Most improvement programs address one area in isolation. Durable, sustained improvement requires all five categories to be designed and installed together — or targeted precisely where the constraint lives. This article describes the model and how each category interacts.
Laboratory improvement programs have a pattern. A consultant arrives, identifies a constraint — usually in throughput, turnaround time, or quality — and installs a solution. Results improve. The consultant leaves. Within twelve months, performance has drifted back to where it started, or close to it.
The reason is almost never that the solution was wrong. It is that the solution was installed in isolation. A new scheduling system without a management cadence to govern it. A standard work program without a performance measurement system to sustain it. A quality initiative without the operating rhythm to embed it into daily practice. Individual interventions, however well-designed, do not hold without the surrounding system.
An operating system, in the laboratory context, is the set of structures, routines, and controls that govern how work is planned, executed, measured, and improved — every day, at every level of the organization. It is not a project. It is not a set of SOPs. It is the architecture that makes execution repeatable and improvement sustainable.
The Meridian House model organises this architecture into five interdependent categories. Each category addresses a distinct dimension of laboratory performance. Together, they form a complete operating system. The categories are not sequential — they interact continuously — but they are distinct enough to be designed, assessed, and improved independently. The full system can be installed in a single engagement, or we can work on one category, a combination, or wherever the constraint is most limiting. Every engagement is scoped to the problem.
Operating architecture is the structural foundation of the laboratory — how work flows, how capacity is understood, and how demand variation is absorbed. It encompasses flow, leveling, capacity planning, and constraint management. Without this architecture, even well-designed processes drift under pressure.
Leveling means smoothing the arrival of work so it reaches the bench at a predictable, manageable rate. Flow means removing bottlenecks and handoff delays so samples move continuously through the value stream. A capacity model does not need to be complex — it needs to be accurate, current, and used. When demand is leveled and flow is optimised, the constraint becomes easier to see and easier to manage.
Execution discipline is the set of operating habits and controls that ensure the right work is done the right way, every time. It encompasses standard work, visual management, workplace organization, and escalation routines. Execution discipline is not about compliance — it is about making the best-known method the default, and making deviation visible before it compounds.
Standard work is the most misunderstood element of this category. In most laboratories, it is treated as a documentation exercise — SOPs written to satisfy an audit, filed in a binder, and rarely consulted. Execution-framed standard work is different: it is written for the person doing the task, located at the point of use, and supported by a control system that makes deviation visible in real time.
Performance management is the management cadence and accountability structure that keeps every level of the organization aligned, informed, and acting on the right information. It encompasses KPIs, daily management, structured problem solving, and accountability cadence. Without a defined rhythm, performance drifts — even when the underlying processes are sound.
A well-designed daily management system is tiered: a brief daily huddle at the bench level, a weekly performance review at the supervisor level, and a monthly operational review at the management level. Each tier has a defined agenda, a defined escalation path, and a defined set of metrics. The cadence is the heartbeat of the system.
Technology enablement is the digital layer of the operating system — LIMS optimization, automation, data analytics, and digital workflow support. The critical design principle is that technology must serve the operating system, not substitute for it. A dashboard that shows throughput data is only useful if there is a management cadence that reviews it and an escalation protocol that responds to it.
Visual management begins with physical boards at the bench and extends through digital dashboards to AI-assisted analytics. The purpose is to give teams the signals they need to self-correct before problems escalate, and to give leaders line-of-sight into where work is stuck, where demand is unleveled, and where capacity is underused. Digital tools without an operating system to govern their use produce information without action.
In most laboratories, quality and operations are managed in parallel. The quality function runs its own processes — CAPA, audit preparation, corrective actions — while operations runs its own. This separation is the root cause of a significant proportion of quality failures. When quality is not embedded in the daily operating rhythm, problems accumulate before they are visible.
Quality integration means embedding CAPA processes, audit readiness, and right-first-time controls into the daily operating rhythm — so quality is a property of execution, not a separate function. CAPA cycles shorten because the underlying operating conditions that generated the problem are addressed directly. Audit readiness becomes a continuous state rather than a periodic scramble.
The five categories are interdependent — and they are grounded in four proven lean principles adapted for the laboratory: leveling, flow, standard work, and visual management. Operating architecture creates the structural conditions for predictable flow. Execution discipline makes reliability the default. Performance management creates the forum in which the system is reviewed and acted upon. Technology enablement makes the whole system visible. And quality integration embeds compliance into execution so it is sustained without external pressure.
"Individual interventions, however well-designed, do not hold without the surrounding system."
This interdependence is why piecemeal improvement programs fail to sustain. A new scheduling system installed without a management cadence to govern it will drift. A quality initiative installed without standard work to embed it will not change daily practice. The layers must be designed together, even if they are installed sequentially.
The right starting point depends on where the primary constraint is. In most laboratories, the first constraint to address is the operating rhythm — because without a functioning management cadence, improvements in any other layer will not be sustained. But in some environments, the capacity model is so poorly understood that it must be built first, simply to establish a shared understanding of the problem.
The diagnostic question is not which category is weakest, but which category's weakness is most limiting overall performance. That is the right starting point. The full system can be installed in a single engagement, or we can work on one category, a combination, or wherever the constraint is most limiting. The right ending point is always the same: a complete system — all five categories designed, installed, and owned by the internal team.
Next step
Start with a diagnostic conversation. No pre-packaged proposals. No junior teams.