/10.2307/3556620. Feldman and Pentland distinguish the ostensive dimension of a routine (the abstract pattern, the launched system) from its performative dimension (actual enactment in practice). Behavioural stabilisation occurs through the performative dimension — through repeated enactment that gradually displaces prior patterns. Go-live is an ostensive event; it announces the new routine but does not enact it. Integration requires the performative dimension to actually take hold, which depends on structural consistency across incentives, authority, and expectation, not on the symbolic timeline of the launch milestone. ↩︎
-
Simons, R. (1995). Levers of Control: How Managers Use Innovative Control Systems to Drive Strategic Renewal. Harvard Business School Press. Simons’ framework of diagnostic and interactive control systems explains the compliance/reversion pattern precisely. Organisations that manage post-launch integration diagnostically — monitoring metrics against targets — create the conditions for oscillation: metrics are met when visible because monitoring creates temporary alignment, but the structural ambiguities that produce reversion (unclear decision rights, misaligned incentives) remain intact between monitoring cycles. Interactive control, which surfaces the tensions that require governance attention, is what prevents oscillation; diagnostic control alone cannot substitute for structural clarity. ↩︎
-
Sterman, J. D. (2000). Business Dynamics: Systems Thinking and Modeling for a Complex World. McGraw-Hill. Sterman’s analysis of feedback delays and policy resistance demonstrates that systems under artificial pressure — delivery deadlines, programme milestones — generate temporary alignment that masks underlying structural tensions. When the pressure is removed, feedback loops that were operating in the background reassert themselves. Post-launch instability is the predictable consequence of this dynamic: the deadline pressure that produced apparent pre-launch alignment disappears at go-live, and the structural contradictions that were suppressed by that pressure become visible in behaviour. This is not a surprise; it is a delayed feedback loop completing its cycle. ↩︎
-
Tushman, M. L., & Romanelli, E. (1985). “Organizational Evolution: A Metamorphosis Model of Convergence and Reorientation.” In L. L. Cummings & B. M. Staw (Eds.), Research in Organizational Behavior, Vol. 7, pp. 171–222. JAI Press. Tushman and Romanelli’s model of convergence periods establishes that structural alignment is the mechanism by which organisations maintain stable performance between reorientations. When structural elements — strategy, architecture, people, and processes — are mutually reinforcing, the system maintains equilibrium without continuous intervention. When they are not, the organisation expends increasing effort to maintain surface stability against the structural pull toward prior patterns. The “continuous and exhausting reinforcement” that characterises incomplete integration is not a feature of change difficulty; it is the energy cost of misalignment. ↩︎