Hospitality Review

Volume 25
Article 6
Issue 2 Hospitality Review Volume 25/Issue 2

1-1-2007

Menu Analysis: A Review of Techniques and

Approaches
James J. Taylor
Central Michigan University, null@[Link]

Denise M. Brown
Michigan State University, shbsirc@[Link]

Follow this and additional works at: [Link]

Recommended Citation
Taylor, James J. and Brown, Denise M. (2007) "Menu Analysis: A Review of Techniques and Approaches," Hospitality Review: Vol. 25:
Iss. 2, Article 6.
Available at: [Link]

This work is brought to you for free and open access by FIU Digital Commons. It has been accepted for inclusion in Hospitality Review by an
authorized administrator of FIU Digital Commons. For more information, please contact dcc@[Link].
Menu Analysis: A Review of Techniques and Approaches
Abstract
This review discusses menu analysis models in depth to identify the models strengths and weaknesses in
attempt to discover opportunities to enhance existing models and evolve menu analysis toward a
comprehensive analytical model.

Keywords
Food Science, Menu Planning, Modeling

This article is available in Hospitality Review: [Link]

 Menu Analysis: A Review of Techniques
and Approaches
By James J. Taylor and Denise M. Brown
This review discusses menu analysis models in depth to identify the models strengths and weaknesses in attempt to discover
opportunities to enhance existing models and evolve menu analysis toward a comprehensive analytical model.
Introduction
The menu drives purchasing and production decisions in all restaurant operations. Menu analysis
models enable restaurateurs to systematically evaluate individual menu items by comparing each menu
item to other menu items based on pre-selected criteria. Historically menu analysis models have included
food cost, contribution margin and popularity or product mix. As many definitions of menu engineering
and analysis exist as there are techniques. For example, LeBruto, Quain, and Ashley (1995) defined menu
engineering as a methodology to classify a menu item by contribution margin and popularity. Cohen,
Mesika, and Schwartz (1998) suggested menu sales mix analysis (often referred to as menu engineering) as
a suitable managerial tool. Atkinson and Jones (1994) considered menu sales mix analysis a generic term
that included approaches attempting to improve menu performance. These early definitions were
expanded into a number of models to assess and manipulate (engineer) menus.
Before the advent of elaborate point of sale (POS) systems, cashiers tallied the sales of specific
menu items for management to evaluate. (Kotschevar, 1994) Management developed a popularity index by
determining the percentage of each menu item sold in a given period in comparison to the total of all
menu items sold during that same period. Only specific predetermined menu items were chosen for
evaluation.
Popularity indexes reflected the popularity of menu items but did not consider any cost
information. Since popularity indexes represented simple percentages, comparison of percentages from
one period to another was difficult. Often only a selection of the menu items was tallied during a specific
production or service period. Total sales of menu items may have varied across these periods resulting in
unequal comparisons based on differing menu item choices.
In an early attempt to improve on simple popularity indexes, Hurst’s menu score was another
menu analysis method frequently used prior to the implementation of POS systems. The Hurst menu
score was obtained from the popularity percentage of a menu item based on similar menu items multiplied
by the gross profit of the selected menu item. The popularity percentage was calculated by taking the total
count of patrons choosing a menu item and dividing it into a total count of similar menu items being
studied within a category of menu items, for example entrees. The resulting total became the menu score.
(Kotschevar, 1994.)
Efforts to model the role and significance of menu composition using a matrix approach were
devised to improve on these early methods of menu analysis. Miller (1987) was the first to develop a
matrix model which focused on food cost and product mix to analyze menu item profitability without
consideration of production costs. Kasavana and Smith (1982) used the Boston Consulting Group
Portfolio Analysis as the basis for the Menu Engineering matrix approach to menu analysis. They
incorporated contribution margin defined as the difference between the sales price of an item and the cost
of food product to produce that item. This approach focused on flow-through dollars rather than on food
cost percentages. Kasavana and Smith (1982) considered high gross profit and low food cost as mutually
exclusive. Pavesic (1983) modified matrix models by both Miller (1987) and Kasavana and Smith (1982) by
using food cost and weighted average contribution margin, which included popularity and contribution
margin. Pavesic did not treat food cost percentage and gross profit as mutually exclusive components and
used weighted gross profit/contribution margin to replace the individual menu item gross profit used by
Kasavana and Smith (1982.) In doing so, Pavesic included popularity as an indirect third variable. To
obtain the weighted gross profit, individual gross profit of each menu item was multiplied by the number
sold. Pavesic’s Cost/Margin model included an option of adding a “supplemental cost” to food cost. This
theoretically could be used for a labor component or for packaging costs of carry-out food.
Hayes and Huffman (1985), LeBruto, Quain, and Ashley (1995), Cohen, Mesika and Schwartz
(1998), and Horton (2001) have all attempted to include labor costs in menu analysis approaches. Hayes

FIU Review Vol. 25 No. 2 Page: 74

and Huffman developed an individual profit and loss statement for all menu components in an attempt to
allocate all costs including labor and fixed costs to individual menu items. LeBruto, Quain, and Ashley
were the first to consider categorical data to measure labor costs by segmenting the Menu Engineering
matrix from four into eight quadrants to reflect high and low labor items. Cohen, Mesika, and Schwartz
(1998) introduced a multidimensional approach that considered food cost, price, labor cost, popularity,
and contribution margin to evaluate menu effectiveness. Attributing labor to specific menu items is inexact
at best. Since multiple tasks are taking place in a restaurant operation at the same time where several
different menu items are prepared simultaneously, allocating exact labor expenses becomes complex.
Bayou and Bennett (1992) proposed that fixed costs such as rent, utilities, and insurance should
be excluded from the evaluation of menu items. Bayou and Bennett developed a profitability analysis
model to evaluate the financial strength of menu items in an attempt to allocate variable costs such as
labor. Horton (2001) segmented the menu prior to the analysis and evaluation of the menu into categories
of comparable items for comparison. He modified the Kasavana and Smith model to include pure variable
labor costs. Horton defined pure variable labor costs as those labor costs that could be calculated in the
production of a menu item. However, Horton did not consider other production costs in the preparation
process into his labor calculation such as the time it takes to cook a batch sauce in a pasta dish.
The Miller Model
The first model, the Menu Analysis model (MAM), was developed by Miller who attempted to
identify menu items that were both popular and low in food cost. He introduced a four quadrant matrix
for analyzing menus. Quadrants were segmented based on the product mix and food cost of the menu.
The average food cost axis is defined as the line of division between high and low quadrants. An
illustration of the Miller menu analysis model is depicted in Figure 1. Originally, Miller did not define a
method to calculate the division line between high volume and low volume (defined as popularity). Later
modifications of the MAM redefined the division axis to match the Kasavana and Smith model discussed
next. Miller segmented high/low volume using the 70/30 percent mark with 30 percent defined as high
volume and the remainder defined as low volume.
The division of the food cost axis between high and low was determined based on the average
food cost of all menu items included in the analysis. Menu items falling into the quadrant named Winners,
the most desirable quadrant, are high in popularity and low in food cost. Menu items located in the
quadrant labeled Marginals II are high in both food cost and popularity. Retooling Marginals II recipes to
decrease food cost or increasing the sales price can bring a Marginals II item into the Winners quadrant.
Marginals II items may also be discarded if recipe modifications are not feasible or practical. Menu items
classified into the Marginals III quadrant are low in food cost but lower in popularity as compared to the
items in the Winners quadrant. Restructuring the menu placement of these items and adjusting advertising
may increase the visibility of Marginals III menu items and popularity moving these menu items into the
Winners quadrant. Finally, menu items grouped into the quadrant labeled Losers are high in food cost and
low in popularity. Generally Losers menu items should be discarded from the menu. Following a menu
analysis with the MAM and subsequent menu changes, a reanalysis of the menu is appropriate.

FIU Review Vol. 25 No. 2 Page: 75

Figure 1: The Miller Matrix

Miller Model

Winners Marginals II

Average Food Cost

Volume

70% average units sold

Marginals III Losers

Food Cost Percentage

Model
Since menu items are categorized into quadrants based on average food cost and popularity, any
retooling of the menu will cause menu items to shift relative to the original analysis. Some previously
suitable menu items will fall into less desirable regions. No ideal distribution of Winners, Marginals II,
Marginals III, or Losers was reported in the literature, however Miller suggested that 60% of the menu
items in the Winners and Marginals III, the low food cost categories, was an appropriate goal.
Kasavana and Smith Model
A modification to the MAM was developed by Kasavana and Smith in 1982, depicted in Figure 2.
Food cost on the x-axis was replaced with individual menu item contribution margin and average
contribution margin replaced average food cost as the axis between quadrants. Contribution margin was
defined as sales price subtracted from direct costs. Because contribution margin only consists of sales
price less food cost in this analysis, contribution margin should be considered gross profit as defined by
Generally Accepted Accounting Principals (GAAP). LeBruto, Quain, and Ashley addressed the impact of
this misrepresentation of gross profit as contribution margin. When using contribution margin (gross
profit) instead of food cost, the quadrants shift locations in the matrix relative to the MAM. The ideal
menu item location in the Kasavana and Smith model is in the upper right quadrant instead of the upper
left quadrant in the Miller model. This shift conceptually occurs because the higher the contribution
margin of the menu item the more profitable the menu item. Conversely, in the Miller model the lower the
relative food cost the more profitable the menu item. Average contribution margin, the x-axis in the
Kasavana and Smith model, was calculated by taking the total weighted contribution margin (gross profit)
and dividing it by the total number of units sold. The total weighted contribution margin (gross profit)
was calculated by summing of the total contribution margin (gross profit) for each menu item and dividing
the contribution margin by the total number of menu items sold. Total contribution margin (gross profit)
for each menu item is calculated by multiplying individual menu item contribution margin (gross profit) by
the number of units sold. The objective in using contribution margin (gross profit) to replace food cost
was a shift in emphasis from raw costs to the profitability potential of a menu item.

FIU Review Vol. 25 No. 2 Page: 76

Figure 2: The Kasavana and Smith Model

Kasavana and Smith Model

Plow horses Stars

Average Contribution Margin

Volume
70% average units sold
Dogs Puzzles

Contribution Margin

An inherent danger of minimizing food cost, however, was pointed out by Pavesic. A menu item
designed from the perspective of achieving the lowest overall food cost percentage may cause an operation
to sacrifice total sales revenues, Low cost items are generally low priced items and typically have low
contribution margins (gross profit). Food cost does not focus management attention on maximizing sales
per guest or on flow through dollars that cover other costs and eventually result in net income.
Kasavana and Smith renamed the quadrants in the revised model which became known as the
Menu Engineering Model (MEM). The premise with the MEM is similar to the MAM with a goal to
increase the profitability potential of the menu. The Winners quadrant was renamed Stars; the most
desirable quadrant. These menu items have a high contribution margin (gross profit) with a high sales
volume. The quadrant named Marginals II by Miller was renamed Plowhorses. These menu items have a
high sales volume but a low contribution margin. The quadrant Marginals III by Miller was renamed
Puzzles. These menu items have a high contribution margin (gross profit) but a low sales volume. The
quadrant named Losers by Miller was renamed Dogs. These menu items exhibit a low contribution margin
(gross profit) and a low sales volume. Menu decisions relative to the recipe and advertising changes based
on quadrant categories are similar to those discussed with the MAM.
The primary limitation of both the MAM and the MEM relates to sales volume. Below an
established minimum sales level, a menu item with a favorable contribution margin cannot generate
sufficient profit to cover the total costs since contribution margin is not weighted by sales volume. (Bayou
and Bennett, 1992; Dearden, 1978.) A second limitation is an inherent danger of using contribution margin
in an effort to increase sales. This strategy may favor higher priced menu items that eventually decrease
demand and total operational profitability.
Pavesic Model
In an attempt to overcome the shortcomings of using either food cost or contribution margin
(gross profit) in menu analysis, Pavesic incorporated a weighted average contribution margin (gross profit),
also called “profit factor,” and food cost into the Cost Margin Analysis Model (CMAM). The profit factor
variable combines contribution margin (gross profit) and sales volume previously addressed in the MEM.
According to Hayes and Huffman and Pavesic, CMAM encompasses the three key elements of sales
volume (popularity, food cost percentage, and dollar contribution margin). Including these three elements
provides an unbiased perspective for making menu-pricing and positioning decisions.
In the CMAM the x and y-axis quadrants were renamed Figure 3. The CMAM model
contribution margin is again defined as sales less food cost. The y-axis values are redefined as the weighted
dollar contribution margin and the x-axis was renamed the potential food cost percentage. The weighted
contribution margin accounts for the popularity of the menu item and the contribution margin (gross

FIU Review Vol. 25 No. 2 Page: 77

profit) by multiplying the contribution margin (gross profit) by the number of items sold. The average
weighted contribution margin is the point of dissection on the y-axis from the high quadrant to the low
quadrant. It is calculated by summing individual contribution margins for all menu items together and
dividing that total by the total number of menu items sold. The potential food cost along the x-axis is
used to segment the high and low food cost quadrants. Potential food cost is calculated by dividing
weighted (total) food cost into the weighted (total) food sales as described with previous models.
Mathematically, potential food cost is actually a weighted average food cost and is often called standard
food cost by other authors. (Miller and Pavesic, 1996.)
Figure 3: The Pavesic Model

Pavesic Model

Primes Standards
Weighted Contribution Margin

Potential Food Cost

Average Weighted
Contribution Margin
Sleepers Problems

Food Cost Percentage

The quadrant named Primes contains those menu items with a low food cost and a high weighted
contribution margin (gross profit). Primes are equivalent to the Winners in MAM and the Stars in MEM.
The quadrant labeled Sleepers identifies menu items with a low food cost and a low weighted contribution
margin. The decision process relative to recipe or menu changes mirrors the quadrant labeled Marginals III
in the MAM and the Puzzles in the MEM. The quadrant with a high food cost and a high weighted
contribution margin is labeled Standards. The menu item decision making process is comparable to the
Marginals II in the MAM and the Plowhorses in the MEM. Finally, the quadrant with the high food cost
and low weighted contribution margin called the Problems in the Pavesic model compares to the Losers in
the MAM and the Dogs in the MEM.
The three models discussed are compared in a decision matrix (Table 1). Quadrant labels and
menu decisions of the three models are included in the figure. The CMAM and MAM quadrants have
similar meanings and locations in the matrix. As previously discussed, the MEM quadrants are transposed
in location relative to the other two models. All three models focus on measures that include food cost
and sales volume in varying degrees. However, none of the models effectively include other costs of
production, most notably absent are labor costs. When labor is subsequently incorporated analysis, menu
items previously grouped into the desirable quadrants in any of the three models may become less
desirable and possibly undesirable from a profitability perspective. Likewise, menu items that group into
less desirable quadrants without consideration of labor costs may move into a more profitable
classification when labor costs are included. A similar argument can be made for the popularity factors
such as menu design, marketing efforts, customer satisfaction, dining ambience, and other factors that are
difficult to quantify. Since many other factors are more qualitative and often difficult to measure they are
frequently excluded in menu analysis models.

FIU Review Vol. 25 No. 2 Page: 78

Table 1:
Model Author Corresponding Quadrant
Menu Analysis Miller Winners Marginals II Marginals III Losers
Menu Engineering Kasavana and Stars Plowhorses Puzzles Dogs
Smith
Cost/Margin Pavesic Primes Standards Sleepers Problems
Analysis
Decisions(s)
Option 1 Keep menu Reduce costs by Promote menu item by Delete menu
item as is retooling recipe advertising and/or item
menu placement
Option 2 Increase price Decrease price
Option 3 Do nothing Do nothing
Option 4 Delete menu item Delete menu item

Hayes & Huffman Model

As an alternative to the matrix model format, Hayes and Huffman developed a method of menu
analysis that focused on meeting net profit goals by creating profit and loss statements for each menu item.
Their goal was to account for fixed and variable costs to overcome flaws of the three aforementioned
matrix menu analysis models. Variable costs were included to assess the profitability of each menu item.
To simplify cost allocation, fixed costs are divided evenly by the number of menu items. The profit and
loss method relied on allocation of both fixed and variable costs. However, the allocation method to assign
variable costs was not explained. The basic formula used by the Hayes and Huffman model was as
follows:
Sales of menu item – (cost of food + fixed costs + variable costs) = net profit for menu item.
In this formula sales were determined as the total sales of a menu item for a predetermined time
period such as a month. Fixed costs in the formula were calculated by dividing total fixed costs from the
operation by the total number of menu items. The same amounts of fixed costs were allocated to each
menu item regardless of any other considerations. Variable costs in the formula were allocated based on an
assigned average variable cost for the entire operation established at 35 percent.
Additionally, Hayes and Huffman argued that the matrix approach and use of averages to
segment menu items into groups was undesirable since minor modifications to the menu could cause
menu items to change ranking. An endless evaluation process would result. They argued that every menu
item must stand on its own from a profitability perspective regardless based upon each menu item
compared with other items on the menu. If an individual menu item met organizational financial goals,
then it remained on the menu. Conversely, if the menu item did not achieve an adequate theoretical return
the menu item was discarded.
LeBruto, Quain and Ashley Model
In an effort to incorporate variable labor costs, LeBruto, Quain, and Ashley modified the MEM.
They proposed that errors in the analysis of menu engineering data could result from ignoring the profit
factors while relying solely on the placement of menu items in the MEM. Additionally they acknowledged
Looft’s (1989) observation about the difficulty in determining labor costs on a menu item basis. As an
alternative to separating labor into fixed and variable components and subsequently attempting to allocate
specific labor costs to each menu item, LeBruto, Quain, and Ashley took the Kasavana and Smith model
and subdivided each quadrant into high and low labor quadrants. The resulting matrix model contained
resulting in a total of eight sectors. Labor was separated in half into high and low segments with the menu
items falling equally into each category Figure 4. They recommend that labor assignment of the menu
items be made by either a food service professional or through employing a jury of execution, a technique
used in qualitative forecasting. The sectors of the quadrants were renamed to reflect the labor component.

FIU Review Vol. 25 No. 2 Page: 79

Figure 4: LeButo, Quain and Ashley Model
LeBruto, Quain, & Ashley Model
High Labor Plowhorse High Labor Star

Average Contribution Margin

Low Labor Tractor Low Labor Shining Star
70% Average Units sold
Volume

High Labor Ultimate Dog High Labor Brain Teaser

Low Labor Dog Low Labor Puzzle

Contribution Margin
Cohen, Mesika, and Schwartz Model
To solve limitations of two dimensional matrix menu analyses, Cohen, Mesika, and Schwartz
developed a multidimensional approach. They included food cost, price, labor cost, popularity, and
contribution margin into the menu analysis by incorporating a multifactor dish approach combined with
normalizing the input data. Normalization of data to a scalar variable ranging from 0 to 10 combined with
analyses of the five menu variables created a polygon dish. The menu variables were evaluated by three
ranges. The ideal range was defined as 8 to 10, the acceptable range was defined as between 4 to 8 and the
unacceptable range was defined as between 0 to 4. Menu items could be either evaluated independently or
compared to other menu items based on the item distribution within the polygon dish. Labor cost was
included as a factor in the multidimensional analysis similar to by Hayes and Huffman, Bayou and Bennett,
and LeBruto, Quain, and Ashley. Although labor was a factor in this model, Cohen, Mesika, and Schwartz
provided no explanation of how the labor cost was measured or calculated. As with others who developed
menu analysis approaches, Cohen, Mesika, and Schwartz did not specify how the variables including labor
would be calculated nor did they consider other factors of production. Incorporating labor as well as other
costs into the menu analysis has been difficult.
Bayou and Bennett Model
Bayou and Bennett argued that the previous menu analysis models for the most part lacked one
or more of three essential ingredients: analysis by menu groups, analysis by meal periods, and not
differentiating between the short run (special order pricing) and long run (make or buy items) profitability
analysis. They argued that allocation of costs was not as simple as evenly assigning costs to all menu items.
Simplification was especially pragmatic when various menu items were sold in different categories at
different meal time periods. Although Bayou and Bennett disagreed with the methodology of the Hayes
and Huffman, conceptually they agreed that labor factors must be considered in the analysis. Bayou and
Bennett stated that cost allocation was essential in any analysis of menu items however; they did not
propose a solution to the allocation problem in the Profitability Analysis Model (PAM). They proposed
first to subdivide the menu into simple parts. Meal periods were assigned as breakfast, lunch or dinner.
Product groups were assigned as appetizers, entrées, and desserts. Finally, the menu items in each group
were analyzed. The PAM included direct fixed costs in the individual menu item analysis but excluded
other fixed costs that Bayou and Bennett defined as common fixed costs from the primary analysis. They
defined direct fixed cost as those cost that can be directly attributed to a meal period. In their study, direct
costs consisted of full and part-time labor. They concluded that advertising and utilities directed to a
specific meal period could also be included. They further defined common fixed costs as those costs
remaining regardless of meal period. Examples of common fixed costs included non-shift specific labor,
utilities, advertising, and total maintenance costs.
Horton Model
Horton proposed another approach to include labor into the matrix analysis. Horton again
modified the MEM by including estimated labor into the contribution margin (gross profit). Horton
measured labor cost by multiplying active labor the time to prepare a menu item by the hourly labor cost
of the production employee performing the task. For example, the time it takes a cook to place a
hamburger on the grill, remove it after cooking and prepare the set-up of bun and assorted
accompaniments is multiplied by the dollar hourly rate for that employee. Horton did not include any of
the other preparation costs associated with the hamburger preparation such as ordering the products,
prepping the hamburger, or prepping accompaniments. Horton compared the menu of an independent

FIU Review Vol. 25 No. 2 Page: 80

restaurant analyzed twice using the Kasavana and Smith model, once with the contribution margin
inclusive of labor and a second time with contribution margin exclusive of labor. The inclusion of labor in
the contribution margin calculation changed classifications of six of the 52 menu items analyzed. Four
menu items dropped in classification while two improved. Horton concluded that labor may be an
important factor to include in future studies.
Alternative Approaches to Menu Analysis
Although these models are useful, more sophisticated approaches to menu engineering are
warranted in response to slim profit margins and increased competition. As an alternative approach to
assign exact labor dollars to an individual menu items must be considered. Managers must be able to
evaluate menu items based on labor attributes assessing all of the steps in the food flow, from the ordering
and receiving of ingredients to the final production and plating. Identification and quantification of these
labor attributes or factors could be used to develop a model that compares and examines menu items
using a multi-dimensional approach. Such a model could assess the relative strength of each menu item in
terms of profitability.
In a full-service restaurant approximately thirty to thirty-five percent of every dollar earned is
spent on labor. The majority of labor costs are associated with kitchen or production staff. Evaluating
labor productivity is common in food service operations. Restaurant operations generally measure or track
labor productivity as a percentage of sales and do not formally consider other cost/productivity factors
such as utilities, equipment efficiencies, pre- or partially prepared goods for production, and training in
organizational productivity analysis. Brown and Hoover (191) recommended institutional food service
organizations utilize total factor productivity or multi-factor productivity as a means of more effectively
measuring labor, materials, and more specifically production costs. They suggested a singular focus on
labor productivity may result in poorer overall productivity and decreased profitability. In the general
business literature there has been a shift from labor productivity to total factor productivity, multi-factor
productivity and DEA. (Reynolds, 2004.)
Although much of the menu analysis research has emphasized the importance of including labor,
minimal success has been reported. The actual dollar costs of labor are extremely difficult to precisely
calculate and assign to single menu item throughout the entire production process. One individual may
prepare many items simultaneously therefore making it difficult to differentiate labor among menu items.
Since menu items have been compared based on a function of food cost and consumption using menu
engineering approaches rather than cumulative dollar values of all costs, an efficiency comparison
approach should be considered. Similar difficulties in cost assignment exists relative to the costs of
inventorying, ordering, receiving and stocking the items are difficult to allocate. (Kiefer and Kelly, 1994.)
Menu items affect labor and other production costs from a number of perspectives. Variables such as
intensiveness of production, complexity of individual recipes, number of ingredients, form of ingredients,
and the number of independent ingredients are all relevant to a comprehensive analysis. To accurately
evaluate the effectiveness of a menu and individual menu items, a number of factors too difficult to obtain
in a quantitative or precise measure should be included using a qualitative and/or categorical level
measure.
Data Envelopment Analysis (DEA) has the capability to incorporate both quantitative and
qualitative factors. DEA has already been used to measure gross national product, modified to evaluate
firm level performance and its methodologies could be adapted to examine menu efficiency. Although not
presently applied to menu engineering and analysis, DEA could be a suitable technique can incorporate
multiple complex sets of factors impacting menu profitability.
A more comprehensive menu analysis model that builds on the strengths of matrix approaches
while permitting a comparison of categorical menu items together by incorporating food cost, popularity,
contribution margin and a well defined labor component is the next step in the evolution of a
comprehensive menu analysis model. Finally to address the dilemma of accurately reflecting labor
components, DEA analysis could be used to rank and measure efficiency of menu items relative to labor
by evaluating qualitatively measure components of labor; eliminating the labor allocation problem that
previous models have encountered. The resulting use of a qualitative menu analysis approach could
provide an operational decision maker with a systematic assessment of menu items to identify the most
efficient menu items. Assessment of the efficiency of all other menu items is also possible. Furthermore, a

FIU Review Vol. 25 No. 2 Page: 81

DEA based menu analysis model can pinpoint specific components such as labor functions or food costs
that require adjustments to improve efficiency.
In addition to capturing and assessing labor costs, DEA could be used to evaluate other variables
such as popularity, customer satisfaction, and marketing. Production costs are as important to the menu
analysis as are the factors that affect menu item popularity. There has been a great deal of research on the
consumer behavior aspect of menu pricing and menu design and how these affect the popularity of
individual menu items. Numerous authors (Carmin and Norkus, 1990; Kiefer and Kelly, 1994; Kreul,
1982; Miller, 1996; and Orkin, 1978) have all examined how menu pricing affects menu item popularity.
Menu design has been researched by Doerfler (1978), Stoner (1988), and Miner (1996.) Popularity can also
include point-of-sale advertisement such as table tents and/or server recommendations and external
marketing activities that could include direct advertisement. The popularity side of the menu item analysis
has yet to be introduced into any menu analysis model. A comprehensive model will need to consider how
menu design and marketing affect menu item effectiveness. However, the first step toward a fully
comprehensive menu analysis is evaluating the impact of labor costs on individual menu items. Once a
qualitative approach is identified for labor, other qualitative factors or variables may be added and a
comprehensive menu analysis model can be derived.
References
Atkinson and Peter Jones, “Menu engineering: managing the foodservice micro-marketing mix,” Journal of Restaurant
and Foodservice Marketing 1, no. 1 (1994).
Bayou and Bennett, “Profitability analysis for table-service restaurants,” Cornell Hotel and Restaurant Administration
Quarterly 22, no 2. (1992).
Brown and Hoover, “Total factor productivity modeling in hospital foodservice operations,” Journal of the American
Dietetic Association 91, no. 9 (1991).
Carmin and Norkus, “Pricing strategies for menus: Magic or myth,” Cornell Hotel and Restaurant Administration Quarterly
31, no. 3 (1990).
Cohen, Mesika, and Schwartz, “A Multidimensional approach to menu sales mix analysis,” Praxis 2, no. 1 (1998).
Dearden, “Cost accounting comes to service industries,” Harvard Business Review 78, no. 5 (1978).
Doerfler, “Menu design for effective merchandising,” Cornell Hotel and Restaurant Administration Quarterly 19 no. 3,
(1978).
Hayes and Huffman, “Menu analysis, a better way,” Cornell Hotel and Restaurant Administration Quarterly 25, no. 4 (1985).
Horton, “Menu analysis: The effect of labor and menu category on menu classifications,” FIU Hospitality Review 19, no.
2 (2001).
Kasavana and Smith, Menu engineering: A practical guide (Lancing, MI: Hospitality Publishers, 1982).
Kiefer and Kelly, “Menu pricing: An experimental approach,” Journal of Business and Economic Statistics 12, no. 3 (1994).
Kotschevar, “Menu analysis: Review and evaluation,” FIU Hospitality Review 12, no. 2 (1994).
Kreul, “Magic numbers: Psychological aspects of menu pricing,” Cornell Hotel and Restaurant Administration Quarterly 23,
no. 2, (1982).
LeBruto, Quain, and Ashley, “Menu engineering: A model including labor,” FIU Hospitality Review 13, no. 1 (1995).
Looft, D., Proceedings from the Second Annual AHFME Research Symposium, Enhancing current menu sales mix analysis for
the 90’s and beyond.(New York, New York. 1989).
Miller and Pavesic, Menu Pricing and Strategy (Reinhold, NY: Van Nostrad 1996).
Miller, Menu Pricing and Strategy (Reinhold, NY: Van Nostrad, 1987).
Miner, “Customer focused menu marketing,” Cornell Hotel and Restaurant Administration Quarterly 37, no. 3 (1996).
Orkin, “An integrated menu-pricing system,” Cornell Hotel and Restaurant Administration Quarterly 19 no. 2, (1978).
Pavesic, “Cost-Margin analysis: A third approach to menu pricing and design,” International Journal of Hospitality
Management, 2, no. 3 (1983).
Reynolds, “An exploratory investigation of multiunit restaurant productivity assessment using data envelopment
analysis,” Journal of Travel and Tourism Marketing 16, no. 3 (2004).
Stoner, “Menus: Design makes the difference,” Lodging Hospitality 42, no. 9, (1988).

About the Authors: James J. Taylor, PhD is in the Department of Family and Consumer Sciences, where
he is Program Director of Hospitality Management at the University of Mississippi, University, MS.
Denise M. Brown, PhD, RD, LD, is an Associate Professor in the Department of Nutrition and Food
Systems at the University of Southern Mississippi, Hattiesburg, MS.

FIU Review Vol. 25 No. 2 Page: 82