Case studies

Real data, actual opportunities.

See how Corundum's analysis framework uncovers win-win-win solutions for owners, guests & users, and the often overlooked ones who make it all happen — the employees.

Revenue StrategyCS.01

When Raising Prices Actually Worked

A growing hometown restaurant

The Problem: Owner implemented a 13.6% menu-wide price increase and needed to know whether it helped or hurt the business.
The Analysis: Weekly POS data split at the hike date. Two-sample t-test on guest counts. Revenue bridge decomposing price effect vs. volume effect.
The Finding: Demand was inelastic (elasticity −0.56). Guests fell 7.6% but price added $2,156/wk while volume only lost $1,403/wk — net positive. The guest decline was actually steeper before the hike.
The Result: Owner confirmed the hike was the right call. Optimization focus shifted to conversion, not reverting the price.

Weekly Revenue Impact

+13.6% menu price increase

−7.6% guest count

+7.3% net revenue gain

Staff PerformanceCS.02

The $100/Shift Opportunity

Full-Service Casual Dining & Bar

The Problem: Management sensed a gap between top and bottom servers but had no data to measure it or set a coaching target.
The Analysis: Shift-level tip aggregation across 6 months of POS exports. Zero-tip shifts excluded. Weekly high/low/avg bands per role over 23 weeks.
The Finding: Shift ceiling: $158.49 — earned by a real server on a real night. Team average: $62.79. Tip rate averaged 18.4% vs. a 22% benchmark ceiling.
The Result: A coaching target ($100/shift) grounded in the restaurant's own data — 63% of what already happened there.

Per-Shift Tip Range

$158.49 shift ceiling

$62.79 current avg

+59% to reach $100 target

OperationsCS.03

What −1 Server Per Shift Does to Earnings

Business Lunch Spot

The Problem: Owner wanted to model the financial impact of running leaner — not arbitrary cuts, but understanding what tighter scheduling means for each remaining server.
The Analysis: 147 multi-server shift days analyzed. Simulated removing lowest-earning server per day, redistributing their tips equally to retained servers.
The Finding: Avg tip rose from $70.37 to $76.80 per shift (+$6.44, +9.1%). On 4-server days: +$20.23/shift (+19.9%). Two servers were the lowest earner on 28 and 27 shifts respectively — a coaching flag.
The Result: A staffing optimization model built from real shift data. No change in guest volume required.

Per-Server Tips: Before vs. After

+$6.44/shift avg gain

+9.1% per-server tips

+19.9% on 4-server shifts

Staff PerformanceCS.04

Finding the Signal in the Noise

Iconic Burger Restaurant

The Problem: Management had gut instincts about server performance but lacked data for coaching conversations or objective evaluations.
The Analysis: Shift-level tips, net sales, guest count, and tip % aggregated per server across 23 weeks. Ranked by avg tip/shift and tip-to-sales ratio. Cross-referenced with low-earner frequency per shift.
The Finding: Bottom-quartile servers identifiable by name and shift frequency. Servers at 22%+ tip rate earned 35–45% more per shift than those at 16–17%. Guest count varied less across the team than tip rate — opportunity is conversion, not traffic.
The Result: A ranked server list with shift-level data behind every name. Evidence for fair, specific, productive coaching conversations.

Tip Rate Distribution

35–45% earnings gap

22%+ tip rate = top earners

+$37/shift avg potential