The Proof

Before (leaking) vs after (engineered).

Anonymized, but structurally real. This is what the diagnostic found — and what we engineered in its place. Run the deltas against your own P&L.

Independent Restaurant Group · 3 units

Prime cost drift, recovered in 90 days

Before · Leaking
Prime cost68.0%
Food cost35.5%
Labor cost32.5%
Theo-vs-actual variance4.8 pts
Est. monthly leak$38K
After · Engineered
Prime cost61.2%
Food cost31.0%
Labor cost30.2%
Theo-vs-actual variance0.9 pts
Est. monthly leak$7K
−6.8 ptsprime cost · ~$31K/mo recovered across three units
What the diagnostic found

No unit-level visibility. Theoretical-vs-actual food variance was masking portioning and waste drift; mid-week labor was scheduled to a weekend pattern.

What we engineered

Par and prep discipline rebuilt around the data, plus demand-based mid-week scheduling. Variance closed from 4.8 to under 1 point.

Boutique Hotel · F&B operation

Labor model rebuilt for the wage floor — service intact

Before · Leaking
Labor cost38.2%
Sales per labor hour$41
Monthly OT hours120
Guest service score88
After · Engineered
Labor cost34.1%
Sales per labor hour$52
Monthly OT hours18
Guest service score89
−1labor cost · zero decline in guest service score
What the diagnostic found

Margin compression from the incoming wage floor, absorbed through overtime rather than scheduling. SPLH ran well below benchmark on soft shifts.

What we engineered

Demand-based scheduling tied to POS sales-per-labor-hour benchmarks, resetting the model for the new cost base without touching service standards.

High-Volume Bar

Beverage program repriced around pour cost

Before · Leaking
Pour cost28.6%
Spec variance6.4 pts
Dead SKUs on menu22
Annual margin at risk$184K
After · Engineered
Pour cost21.5%
Spec variance1.1 pts
Dead SKUs on menu4
Annual margin recovered+$184K
−7.1 ptspour cost · +$184K annual margin, no guest-facing price increase
What the diagnostic found

Pouring on instinct, not spec. Six points of variance between recipe and actual pour, and a long tail of SKUs that never moved.

What we engineered

Menu engineering, supplier renegotiation, and spec-driven pour controls repositioned the program around its highest-margin SKUs.

In aggregate

The numbers don’t round up.

6.0 pts
Average prime / pour cost recovered
90 days
Average time to engineered state
$0
Guest-facing price increases required

Your P&L has a number like this in it too.

The only question is which line item it’s hiding in. A free diagnostic call is where we start looking.

Start My Diagnostic →