Chapter 24 — Key Takeaways
Revenue Management: Covers, Seat Turns, RevPASH, and Pricing the Room
The core claims
1. A seat-hour is perishable inventory, and it is the only inventory you never count. Bellwether manufactures 2,244 seat-hours a week and 116,688 a year, whether anyone sits in them or not. It sells about 45%. The rest are destroyed on the clock. The cost was sunk before the doors opened.
2. Covers measure traffic. RevPASH measures whether the building is working. Saturday brunch does 110 covers to Tuesday dinner's 62 — and \$212 less revenue. A cover count will lead you to praise the wrong service.
3. RevPASH has exactly three levers. Capacity utilization, average check, and dine time. Any revenue proposal is one of those three or it is not a proposal. Test every idea by asking which lever it moves and what it costs in the other two.
4. Duration management only pays where the seat is the binding constraint. At Bellwether it never is. Three nights are demand-constrained; two are hearth-constrained. A faster table on a Tuesday produces an emptier room; a faster table on a Saturday hands you a party the fire cannot cook for. Chapter 14 was right: wave-splitting creates zero capacity.
5. The most valuable cover in the week is the one on the slowest night. The fixed labor floor is already paid, so an incremental Tuesday cover contributes close to the full \$33.23 with no additional labor. This is the revenue-side answer to Chapter 19's finding that a percentage-of-sales labor model under-funds the empty nights.
6. Discount the trough; never surcharge the peak. The two are algebraically identical and operationally opposite. Set your menu price at what the peak should pay and build every instrument as a reduction from it.
7. An unfenced discount is not revenue management — it is a price cut you did not decide to make. Your social-media followers are your regulars, and your regulars come on the busy nights.
8. Size a no-show policy to the exposure, and charge for silence, not for cancellation. A cancellation returns sellable inventory. Penalizing it teaches guests to go quiet, which makes the problem worse.
9. Table mix is decided once and constrains you forever. Almost every American dining room is short two-tops. Three deuces a night at four-tops on a weekend is \$28,704 a year that will never appear on a P&L, because revenue you did not earn is not a line item anywhere.
10. Never close a plan with the lever you are counting on to save you. Bellwether's bridge takes the itemized version, not the turns version — not because the turns dollars are worse (they are better) but because they are the contingency, and a plan that spends its contingency in the base case has no base case.
The key formulas
$$\text{Seat turns} = \frac{\text{Covers}}{\text{Seats}} \qquad \text{Capacity utilization} = \frac{\text{Covers} \times \text{dine time (hrs)}}{\text{Seats} \times \text{hours open}}$$
$$\boxed{\ \text{RevPASH} = \frac{\text{Revenue}}{\text{Available seats} \times \text{Hours open}}\ }$$
$$\text{RevPASH} = \text{utilization} \times \frac{\text{average check}}{\text{average dine time (hours)}}$$
Bellwether dinner, on plan: $44.2\% \times (\$46.00 \div 1.583\text{ hr} = \$29.05) = \mathbf{\$12.85}$
The numbers to remember
| Seat-hours, one dinner service | 68 × 5 = 340 |
| Seat-hours, one week | 1,700 dinner + 544 brunch = 2,244 |
| Seat-hours, one year (indoor) | 116,688 |
| Dinner RevPASH, week | **\$12.85** (Tue \$8.39 → Sat \$16.64) |
| Brunch RevPASH | \$9.71 |
| On-premise RevPASH with patio | \$11.60 (the patio dilutes it, and is still worth having) |
| Capacity utilization, dinner week | 44.2% — Saturday peaks at 57.3% |
| Revenue per occupied seat-hour, dinner | \$29.05 — nearly identical every night |
| Contribution per cover, à la carte | \$33.23 at 27.8% blended COGS |
| One extra dinner cover a night, all year | \$11,960 |
| Two guests at a four-top, peak night | 3.5 destroyed seat-hours = \$92 |
| The revenue bridge | \$139,240, closed |
Rules of thumb
- Compute RevPASH by daypart weekly, alongside the flash report. Set a floor per service.
- Keep off-premise revenue out of the numerator. RevPASH is a dining-room metric; putting takeout in it makes the number rise while the room does nothing.
- Never use RevPASH to decide whether to add capacity. It is a comparison metric within a fixed capacity. Adding cheap seat-hours lowers it and raises profit. Capacity decisions are contribution decisions.
- Price a buyout against the RevPASH of the hours it eats, never against food cost. A four-hour Friday buyout at Bellwether must clear about \$4,417 to be neutral.
- Fence every discount three ways where you can — time, day, and product.
- The free no-show tools come first: an active confirmation at 24 hours and a same-day reminder. Do those before anything punitive.
- Give the manager on duty authority to waive a forfeiture without asking, and require only that they log the reason.
Key terms
RevPASH · seat turns · capacity utilization · table mix · yield management · demand-based pricing · fencing · no-show and deposit policy · shoulder hours · prix fixe economics · the shape of the week
What you should be able to do Monday morning
Pull last week's sales by service out of the POS, divide each one by that service's available seat-hours, and put the seven numbers on one page. Then find the two hours in your week where the building earns least, name the reason each is empty — demand, kitchen, or table mix — and pick the one instrument that addresses that specific reason. Fence it. Put a date on it. And do not close any revenue gap in your plan with the covers you are counting on to save you if something else goes wrong.