Plateau Review Cadence¶

A recurring review ritual — instantiates Plateau Detection and Switching

A standing review at a fixed interval where a team looks at each active path and decides to keep escalating, gather more evidence, or switch — so plateau calls happen on schedule, not only in a crisis.

Plateau Review Cadence is the time-driven counterpart to an event-driven stop. It is a recurring ritual — a standing slot on the calendar — where a team deliberately looks at every path still receiving effort and asks of each: keep going, gather more evidence, or switch? Its distinguishing feature is rhythm. A threshold trip catches a fast collapse, but many plateaus creep in slowly and never cross a hard line cleanly; the cadence catches those by forcing a periodic, deliberate look whether or not anything alarming has happened. The interval itself is the observation window — long enough that each path has had a fair stretch to show movement, short enough that a spent path isn't fed for another two quarters unexamined.

Example¶

A product-engineering org runs a monthly Initiative Review. Every active initiative appears on the same board with its marginal-progress trend since last month and a one-word ask: keep, evidence, or switch. This month, a performance-optimization initiative shows a third straight month of flattening — a slow bend, never sharp enough to trip an automatic stop. In the crisis-free calm of a scheduled review, that slow flattening is visible next to its peers, and the room applies the standing rule: three flat cycles moves an initiative from keep to evidence-or-switch by default. The team doesn't kill it on the spot; the cadence's job is to surface it and force the conversation, which lands on running a small test before committing to a rebuild. Without the ritual, the initiative would have kept its funding by inertia, because nothing ever forced anyone to look.

How it works¶

Fix the interval as the window. The review period doubles as the observation window — each path is judged over a full, consistent stretch, not a cherry-picked snapshot.
Review every active path, not just the troubled ones. Bringing all paths to the same table is what surfaces the slow plateaus that never raise an alarm.
Apply a standing default rule. A pre-agreed rule (e.g., N flat cycles ⇒ evidence-or-switch) removes the need to re-argue from scratch and blunts optimism.
Resolve each path to a branch. Every path leaves the review labeled keep, gather-evidence, or switch — the ritual's output is decisions, not discussion.

Tuning parameters¶

Cadence length — how often the review runs. Frequent reviews catch plateaus sooner but add overhead and can react to a single noisy period; sparse reviews are cheap but let spent paths run.
Default-rule strictness — how many flat cycles auto-escalate a path. Strict defaults act decisively but on thin evidence; lenient ones give paths room but tolerate drift.
Scope of the review — all initiatives vs. only flagged ones. Reviewing everything catches slow creep at the cost of meeting length.
Decision bindingness — whether the review's call is advisory or committing. Binding calls prevent backsliding; advisory ones preserve flexibility but let inertia win.

When it helps, and when it misleads¶

Its strength is catching the slow plateau — the gradual flattening that never trips a threshold and so would otherwise be funded by inertia indefinitely. The regular, low-drama slot also makes stopping a normal act rather than an admission of failure, which lowers the political cost of switching.

Its failure mode is ritual hollowing: a review that becomes a status-reading formality where everything is nodded through as "keep," or, at the other extreme, a churn-machine that switches paths every cycle before any has had a fair trial. It can also become theater — a meeting that documents decisions already made elsewhere. The discipline is a real standing default rule the room must actively overrule (not silently ignore), a cadence matched to how long the response genuinely takes to move, and a bias toward evidence as the middle branch so switching isn't the only alternative to drift.

How it implements the components¶

Plateau Review Cadence fills the periodic-governance subset — scheduled judgment rather than automatic enforcement:

observation_window — the review interval is the window over which each path's plateau evidence is gathered and judged.
escalation_stop_rule — the standing default rule that, on schedule, moves a flat path off "keep more of the same."
switch_trigger — the review is the moment where accumulated plateau evidence is turned into a keep/evidence/switch decision.

It does not compute the marginal signal (marginal_response_metric → Diminishing Returns Detection), nor design the replacement path (alternative_strategy → Strategy Switch Decision Tree); it is the scheduled forum where those inputs become decisions.

Instantiates: Plateau Detection and Switching — it is the recurring forum that keeps plateau decisions from being deferred indefinitely.
Sibling mechanisms: Escalation Stop Workflow · Diminishing Returns Detection · Marginal Gain Dashboard · Controlled Experiment After Plateau · Ad Fatigue Switching · Training Plateau Adjustment · Product Growth Plateau Response · Process Redesign After Plateau · Saturation-Aware Resource Allocation · Strategy Switch Decision Tree · Expert-Governed Modality Change

Notes¶

The scheduled half of a pair: the cadence revisits paths on a fixed rhythm, whereas Escalation Stop Workflow fires the instant a stop threshold is crossed. The two are complementary, not redundant — run both so neither slow creep nor fast collapse slips through.