Cornell University School of Hotel Administration The Scholarly Commons

Articles and Chapters School of Hotel Administration Collection

6-1998

Restaurant Revenue Management: Applying Yield Management to the Restaurant Industry

Sheryl E. Kimes Cornell University, [email protected]

Richard B. Chase University of Southern California

Sunmee Choi Cornell University, [email protected]

Philip Y. Lee Cornell University

Elizabeth N. Ngonzi Cornell University

Follow this and additional works at: https://scholarship.sha.cornell.edu/articles

Part of the Food and Beverage Management Commons

Recommended Citation Kimes, S. E., Chase, R. B., Choi, S., Lee, P. Y., & Ngonzi, E. N. (1998). Restaurant revenue management: Applying yield management to the restaurant industry [Electronic version]. Cornell Hotel and Restaurant Administration Quarterly, 39(3), 32-39. Retrieved [insert date], from Cornell University, School of Hospitality Administration site: http://scholarship.sha.cornell.edu/articles/460/

This Article or Chapter is brought to you for free and open access by the School of Hotel Administration Collection at The Scholarly Commons. It has been accepted for inclusion in Articles and Chapters by an authorized administrator of The Scholarly Commons. For more information, please contact [email protected].

If you have a disability and are having trouble accessing information on this website or need materials in an alternate format, contact [email protected] for assistance. Restaurant Revenue Management: Applying Yield Management to the Restaurant Industry

Abstract The crucial element in a strategy for boosting restaurant revenues may be to relate prices to the length of time guests spend at the table. But, as the Witch of the West told Dorothy, the issue is how to do it.

Keywords restaurant industry, revenue management, yield management, strategies

Disciplines Food and Beverage Management

Comments Required Publisher Statement © Cornell University. Reprinted with permission. All rights reserved.

This article or chapter is available at The Scholarly Commons: https://scholarship.sha.cornell.edu/articles/460 Restaurant Revenue Management

Applying Yield Management to the Restaurant Industry

The crucial element in a strategy for boosting restaurant revenues may be to relate by Sheryl E. Kimes, prices to the length of time guests spend at the table. But, as the Witch of the West told Richard B. Chase, Sunmee Choi} Dorothy, the issue is how to do it. Philip Y Lee, and Elizabeth N. Ngonzi p ■ wk esearch in revenue manage- restaurant operators to gain the ment has heretofore addressed the benefits of strategic revenue man- theoretical and practical problems agement that they currently lack. facing airlines and hotels, among Our objective in this paper is to other industries, but has given little develop the framework for such a consideration to the restaurant in- theory. We discuss the necessary dustry.1 The restaurant business is conditions for revenue management, similar enough to hotel and airline the strategic levers available for rev- operations that restaurants should be enue management, how they have able to apply revenue-management been applied in traditional revenue- principles. Indeed, many restaurants use various revenue-management— Sheryl E. Kimes, Ph.D., is an type practices, but the application associate professor at the Cornell has so far been mostly tactical. We University School of Hotel Administra- believe that a broad theory of rev- tion. Richard B. Chase, Ph.D. , is a enue management would permit professor at the University of Southern California—Marshall School of Business. 1 For a review of revenue-management litera- Sunmee Choi is a Ph.D. candidate ture, see: Sheryl E. Kimes,“Yield Management: A Tool for Capacity-Constrained Service Firms,” at Cornell. Philip Y. Lee holds a Journal of Operations Management, Vol. 8, No. 4 Master of Management in Hospitality (1989), pp. 348-363; or Lawrence R. Weatherford degree from Cornell, and Elizabeth N. and Samuel E. Bodily,“A Taxonomy and R e- search Overview o f Perishable-Asset Revenue Ngonzi, M.M.H., is a master’s degree Management: Yield Management, Overbooking candidate at Cornell. and Pricing,” Operations Research, Vol. 40, No. 5 (1992), pp. 831-844. © 1998, Cornell University

32 HOTEL AND RESTAURANT ADMINISTRATION QUARTERLY FOCUS ON FOOD SERV management settings, and how they, taurants have some flexibility to requires an effective computerized along with some tactical tools, can crowd a table with an additional seat or manual reservation system. be applied to restaurants. if necessary, and the restaurant’s cost Perishable inventory. One of adding additional capacity in the might think of a restaurant’s inven- Defining Revenue Management form of tables or seats (say, by tory as its supply of raw food, but Revenue management is the appli- reconfiguring the dining room) is most of that is not perishable until it cation of information systems and lower than that of many businesses is removed from the freezer or sit- pricing strategies to allocate the that typically use revenue manage- ting on the receiving dock. Instead, right capacity to the right customer ment. Most restaurants have a fixed restaurant inventory should be at the right place at the right time.2 number of tables, but can vary the thought of as time—or, in this case, In practice, revenue management number of seats depending on the the time during which a seat or has meant determining prices ac- mix of party sizes. In addition, some table is available. If the seat or table cording to predicted demand levels restaurants might increase capacity is not occupied for a period of time, so that price-sensitive customers during warm weather by using out- that part of the restaurateur’s inven- who are willing to purchase at off- door dining. tory perishes. This is the key to our peak times can do so at favorable Adjusting kitchen capacity is strategic framework, and it is the prices, while price-insensitive cus- generally more expensive than add- element we believe has been missing tomers who want to purchase at ing tables, although kitchen output in most approaches to restaurant peak times will be able to do so. can often be increased by changing revenue management. Instead of The application of revenue manage- the menu (e.g., by selling only items counting table turns or revenue for ment has been most effective when that can be prepared quickly) or by a given day part, restaurant operators it is applied to operations that have increasing staffing levels so that should measure revenue per avail- the following characteristics: rela- more food can be prepared. Still, able seat hour (RevPASH). This tively fixed capacity, predictable one can consider the kitchen capac- measure captures the time factor demand, perishable inventory, ap- ity to be relatively fixed because of involved in restaurant seating. propriate cost and pricing structure, the expense of adding equipment. Many restaurants evaluate manag- and demand that is variable and Although service capacity can be ers and servers based on average uncertain.3 Those attributes are increased by adding staff or by re- sales per customer. This is equivalent generally found in some form or ducing meal duration, the limita- to hotels’ measuring effectiveness by another in the restaurant industry. tions of the kitchen and dining ADR without paying attention to Relatively fixed capacity. Res- room may make that change fruit- occupancy. While knowing sales per taurants’ capacity can be measured less. Thus, although restaurant op- customer or contribution margins by seating, kitchen size, menu items, erators can tweak their capacity, it is of menu items is valuable, those and staffing levels. Most restaurant essentially fixed. measures do not provide the infor- operators’ approaches to optimizing Predictable demand. Setting mation on revenue generation that revenue primarily involve filling the aside carry-out activities as a sepa- R evPASH would give. seats to capacity and turning tables rate business, restaurant demand Appropriate cost and pricing as quickly as possible, but that effort consists of guests who make reserva- structure. Like hotels, restaurants can be limited by the kitchen, by tions and guests who walk in. Both have a cost structure that features the menu design, or by staff mem- forms of demand can be managed, relatively high fixed costs and fairly bers’ capabilities. but different strategies are required. low variable costs, although it’s true Seating capacity is generally fixed In sum, guests who make reserva- that an item’s food-cost percentage over the short-term, although res- tions and those who walk in consti- is usually higher than the cost of tute an inventory from which man- opening a hotel room compared to agers can select the most profitable the revenues from that room. Like 2 Barry C. Smith, John E Leimkuhler, and Ross M. Darrow, “Yield Management at Ameri- mix of customers. To forecast this hotels, restaurants must generate can Airlines,” Interfaces•, Vol. 22, No. 1 (1992), demand and manage the revenue it sufficient revenue to cover variable pp. 8—31. generates, a restaurant operator costs and offset at least some fixed 3 Kimes, pp. 348-363; and Robert G. Cross, Revenue Management (New York: Broadway needs to compile information on costs. Nevertheless, restaurants’ rela- Books, 1997). See also: Robert G. Cross, the percentage of reservations and tively low variable costs allow for “Launching the Revenue Rocket: How Rev- walk-ins, guests’ desired dining some pricing flexibility and give enue Management Can Work for Your Business,” Cornell Hotel and Restaurant Administration Quar- times, and likely meal duration. operators the option of reducing terly, Vol. 38, No. 2 (April 1997), pp. 32-43. Tracking customer-arrival patterns prices during low-demand times.

June 1998 • 33 Time-variable demand. Cus- unpredictable duration of customer Exhibit 1 tomer demand varies by the time of use, which inhibits their ability to Methods of managing year, by the week, by the day, and by manage revenue. To allow for better duration the day part. For restaurants, dinner revenue-management opportunities, demand may be higher on week- restaurant managers must increase ends, during summer months, or at control over the length of time Uncertainty of arrival particular times during the lunch or customers are occupying their seats. Internal measures dinner periods. Restaurant operators To do this, they can refine the defi- Forecasting must be able to forecast time-related nition of duration, reduce the un- Overbooking demand so that they can make ef- certainty of arrival, reduce the un- fective pricing and table-allocation certainty of duration, or reduce the External measures decisions to manage the shoulder amount of time between customers’ Guaranteed reservations periods around high-demand meals (see Exhibit 1). Reconfirm reservations Service guarantee periods. Redefining duration. The A special factor for restaurant length of time that guests use a Uncertainty of duration operators is that they have to reckon table is usually measured either by Internal measures with the length of time a party stays the number of minutes or hours once it is seated. This is analogous that they actually occupy that table Menu design Process analysis to hotels’ having to forecast the or by the events relating to a meal Labor scheduling number of guests who will stay an (e.g., by the course or by the full Communication systems additional (unscheduled) night, but meal). In either case, the restaura- the hotel still is selling an integral teur must know how long a typical External measures room-night, rather than the some- guest will stay at a table for a given Pre-bussing Check delivery times unpredictable period that day part or meal. When duration is Coffee and dessert bar diners will stay at a table. If restau- defined as a meal rather than as the Visual signals rant managers can accurately predict time to complete a meal, the opera- Reduce time between customers meal duration, they can make better tor must be able to forecast meal Process analysis Communication systems reservation decisions and give better length so that selling a meal essen- estimates of waiting times for walk- tially becomes selling a certain in guests. length of time in the restaurant. The conventional wisdom is that Managing Demand: Strategic Levers restaurants sell meals, rather than Restaurants appear to possess the explicitly selling time—although a conditions necessary for revenue few restaurants actually do sell management, but we see little evi- blocks of time (e.g., seating parties dence of restaurants’ using a strate- every two hours, with a reminder gic approach for using the demand- to leave when the time is up). management mechanisms at hand. Rather than explicitly selling time, A successful revenue-management however, most restaurant operators strategy is predicated on effective will continue to sell time in the control of customer demand. We form of meals of predictable length. have alluded to the two strategic This could be done directly, in levers that restaurant managers have theory, by asking customers how at hand to manage demand and, long they will need the table when thus, revenue. Those are duration they make a reservation or request a management and demand-based table, but such an approach would pricing.4 require a radical change in thinking Duration management. Res- for both management and custom- taurant operators typically face an ers. Even though that approach would help change the definition 4 As explained further in: Sheryl E. Kimes and of duration from the meal itself to Richard B. Chase, “The Strategic Levers of Yield Management,” Journal of Service Research, 1998, the time involved in eating the in press. meal, the tactic might put off most

34 HOTEL AND RESTAURANT ADMINISTRATION QUARTERLY FOCUS ON S E R V guests—other than those who have such things as asking for deposits or untary rescheduling system when a specific date or appointment after guaranteeing reservations with they believe they will be oversold by the meal. credit cards. asking willing patrons to switch Instead, most restaurant operators Overbooking. Many capacity- their meal to a slower time. A call will have to keep track of the length constrained service industries use far in advance to customers who of time that guests occupy a table overbooking to protect themselves have reservations at the congested during given day parts. From those against no-shows. Restaurants have time with the offer of a discounted observations the restaurateur could typically not used overbooking or free meal for switching to an determine an average meal length, in this way but have instead relied open period could increase cus- while also noting any variation in on walk-in business as a buffer— tomer goodwill and increase profit. meal length. That is, the restaurant although this strategy works only if While proposing to displace cus- operator needs to know the average enough walk-ins arrive at the right tomers can make them angry, mak- length of a meal, plus how close to time. ing them wait without warning the average most diners come. Wide The key to a successful over- could well be worse. In the latter variation of meal lengths (i.e., some booking policy is to obtain accurate situation, many restaurants attempt very short, but others very long) information on no-shows, cancella- to compensate for the inconve- makes forecasting more difficult tions, and walk-in guests to set levels nience of an unexpected wait by and perhaps calls for management of overbooking that maintain an offering guests a discounted meal efforts to make the duration more acceptable level of customer service. or free items. consistent. A manager can use simple math- Shifting to the customer the con- Uncertainty of arrival. Restau- ematical models to develop appro- sequences of not arriving is a prac- rant managers have always struggled priate overbooking policies by time tice that has gained currency in with not knowing whether or how o f day, day o f week, and time o f many service industries, although many guests will show up. Some year. A good overbooking policy restaurants have been slow to adopt rely solely on walk-in business and balances the cost of unused tables it. Hotels and airlines have used depend on queues to manage guest with the cost of inconveniencing or guaranteed reservations for many arrival (ensuring a steady flow of displacing a party—bearing in mind years and have thereby been able to guests during busy times). Restaura- that a guest denied a reserved table reduce the number of no-shows. teurs have long taken reservations to may not be especially forgiving. A The American Express No-Show gain a forecast of arrivals, but that restaurant that attempts an over- Initiative is attempting to accom- does not eliminate uncertainty of booking approach must develop plish the same thing for restaurants. arrival, since not all guests honor good internal methods for selecting Under this program, patrons are their reservations and others arrive and handling displaced guests. Many asked to give their credit-card early or late. In any of those cases industries base their displacement number with a reservation and are the restaurant faces the prospect of decision on time of arrival (if cus- warned that they will be charged if unused seat-hours. It is worth not- tomers are late, their reservation is they do not come. If guests fail to ing that in many countries a reser- no longer honored), frequency of honor their reservation (and give no vation for a specific time is regarded use (regular customers are never notice), they are charged a fee (typi- as a firm commitment, and guests displaced), or perceived importance cally, $15 to $25 per guest).5 This from those cultures would be ap- (important, high-spending custom- program, in conjunction with edu- palled at U.S. restaurants’ sometimes ers are never displaced). cational advertising on the impact cavalier treatment of reservations. Following the approach of air- of restaurant no-shows, has helped Restaurants taking reservations lines, restaurant operators might ask increase customer awareness of the can involve their customers in ar- customers who arrive during an need to cancel unwanted reserva- rival management (external ap- overbooked time whether they tions. Program managers report that proaches) or not involve customers would voluntarily relinquish a table only a small percentage of the guar- (internal approaches) to decrease and move to a different time (with anteed reservations have turned into arrival uncertainty. Perhaps the fore- an appropriate incentive). Airline no-shows that incurred a penalty— most internal approach is over- customers seem to have accepted an indication that the program has booking, which most restaurants this practice, and some travelers succeeded. avoid. The primary external ap- even seek out the opportunity for a 5 Brian Breuhaus, “Handling No-shows: proach is to shift the liability for free trip when flights are oversold. Operators React to Reservation Plan,” Restaurant arrival to the customer by doing Some upscale restaurants use a vol- Business, Vol. 16, No. 1 (1998), p. 13.

June 1998 • 35 Restaurants have long required vation requests to accept, and res- a deposit for special meals (e.g., taurants with a large walk-in trade Mother’s Day, New Year’s Eve), al- will be better able to provide accu- though the practice may meet with rate estimates of waiting time for customer resistance during times of guests in the queue. As with arrival low demand. Again, a deposit helps time, restaurant operators can exert A dessert and coffee bar, to ensure that customers will honor control over meal duration. Internal their reservation and also protects approaches in this case revolve where guests can move to restaurants against last-minute around making the meal length cancellations. more consistent, while the external Rather than require deposits in approaches involve encouraging chat after a meal, provides a any form, some restaurants use a less guests to give up their table even if obtrusive, more service-oriented they choose to linger elsewhere in customer-friendly way to free method of reducing no-shows. the restaurant. These operators call their customers By reducing time variability, dinner tables for the next during the day to reconfirm their managers will be better able to give reservations. The call reminds the accurate estimates of waiting time dining parties. customer of the reservation and and determine whether and for gives the customer a chance to can- what time reservations should be cel on the spot, if need be. The calls accepted. A restaurateur can work also create a reasonably solid forecast toward consistency of duration of the numb er of parties who in- through menu design, process de- tend to honor their reservations. sign, labor scheduling, and commu- For this approach to be successful, nication tools. Some restaurants the incremental personnel cost asso- have redesigned or established their ciated with calling customers should menus according to the preparation be offset by the increased revenue and consumption time for each associated with a reduction in menu item. Menu items that exceed no-shows. the established target for prepara- Quid pro quo. To encourage cus- tion or consumption are either tomers to arrive on time, restaurants reconfigured or eliminated from the could offer a service guarantee. This menu. Likewise, menu items that is another practice that is not wide- cause customers to Unger can be spread in the restaurant industry, but eUminated if they do not contr ibute has gained strength in other busi- to an increase in revenue per avail- nesses. Some golf courses, for in- able seat-hour (particularly if the stance, offer free or discounted play restaurant cannot encourage guests to patrons who must wait more than to move to another area of the res- ten minutes after their scheduled tee taurant for after-dinner drinks or time. Restaurateurs could adapt the munchies). Some chain restaurants idea by offering discounted or free have used the same approach to items to diners who are not seated reconfigure or eliminate food- within a specified length of time of preparation and service processes to their reservation. reduce the variabiUty in service time. Uncertainty of Duration The redesign of the menu and A restaurant operator who has dealt processes, in conjunction with im- with the arrival-time issue must still proved forecasts of customer arrivals, be able to forecast meal length ac- should improve labor scheduling, curately, because this variable is the which is a key element in control- one that controls the number of Ung meal duration. Restaurateurs’ tables available. With this informa- common desire to minimize labor tion, operators of reservation-based costs may backfire if reduced staffing restaurants can decide which reser- leads to slower table turnovers and

36 CORNELL HOTEL AND RESTAURANT ADMINISTRATION QUARTERLY FOCUS ON FOOD SE C E longer meal times. The increased A restaurant can use both implicit management system to find out revenue resulting from faster table and explicit signaling devices to when a party is leaving, so that they changeovers made possible by extra remind guests and servers that the can clear and reset the table, thereby bussers or servers may more than meal is over. Many restaurants use increasing the revenue per available compensate for the increased per- subtle implicit approaches such as seat-hour.7 A similar result may be sonnel costs. A revenue-manage- bussing the table, dropping off the obtained by developing a good ment strategy would help a restau- check, or offering valet service. In a communication system among serv- rant operator determine appropriate few restaurants, customers are asked ers and bussers so that employees staffing levels. to specify how long they plan to know when tables are ready to be Some restaurants have improved stay, but that is rare. Instead, the cleared and reset. It goes without communication systems among restaurant manager must rely on the saying that management could ana- employees and have increased con- timing of the courses and other lyze and streamline the process of trol over duration by tracking the implicit signals to remind the cus- clearing and resetting tables to connection between food prepara- tomer that the meal has ended. minimize changeover time. tion and food delivery. By setting up Explicit approaches risk customer appropriate communication mecha- ire. The manager obviously cannot Price Management nisms, kitchens can notify servers ask customers directly to leave, but People commonly connect revenue that a course is ready for pick up the restaurant could attempt other, management with offering dis- and servers can notify bussers that less offensive methods of turning counts, but discounting is only part a table is ready to clear, thereby the table. Some restaurants in the of the revenue-management story. speeding the meal service (usually theater district of New York City, When price is used as a tool of rev- to the guests’ delight) and making for instance, place an hourglass on enue management, managers must it possible to improve revenue per each party’s table. When the sand in think beyond happy hours and two- seat-hour. To assist with employee the hourglass is gone, patrons have a for-one specials and develop meth- communication, restaurants have visual cue to finish dinner and leave ods for offering differential prices available to them information tech- so that they will not be late to the that make sense for the demand nology, such as table-management theater. level at a given time. Hotels and systems. Some restaurants provide a des- airlines use various rules, sometimes Selling time. People generally sert and coffee bar where guests can known as price fences, to offer dis- do not think of restaurants as selling move to chat. This provides a cus- counts on inventory that might a time-related product, except per- tomer-friendly way to hasten guest otherwise not be sold at all to cus- haps during lunch. Hotel guests departure and free the dinner table tomers who might otherwise not who stay over expect to be charged for the next party. purchase—while at the same time for the additional night, but restau- Reducing changeover time. preventing customers who were rant customers do not expect to pay Reducing the amount of time be- going to buy anyway from taking extra if they stay at a table after the tween customers (changeover time), advantage of a discount that they meal is over. Indeed, in many parts increases capacity and revenue. This did not actively seek.8 Thus, coach of the world once the guest is seated tactic will not offend a departing passengers flying from Chicago to that table is hers for as long as she customer and should please the Los Angeles may pay nothing (for chooses. That guest, however, usu- customers who are waiting to be those using frequent flyer vouchers) ally pays an explicit cover charge, seated. Reducing changeover time or over $1,500 for the same seat. something that is uncommon in the has become a common strategy for The fares vary according to the time United States. This discussion ap- airlines. Southwest Airlines and the of the reservation, the days that the plies to the American-style ap- Shuttle by United both boast 20- individual is flying, and group or proach of selling a meal with a table, minute aircraft turnarounds and company affiliations that the flyer rather than the European attitude of have thereby been able to increase might have. The fences can com- selling a table with a meal. plane use.6 Some restaurants have Part of duration management instituted table-management systems 7 Alan Liddle,“New Computerized Table involves finding ways to signal to that track tables in use, the progress Management Reduces Guests’ Waits, Empty Seats,” Nation’s Restaurant Neu/s, August 5,1996, guests that it is time for them to of the meal, and when the bill is p. 22. relinquish their table. Customers paid. Bussers check the table- 8 See: Richard D. Hanks, Robert G. Cross, and who unexpectedly linger after their R. Paul Noland, “Discounting in the Hotel 6 Sheryl E. Kimes and Franklin Young, “The Industry: A New Approach,” Cornell Hotel and meal is completed may prevent the Shuttle by United,” Interfaces, Vol. 27, No. 3 Restaurant Administration Quarterly, Vol. 33, No. 3 manager from seating the next party. (1997), pp. 1-13. (June 1992), pp. 40-45.

June 1998 • 3 7 other unfavorable times rather than but more involved items from a full Exhibit 2 levy surcharges in the face of strong menu. The restaurant could also Types of price fences demand. However, they noted that charge extra for customers who most customers expect restaurants to desire a visit from the chef, under Physical price fences charge relatively modest prices that the rubric of a “meet the chef” Table location allow the restaurant to fill to over- night, or for those who prefer a par- Party size flowing, with the resulting excess ticular server. Menu type Absence or presence of demand managed by queues. That The purpose of intangible rate certain amenities strategy was once used by airlines, fences is to shift demand from busy times to slow periods, to reward Intangible price fences but more than a decade of airline revenue management has changed regular and reliable customers, and Group membership or affiliation Time of day or week that approach, and the stand-by pas- to schedule the highest-margin busi- Duration of use senger is nearly extinct. The Kahne- ness at the busiest times. When reservation booked man group concluded that the tim- The supply of restaurant seats Walk-in versus reservation ing of the sale (i.e., busy night versus exceeds the demand in most mar- Type of reservation (guaranteed or not) slow night) is not viewed as a “fair” kets, and managers frequently offer prise almost any set of rules as long reason to change prices. In deter- discounted prices in an attempt to as they somehow make sense to the mining who pays which price, man- fill empty tables. The point of rev- customer. In contrast, most restau- agers must develop logical selection enue management is that the dis- rants offer the same menu prices methods that make sense to custom- counts should fit the restaurant’s regardless of the customer’s demand ers without offending their sense of overall strategy. As long as the vari- characteristics. Perhaps the question fairness, because, as the Kahneman able costs of the meal are covered, for restaurateurs is whether they group discovered, customers will go managers should consider offering could implement some kind of out of their way to punish a business discounts and other benefits for din- pricing differential for busy times that has acted unfairly. ing during off-peak times. (e.g., Saturday night) and slack Fences. Restaurant operators Broad-scale discounting is dan- times. Early bird specials are a step might take into account the follow- gerous because price is one of the in this direction, as are special prices ing attributes in developing price methods that customers use to de- for affinity groups and frequent- fences (see Exhibit 2). Physical at- termine the perceived value of a diner clubs. The next step is to cre- tributes include table location, party service—making fences important. ate an overall demand-management size, menu type, and amenities, while Most quick-service restaurants face a program based in part on time- intangible rate fences include group problem of perceived value. Having sensitive pricing. membership or affiliation, time of once “told” the public that its Charging price premiums is gen- day or week, meal duration, presence burgers (and seven other items) are erally not acceptable, as revealed by or timing of the reservation (e.g., worth just 88 cents, McDonald’s is the work ofKahneman, Knetsch, whether the party is a walk-in), thereafter hard-pressed to raise and Thaler.9 Although it is possible and whether the reservation is prices—at least on those items. One to charge a higher price solely based guaranteed. tactic to offset this difficulty is to on high demand, customers may Restaurateurs may wince at the offer early bird specials or two-for- resent being charged different notion of charging higher prices for one offers on a separate menu. This prices for essentially the same menu customers seated at more desirable approach builds demand for slow item, unless they perceive a “fair” tables (perhaps those with a view), periods, but does not diminish the reason for the price differential. The but an informal differential pricing customer’s perception of the regular Kahneman group suggests that con- system already exists in some restau- menu. Beyond early bird specials and sumers consider it more fair to set a rants in the form of gratuities given the like, some restaurants offer spe- “full” price during high times and to the maitre dy or host to ensure a cialty meals, wine-tastings, or five then offer discounts during slack or preferred table. Some restaurants music during slow periods to attract effectively charge by party size by business. Restaurants with meeting 9 D. Kahneman,J. Knetsch, and R. Thaler, including a mandatory service facilities solicit group and corporate “Fairness as a Constraint on Profit Seeking: Entitlements in the Market,” American Economic charge for large parties. Parties or- business for slow times. Other res- Review, Vol. 76, No. 4 (1986), pp. 728-741. For a dering from a reduced menu of taurants have developed frequent- summary of this work, see: Glenn Withiam, items that are quickly prepared and eater clubs that offer bonus points “What’s Fair (and What Ain’t Fair),” Cornell Hotel and Restaurant Administration Quarterly, Vol. 27, consumed can be offered a lower for customers who dine during off- No. 3 (November 1986), p. 7. price than parties ordering similar peak periods.

38 CORNELL HOTEL AND RESTAURANT ADMINISTRATION QUARTERLY FOCUS ON FOOD SE E

menting strategic differential pricing. A Typology of Revenue-Management The accompanying box summarizes the strategic revenue-management Different industries are Price subject to different positions of several industries by combinations of duration Fixed Variable comparing their ability to control control and variable duration and to vary price. pricing (see accompany Quadrant 1 .Quadrant 2 ing table). Industries o> Restaurants can adapt the prin- .o traditionally associated CO Movies | Hotels ciples of revenue management to with revenue manage o Stadiums and arenas i Airlines increase revenue per available seat- Convention aenters Rental cars ment (hotels, airlines, o hour (RevPASH) by emulating cer- car-rental firms, and i Cruise lines cruise lines) are able to i tain attributes of the industries that apply variable pricing for I i use revenue management success- a product that has a I fully (i.e., those in Quadrant 2 of the specified or predictable i box). The key elements are being duration (Quadrant 2). I a> Quadrant 3 \ Quadrant 4 Movie theaters, arenas, n able to predict the duration of a S performing-arts centers, o Restaurants I Continuing care customer’s visit and to establish vari- and convention centers mS Golf courses | Hospitals able prices based on a customer’s £ charge a fixed price for Q. Internet service . demand characteristics. Restaurant a product of predictable C providers ' duration (Quadrant 1), 3 operators can make duration more while restaurants, golf predictable by reducing the uncer- courses, and most ______L tainty of when (or whether) cus- internet service providers charge a fixed price but face a relatively unpredictable duration tomers will arrive and by reducing of customer use (Quadrant 3). Many health-care businesses charge variable prices (e.g., Medicare versus private pay), but do not know the duration of patient use, even though the variability of the length of the some may try to control that duration (Quadrant 4). The lines dividing the quadrants are meal. Operators can apply differen- broken because in reality no fixed demarcation point exists between quadrants. Thus, an tial pricing and logical rate fences to industry (such as restaurants) may have attributes from more than one quadrant. Successful revenue-management applications are generally found in Quadrant-2 build demand during off-peak peri- industries, because they can manage both capacity and price. Restaurants can shift to ods and to establish appropriate Quadrant-2 strategies by manipulating duration and price. prices for busy periods. Although many restaurants use some of the tools described in the accompanying Many restaurants practice some of article, a strategic framework with which to evaluate and position such efforts has not yet been developed.— S.E.K., R.B.C., S.C., P.Y.L., and E.N.N. the revenue management approaches described in this paper, but do not yet have a strategic framework by Customers and groups who pro- embodied by menu engineering. which to coordinate these practices. vide a substantial amount of business Menu engineering does consider The intent of this paper was to ex- should be rewarded with benefits the demand for a menu item, so it plain the elements of such a strategy. and discounts. Some restaurants offer has an element of demand pricing, The strategies themselves will be as discounts to customers who are but an item’s contribution margin divergent as the many types of res- associated with particular groups (in some form) is also a strong fac- taurant. Thus, researchers are chal- (e.g., AARP) or who are employees tor.10 Perhaps adding a time-based lenged to use this framework to of certain corporations. Regular factor to the analysis would make assist restaurant managers in identi- customers are sometimes guaranteed menu engineering an effective tactic fying revenue-management oppor- desirable tables and times, and res- for revenue management. tunities and to develop appropriate taurants can offer discounts to cus- duration-management and differen- tomers who make guaranteed reser- Unified Approach tial-pricing approaches. In the long vations by a specified advance time. Few of the ideas connected with run, achieving the full potential from A final consideration about pric- revenue management that we have revenue management lies in man- ing is that restaurateurs have tradi- presented here are novel or untried. agement’s ability to market and tionally used a cost-plus approach to In fact, our examples are drawn manage every available moment of pricing, typically setting prices so from actual restaurants’ activities. the restaurant as a unique product. that the menus average food cost is What is needed is a unified frame- This in turn requires that restaura- capped at some predetermined per- work for developing and imple- teurs treat the time and length of the centage (say, 30 percent). Individual meal as a variable that should be as 10 For example, see: Michael Kasavana and Don items deviate from this average, of Smith, Menu Engineering (Lansing, Michigan: carefully managed as the service course, giving rise to the analysis Hospitality Publishers, 1982). process itself. CQ

June 1998 • 39