Saturation-Aware Resource Allocation¶

A resource-allocation policy — instantiates Plateau Detection and Switching

Moves budget or effort off a channel whose marginal return has saturated and onto a less-saturated one — allocating by where the next unit buys the most, not by where the spend already sits.

Saturation-Aware Resource Allocation is the archetype applied across a portfolio of paths at once. Where a single-path switcher rotates or redesigns one plateaued channel, this mechanism's distinguishing feature is that it never designs a new path — it redistributes a fixed pool of budget or effort across existing paths according to their marginal returns. Its governing rule is comparative: a plateau on one channel matters only relative to what the same unit would earn elsewhere, so it continuously shifts resource off saturated paths and onto ones with steeper slope until the next unit buys roughly the same everywhere. The plateau signal here is not "stop" but "reallocate" — the forgone return on the saturated path is exactly the resource the fresher path is starving for.

Example¶

A hospital runs seasonal flu-shot outreach across a dozen neighborhoods, with a fixed outreach budget. Early on, every neighborhood's spend lifts vaccination rates. By mid-season three early-targeted neighborhoods have saturated — their marginal uptake per outreach dollar has flattened near zero — while three later-reached neighborhoods are still climbing steeply. Saturation-Aware Resource Allocation reads the portfolio, not any single path: it flags that a dollar in the saturated neighborhoods now buys ≈0 additional shots while a dollar in the fresh ones buys meaningfully more, and shifts budget accordingly. It doesn't invent a new outreach method — that would be a different mechanism — it moves the existing method's budget to where its marginal return is highest. Total shots delivered rise from the same budget, because resource stopped piling into paths that had stopped responding.

How it works¶

Compare marginal returns across paths. Every channel's next-unit yield is put side by side, so allocation is driven by relative slope, not by history or habit.
Read each path's saturation. A flattening marginal curve marks a path as saturated and its resource as a candidate to move.
Price the opportunity cost of staying. The gap between a saturated path's next-unit yield and the best alternative's is the cost of leaving resource where it sits — and the size of the case for moving it.
Shift toward equal marginal return. Reallocate until the next unit buys roughly the same on every funded path; that balance is the stopping point.

Tuning parameters¶

Reallocation frequency — how often budget is rebalanced. Frequent shifts track saturation closely but incur switching costs and churn; infrequent ones are stable but let imbalance persist.
Move size — how much resource shifts per adjustment. Big moves rebalance fast but can overshoot a channel's absorptive capacity; small moves are safe but slow.
Minimum-floor policy — whether saturated paths keep a maintenance allocation. Floors preserve a channel's efficient core and avoid abandonment; zero floors maximize marginal efficiency but can let a path decay.
Comparison baseline — whether "best alternative" is the single steepest path or a diversified spread. Concentration maximizes marginal return; spreading hedges against a mis-estimated curve.

When it helps, and when it misleads¶

Its strength is turning a plateau into a reallocation rather than a loss: instead of stopping at a saturated channel, it puts the forgone resource to work where the marginal return is highest — the equimarginal principle, that a fixed budget yields the most when the last unit spent on each path returns the same.^[1]

Its failure mode is chasing noisy marginal estimates — over-reacting to a temporary dip and yanking budget from a path that would have recovered — and ignoring lag, so a slow-maturing channel looks saturated before its returns arrive. Stripping a path to zero can also forfeit fixed setup or relationship value that made it cheap to keep alive. The classic misuse is dressing up a favoured channel's budget grab as "efficiency," reverse-justifying a reallocation someone already wanted. The discipline is to smooth the marginal estimates before acting, respect each path's response lag, and keep a maintenance floor where abandonment is costly to reverse.

How it implements the components¶

Saturation-Aware Resource Allocation fills the portfolio-reallocation subset — it redistributes across paths by comparative marginal return:

opportunity_cost_signal — its core: the gap between a saturated path's next-unit yield and the best alternative's, which quantifies the cost of leaving resource in place.
capacity_context_signal — reads which paths have saturated (flat marginal curve) versus which still have absorptive slope.
input_pathway_scope — defines the set of channels being compared, so the reallocation is over a bounded, like-for-like portfolio.

It does not design a new intervention or rotate a single channel's creative (alternative_strategy → Ad Fatigue Switching, Product Growth Plateau Response); it moves resource among existing paths rather than inventing one.

Instantiates: Plateau Detection and Switching — it is the portfolio-reallocation instance of stop-and-switch.
Consumes: Diminishing Returns Detection supplies the per-path marginal curves it compares.
Sibling mechanisms: Ad Fatigue Switching · Product Growth Plateau Response · Diminishing Returns Detection · Marginal Gain Dashboard · Controlled Experiment After Plateau · Escalation Stop Workflow · Plateau Review Cadence · Training Plateau Adjustment · Process Redesign After Plateau · Strategy Switch Decision Tree · Expert-Governed Modality Change

References¶

[1] The equimarginal principle — a fixed resource pool yields the greatest total return when it is allocated so that the marginal return of the last unit is equal across all paths. A saturated path is one whose marginal return has fallen below that of alternatives, signalling reallocation. ↩