The Bench Read · A Worked Demonstration

ASKED · “A coaching read on one store’s leadership team — who to coach, on what, and what is store process rather than person.” — SYNTHETIC · every store, name, and number on this page is fictional, modeled on a real engagement.

The store’s promise failure is two-thirds store process and one-third one shift leader — and every graded leader on the bench is trending up.

The demonstration store — call it Store 21 — keeps fewer hot-shelf promises than any of its fleet’s forty stores, and the question for its GM is where a week of attention actually moves the number. Graded against the store’s own hour-by-hour break rates — a staffing-fair expectation, not a fleet league — the answer splits cleanly: one shift leader carries the largest recoverable slice, and the store’s midday process fails under every leader who works it. Both facts are true at once; a league table shows neither.

Hold one leader’s hours at his peers’ rates and the store still fails — the counterfactual recovers a third of the gap, no more.

Ten fiscal weeks · one store, six leaders on watch · synthetic demonstration data

Store promises kept

57.8%

Lowest of 40 stores; fleet median 83.9%

Recoverable / week

24.1

If leader M.O.’s hours ran at peers’ rates

Leader share of gap

7.4 of 22.8 pp

Remainder is store-level process

Graded leaders

4 of 6

Two below the volume floor — not graded

Grade each leader against the store’s own clock, then say both halves out loud.

Each leader’s promise volume is scored against the store’s own break rate for the same hour and weekday cells — so nobody is punished for drawing the lunch rush, and nobody hides behind it. The counterfactual then re-prices the flagged leader’s cells at his peers’ pooled rates: 24.1 recoverable promises a week, store 57.8% → 65.2% — a third of the 22.8-point gap to its own market’s median store (80.6%). The rest belongs to the store’s midday process, and it fails under everyone.

Leader	Role	Hours led	Kept vs expected	Keyed comps	Trend · first 3 vs last 3 wks
M.O.	Shift leader	228	31.5% vs 47.2%	4 vs 12.6 exp	28.9 → 39.8
L.V.	Manager	221	72.1% vs 59.7%	26 vs 24.1 exp	64.0 → 79.5
S.B.	Shift leader	198	54.6% vs 55.0%	29 vs 15.2 exp	51.2 → 60.7
J.R.	Shift leader	164	63.8% vs 60.9%	6 vs 14.3 exp	60.1 → 66.9
T.K.	Shift leader	71	53.4% vs 52.6%	2 vs 5.5 exp	4 wks — n/a
C.D.	Manager	6	not yet graded — below the volume floor

Why this grade is fair enough to say out loud. A fleet league punishes whoever works the hardest hours; a raw kept-rate punishes whoever works at the weakest store. The staffing-fair expectation does neither — it asks only “did this leader’s watches beat the rates this store itself runs in these same hours?” — and then reports the absolute number beside it so a failing store cannot grade itself on a curve. The counterfactual is stated with its limit in the same breath: fixing the flagged leader recovers a third of the gap, and the store still fails. That sentence is what keeps a bench read from becoming a blame instrument.

What a read like this cannot say: whether a flagged leader’s gap is skill or circumstance (the split-half persistence test narrows it, never settles it); what any individual comp was for; and anything at all about leaders below the volume floor. Flag counts are not defect counts — the grading runs with no multiplicity correction, and it says so.

Provenance — This is a worked demonstration: every store, leader, and number on this page is synthetic, authored to be internally consistent (the counterfactual, the daypart cuts, and the staffing fit all reconcile against the same fictional ledger), and modeled on the shape of a real ten-week engagement for a multi-unit pizza franchisee. In production, the same read runs on watch-grain shift attribution from the operator’s timeclock joined to order-grain POS promises — per-leader hours, own-store hour-cell expectations, keyed-comp and waste lanes, volume floors, and split-half persistence, exactly as shown. Names are initials by design. Prepared by SixTop as a demonstration of the bench instrument.