Restaurant Menu Pricing

RevPASH, item classification, and menu psychology

Restaurants / HospitalityIntermediate

Restaurant Menu Pricing

Restaurant pricing operates under a distinctive set of constraints: perishable capacity (empty seats generate zero revenue), heterogeneous demand across dayparts, and menus that serve simultaneously as price lists, marketing tools, and psychological choice architectures. The field of restaurant revenue management, pioneered by Kimes (1998), adapts the principles of airline and hotel yield management to the unique economics of foodservice operations.

Restaurant Revenue Management

A restaurant’s revenue in any period depends on three factors: the number of available seats, the duration each party occupies a table, and the revenue generated per cover. Unlike hotels (which sell room-nights) or airlines (which sell seat-departures), restaurants face a dual duration problem: both meal duration and table turnover affect capacity utilization. A four-top occupied for three hours by a party of two ordering appetizers generates far less revenue than the same table turned twice with full parties ordering entrees and dessert.

Kimes (1998) identifies five strategic levers for restaurant revenue management: (1) pricing by meal duration, (2) managing meal duration directly, (3) forecasting demand by daypart, (4) overbooking tables, and (5) optimizing the menu mix. This topic focuses on the last of these—menu engineering and menu pricing—along with the RevPASH metric that ties individual menu decisions back to overall capacity utilization.

Definition — Restaurant Revenue Management

The application of disciplined analytics to predict consumer behavior at the micromarket level and optimize product availability and price to maximize revenue growth. In the restaurant context, the “product” is a combination of the physical seat, the time window, and the menu offering. The goal is to maximize RevPASH=Revenue/(Seats×Hours)\text{RevPASH} = \text{Revenue} / (\text{Seats} \times \text{Hours}) across all operating periods.

The menu engineering matrix, introduced by Kasavana and Smith (1982), is the foundational tool for analyzing menu item performance. It classifies every item along two dimensions: contribution margin (price minus food cost) and popularity (share of total units sold). The intersection of these two dimensions creates four quadrants, each with a distinct strategic implication.

Definition — Menu Engineering Matrix

A 2×2 classification of menu items based on two metrics. Let item ii have price pip_i, food cost cic_i, and units sold nin_i. The contribution margin is mi=pi−cim_i = p_i - c_i and the popularity share is si=ni/∑jnjs_i = n_i / \sum_j n_j. Let mˉ\bar{m} denote the average contribution margin and sˉ=1/N\bar{s} = 1/N the equal-share benchmark. An item is classified as:

  • Star if mi≥mˉm_i \ge \bar{m} and si≥0.7sˉs_i \ge 0.7 \bar{s}
  • Plowhorse if mi<mˉm_i < \bar{m} and si≥0.7sˉs_i \ge 0.7 \bar{s}
  • Puzzle if mi≥mˉm_i \ge \bar{m} and si<0.7sˉs_i < 0.7 \bar{s}
  • Dog if mi<mˉm_i < \bar{m} and si<0.7sˉs_i < 0.7 \bar{s}

The 70% threshold (rather than 100%) follows the original Kasavana-Smith convention, which accounts for the fact that a perfectly uniform menu mix is unrealistic.

Each quadrant suggests a different pricing action:

  • Stars (high margin, high popularity): Maintain current pricing and prominent menu placement. These items drive both volume and profit.
  • Plowhorses (low margin, high popularity): Customers love these items but they contribute less per unit. Consider modest price increases, portion adjustments, or ingredient substitutions to improve margins without sacrificing volume.
  • Puzzles (high margin, low popularity): Profitable when ordered, but rarely chosen. Reposition on the menu, rename, or pair with a popular item to increase visibility.
  • Dogs (low margin, low popularity): Neither profitable nor popular. Candidates for removal, repricing, or reformulation.

The Contribution Margin

The contribution margin for menu item ii is:

mi=pi−cim_i = p_i - c_i
(1)

where pip_i is the selling price and cic_i is the food cost (the direct cost of ingredients). Note that this is not the full profit margin—it excludes labor, rent, and other fixed costs. However, because those costs are largely independent of which items a customer orders, contribution margin is the relevant metric for menu-level decisions: every additional dollar of contribution margin from an order flows directly to the bottom line.

Interactive Matrix

The chart below plots eight representative menu items on the Kasavana-Smith matrix. Select any item from the dropdown and adjust its price to see how the change affects its contribution margin, classification, and position relative to the threshold lines. Notice how raising the price of a Plowhorse can promote it to a Star, while an aggressive price increase on a popular item may reduce its sales share and push it into Puzzle territory.

Repricing a Plowhorse

Suppose the House Burger has a food cost of $5 and is initially priced at $15, yielding a contribution margin of $10. If the average margin across all items is $14, the burger falls below the threshold and is classified as a Plowhorse. Raising the price to $20 increases the margin to $15—above average—reclassifying it as a Star, provided popularity remains above the 70% threshold. Use the slider above to verify this transition.

Revenue Per Available Seat Hour

While menu engineering focuses on the profitability of individual items, Revenue Per Available Seat Hour (RevPASH) captures the productivity of the restaurant’s physical capacity over time. It is the restaurant analog of RevPAR (Revenue Per Available Room) in the hotel industry.

Definition — RevPASH

Revenue Per Available Seat Hour is defined as:

RevPASH=Revenue in periodSeats×Hours in period\text{RevPASH} = \dfrac{\text{Revenue in period}}{\text{Seats} \times \text{Hours in period}}

RevPASH decomposes into three factors: average check size, seat occupancy rate, and table turn rate. A restaurant can increase RevPASH by raising prices (higher check), filling more seats (higher occupancy), or turning tables faster (shorter meal duration).

Formally, let RtR_t denote revenue in hour tt, SS the number of seats, and HH the number of hours in the operating period. Then:

RevPASH=∑t=1HRtS⋅H\text{RevPASH} = \frac{\sum_{t=1}^{H} R_t}{S \cdot H}
(2)

This can be further decomposed. Let ntn_t denote the number of covers (guests served) in hour tt and vˉt\bar{v}_t the average check per cover. Then:

RevPASH=1S⋅H∑t=1Hnt⋅vˉt\text{RevPASH} = \frac{1}{S \cdot H} \sum_{t=1}^{H} n_t \cdot \bar{v}_t
(3)
RevPASH Decomposition

RevPASH can be expressed as the product of three operational levers:

RevPASH=vˉ⏟avg check×nˉS⏟occupancy rate×1⏟per hour\text{RevPASH} = \underbrace{\bar{v}}_{\text{avg check}} \times \underbrace{\frac{\bar{n}}{S}}_{\text{occupancy rate}} \times \underbrace{1}_{\text{per hour}}

where vˉ\bar{v} is the average check and nˉ\bar{n} is the average covers per hour. Improving any one factor raises RevPASH proportionally, but the three factors interact: raising prices (increasing vˉ\bar{v}) may reduce covers if demand is elastic.

Daypart Variation

RevPASH varies dramatically across the week and within each day. Dinner service on Friday and Saturday typically generates three to five times the RevPASH of a Tuesday afternoon. This variation is the restaurant equivalent of airline peak/off-peak demand, and it creates opportunities for time-based pricing strategies: early-bird discounts to shift demand into low-RevPASH periods, premium pricing during peak windows, and happy-hour promotions to fill otherwise idle capacity.

The heatmap below displays RevPASH across a typical week. Adjust the seat count to observe how capacity affects the metric: more seats spread the same revenue over a larger denominator, reducing RevPASH unless the additional capacity attracts proportionally more covers.

Capacity Planning with RevPASH

An 80-seat restaurant generating $450 in revenue during the 7 PM hour on a Saturday achieves a RevPASH of 450/(80×1)=$5.63450 / (80 \times 1) = \$5.63 per seat-hour. If the restaurant expands to 120 seats but Saturday dinner revenue only increases to $550 (because the kitchen cannot produce more meals), RevPASH drops to 550/120=$4.58550 / 120 = \$4.58—a 19% decline despite higher absolute revenue. Use the seat slider above to see this tradeoff across all periods.

Beyond the analytical tools of menu engineering and RevPASH, the physical design of a menu influences what customers order. Research by Yang, Kimes, and Sessarego (2009) demonstrates that the way prices are presented on a menu has a measurable effect on spending behavior. Their study found that guests presented with menus using numeral-only prices (no dollar signs or decimals) spent significantly more than those given menus with traditional price formatting.

The key principles of menu price presentation are:

  • Currency symbol omission: Dollar signs activate the “pain of paying”—a concept from behavioral economics in which any cue reminding the customer that they are spending money reduces willingness to pay.
  • Round pricing: Prices like “32” rather than “$31.99” signal quality and reduce the impression that the restaurant is competing on price.
  • Nested placement: When prices are embedded at the end of a dish description rather than right-aligned in a column, customers process the price as an attribute of the dish rather than as a standalone cost to be compared across items.
  • Typographic de-emphasis: Setting the price in the same font size and weight as the description prevents it from becoming a visual anchor that dominates the selection decision.

The demonstration below shows the same set of dishes under two formatting regimes. Toggle between them to see how each principle changes the visual experience.

Effect Size of Price Format

In the study by Yang, Kimes, and Sessarego (2009), guests who received menus with numeral-only prices (no dollar signs) spent an average of $3.80 more per person than those receiving the traditional format with dollar signs and decimals. On a per-table basis with an average party of 2.5, this translates to $9.50 in additional revenue per table—a meaningful lift for a restaurant turning 40 tables per evening.

References

  • Kasavana, M. L. & Smith, D. I. (1982). Menu Engineering: A Practical Guide to Menu Analysis. Hospitality Publications.
  • Kimes, S. E. (1998). “The Bases of Restaurant Revenue Management.” Cornell Hotel and Restaurant Administration Quarterly, 39(5), 12–18.
  • Yang, S. S., Kimes, S. E. & Sessarego, M. M. (2009). “Menu price presentation influences on consumer purchase behavior in restaurants.” International Journal of Hospitality Management, 28(1), 157–160.
Previous
Bundling & Product Line Design
Next
From List Price to Pocket Margin