Before/After Constraint Monitoring¶
Metric / dashboard — instantiates Bottleneck Capacity Shadowing
Tracks, after a relief action, whether performance actually moved, where the new limiting constraint appeared, and whether the gain leaked downstream.
Before/After Constraint Monitoring is the post-hoc, empirical instrument of the archetype. Every other mechanism here estimates before acting; this one watches after. Once a relief action lands, it tracks the live metrics over time to answer three questions the up-front models cannot: did the objective actually improve, where did the binding constraint migrate to, and did relieving one stage merely shove overload into the next. Its defining trait is that it measures realized outcomes rather than predicted ones — it is backward-looking and observational, the reality check that keeps a shadow value from going stale. It exists because relief changes the system: the old bottleneck may stop binding, a new one appears, and continued investment in the old one becomes waste the moment the constraint moves.
Example¶
A state unemployment-insurance agency, told that claim review was the binding constraint, hired twenty additional caseworkers. Before/After Constraint Monitoring dashboards the result week over week against the pre-hire trend. Throughput rises about 15% and then plateaus — the dashboard shows the queue has migrated to identity verification, which is now where claims stack up. At the same time it watches the downstream stages, and the first-pass error rate ticks up: faster review has pushed more imperfect claims into appeals and correction, adding rework the celebration would have missed. The monitor's verdict is concrete: stop adding caseworkers, whose marginal value has fallen toward zero, redirect relief to identity verification, and close the quality leak before it grows. Without the "after" view, the agency would have kept funding a constraint that no longer binds.
How it works¶
- Instrument both sides. Measure the objective at the relieved stage and at candidate downstream stages, capturing a "before" window as well as "after."
- Compare against a baseline trend, not a single point. Judge the change against the pre-relief trajectory (or an untouched comparison unit), so noise and natural drift aren't credited to the intervention.
- Watch where the wait moves. Rising queueing or wait time at a new stage flags the migrated binding constraint.
- Audit downstream. Track errors, rework, and spillover to catch relief that improved one stage by degrading the next.
Tuning parameters¶
- Observation window — long enough to separate real signal from noise, short enough to act on; too brief and a lucky week reads as success.
- Baseline model — a single before-measurement versus a trend line or control group; the latter is the main guard against crediting a naturally-easing queue.
- Downstream breadth — how many downstream stages to instrument; wider coverage catches more spillover but costs more to maintain.
- Migration sensitivity — how large a shift in wait time counts as evidence the binding constraint has moved.
- Alert cadence — a continuously refreshed dashboard versus a periodic review; faster feedback against ceremony and alarm fatigue.
When it helps, and when it misleads¶
Its strength is catching what every ex-ante estimate is blind to: constraint migration and system-level leakage. It is the mechanism that stops a team from re-investing in yesterday's bottleneck and that verifies a relief action delivered the gain its shadow value promised.
Its central failure mode is misattribution — crediting the relief for an improvement that was really regression to the mean[1], a seasonal dip, or some co-occurring change. A short window and a single before-point make this almost inevitable; a queue that spiked and would have subsided anyway gets read as a triumph of the intervention. The classic misuse is declaring victory on the first good week. The discipline that guards against it is to judge the "after" against a baseline trend or an untouched comparison unit, hold the observation window open long enough for signal to separate from noise, and always instrument downstream so a local win that is really a system-level loss can't hide.
How it implements the components¶
constraint_migration_monitor— tracks where the binding constraint moved once relief landed, so the next decision targets the new limiter rather than the old one.downstream_effect_audit— instruments the stages after the relieved one to catch relief that merely pushed overload along instead of removing it.
It does not estimate value before acting — it computes no dual price and certifies no binding status from a model (shadow_value, binding_constraint) — that's Shadow Price Analysis; and unlike Constraint Sensitivity Report, which predicts response by varying a model over a range (baseline_solution, marginal_response_range, relief_cost_model), this instrument measures the response actually realized after a real relief action. It does not lay out or rank the relief options (relaxation_option, relief_priority) — that's Bottleneck Valuation Map.
Related¶
- Instantiates: Bottleneck Capacity Shadowing — the monitor closes the archetype's loop, feeding realized outcomes back into the next relief decision.
- Consumes: Shadow Price Analysis supplies the predicted marginal value this monitor checks the realized outcome against.
- Sibling mechanisms: Shadow Price Analysis · Constraint Sensitivity Report · Bottleneck Valuation Map · Capacity Investment Analysis · Marginal Capacity Value Review
Editorial Notes¶
Form Classification¶
Form family: Monitoring, Sensing & Alerting
Rationale: Tracks, after a relief action, whether performance actually moved, where the new limiting constraint appeared, and whether the gain leaked downstream, making its operative form repeated observation of actual state that emits measurements, status, or alerts.
Independent corroboration: The frozen evidence defines Before/After Constraint Monitoring as 'Tracks, after a relief action, whether performance actually moved, where the new limiting constraint appeared, and whether the gain leaked downstream', so its operative form is Monitoring, Sensing & Alerting.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Constraint and throughput analysis checks whether relieving the binding stage increased system performance or merely moved the bottleneck.
Related originating lineages:
- Organizational & Management Science — Theory of Constraints operationalizes repeated identification and exploitation of the current constraint.
- Statistics & Experimental Design — Baseline trends and controls guard against regression-to-mean attribution.
Review resolution: Operations research is the agreed primary lineage through constraint and throughput analysis. Organizational management operationalizes the Theory of Constraints and experimental design guards the pre/post attribution, so both are retained as formative alternates.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
This is the mechanism the archetype names as its protection against stale shadow values. A shadow price is only true until the relief that acts on it changes the system; the "after" view is what tells you the estimate has expired. Skipping it is the quiet way a disciplined relief program drifts into repeatedly optimizing a constraint that stopped binding some quarters ago.
References¶
[1] Shadish, W. R., Cook, T. D., & Campbell, D. T. Experimental and Quasi-Experimental Designs for Generalized Causal Inference. Houghton Mifflin (2002). Supports regression to the mean and concurrent events as alternative explanations for an apparent before/after improvement. registry ↩