p–ISSN: 2723 - 6609 e-ISSN: 2745-5254

Vol. 5, No. 8 August 2024 http://jist.publikasiindonesia.id/

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2894

A Proposed Menu Engineering-Based Business Intelligence

Design Using K-Means Algorithm

Jonathan Shinray Fang

, Dwi Hosanna Bangkalang

Universitas Kristen Satya Wacana, Indonesia

Email:

[email protected]

*Correspondence

ABSTRACT

Keywords: menu

engineering, k-means,

business intelligence,

MSMEs.

Business continuity is greatly influenced by the menu,

especially in the culinary industry. Micro, Small, and

Medium Enterprises (MSMEs) account for 64.1 million

units or around 99% of all businesses in Indonesia. For many

years, MSME has been using menu analyses to keep its menu

optimized. However, this is not enough since 50% of

MSMEs are failing in their first 5 years due to poor decision-

making as a result of a lack of knowledge. Based on that

problem, there is a necessity for menu analysis and tools to

assist in decision-making, also called Business Intelligence.

The method used consists of three stages: data collection,

business intelligence design, as well as analysis and results.

The BI design focuses on menu engineering using the K-

Means algorithm to divide menu items into four unique

clusters according to Kasavana-Smith's menu engineering

concept. After validating its findings with the Davies-

Boudlin Index evaluation, it concludes that a four-cluster

solution is most optimal among other value-cluster. This

study aims to assist business owners in making better

decisions, and it may be used as a reference for business

owners by providing suggestions based on the menu review

analysis.

Introduction

The menu is one of the things that determines the continuity of a business,

especially in the culinary business. (de Riandra & Islam, 2021) A well-crafted menu can

provide large profits for the company and better product information to customers. The

rapid business competition makes menu optimization necessary to maintain company

suitability, considering that most aspects of the industry have a risk of going out of

business by almost 50% in the first 5 years (Zebua et al., 2023). One of the main causes

of business failure is menu incompatibility. Hence, the performance of menu

requirements must be carried out to maintain business continuity (Sutaguna et al., 2023).

Micro, Small, and Medium Enterprises (MSME) are the type of company that

mostly runs in Indonesia. Based on data from the Ministry of Cooperative and Small

Medium Enterprises (SME), the number of MSMEs in Indonesia has reached 64.1 million

A Proposed Menu Engineering-Based Business Intelligence Design Using K-Means Algorithm

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2895

units, or around 99 percent of the total number of business actors in this country (MM,

2007). For many years, MSME has conducted menu analyses as part of its operations to

keep its menu optimized and according to customer needs. Nowadays, POS systems are

used as cashier systems by the majority of MSMEs (Bahwita, 2022). It can compile sales

information and generate a sales report based on transactions. However, the majority of

MSMEs that neglected to conduct strategic planning or menu analysis typically fail as a

result of a lack of knowledge leading to poor decision-making (FADLY, 2022).

Without loss of generality, better menu analysis output can be achieved when there

is more data provided as input. While the POS System can access historical transaction

data, the MSME can use it to gain a more precise menu analysis to increase profitability

and customer satisfaction (Laeliyah, 2017). But even though MSME has previously

adopted the POS System to provide more data in menu analysis, it cannot ensure MSME's

survival. There are still certain issues with operations and business. If the business owner

is not skilled enough or doesn’t even have the employee to process and analyze the data,

the MSME is likely to make poor decisions and, worse, it will fail (Haryadi, Rojali, &

Fauzan, 2021). All of the effort put into digitalization with the POS system will be in vain

if it continues in this manner—it will only appear to be a fancy receipt. All of that comes

to the needs of BI, the BI will help businesses in menu analysis by interpreting the

transaction data, so it will come in handy for SME. Also considering the Point of Sales

(POS) system adaptation in MSME, it is very wise to build a system while still making it

an option to be integrated with the POS system (Sumarto, 2023).

One of the most used menu analysis techniques in the business is Kasavana-Smith’s

menu engineering. It is a special technique that is known to be able to see the performance

comparison of items on the menu. By knowing the performance of the menu, it is possible

to estimate future sales and make decisions based on marketing strategies. As introduced

by Kasavana and Smith, this technique grouped items on the menu in four quadrants

formed in a 2x2 matrix: Star, Plowhorse, Puzzle, and Dog. The 2x2 matrix of menu

engineering can be seen in Fig. 1

Fig. 1. Menu Engineering Matrix

The analysis of menu engineering focuses on two elements: menu mix analysis and

item contribution margin. The menu mix represents each menu item's popularity level,

and the contribution margin represents the difference between the selling price of the

menu and the cost of goods.

Jonathan Shinray Fang, Dwi Hosanna Bangkalang

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2896

One of the latest research projects conducted by Atmaja et al. aimed to assist the

management of Swell Bar & Restaurant in evaluating menus by analyzing the relationship

between the level of popularity and profitability of each menu item. The research used

saturated sampling with 50 menu items. The research recommends using regular analysis

to enhance revenue and improve the menu by minimalizing the number of those two

categories.

Another menu engineering research that Attwood. et al. conducted regarding price-

based decoys in menu engineering to promote menu items by marking up the prices. The

result is that the decoy didn’t significantly influence customer choice. Also, the study

highlights that further research is needed to determine which attributes, such as taste,

portion size, or signature ingredients, are effective in promoting menu items.

Based on the current problems, a BI must exist to support business owners in menu

engineering. The purpose of using the K-Means approach in menu engineering clustering

is to improve efficiency by speeding up and automating the process. Besides the data

visualization, menu review should also be included in the BI to assist business owners by

making suggestions on menu items. Where this decision is based on current data or at a

certain period. The analysis is based on market trend data to increase company sales and

income. So, it is not only based on individual assumptions and hunches.

This research aims to propose a software design like BI that can provide insight into

menu sales and can be integrated with the POS system. Implementing BI for menu

engineering will help business actors, especially SMEs, to make better decisions for a

better chance of survivability in today’s highly competitive business world and

experience an increase in sales.

Research Methods

This study uses the dataset from a POS system in an SME. The development steps

that are used in this study are described in Fig. 2.

Fig. 2. Development Step

Data Collection

A Proposed Menu Engineering-Based Business Intelligence Design Using K-Means Algorithm

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2897

MSMEs, or in Indonesia, Usaha Mikro Kecil dan Menengah (UMKM), are

businesses owned by individuals, households, or business entities with assets and a yearly

turnover below Rp 10 billion.[19] While the income generated from this kind of business

is relatively small, it’s not surprising that most MSMEs run from their own homes.[20]

Businesses that are categorized as MSMEs can be anything from food carts to service

businesses. MSME itself has a massive impact on the Indonesian national economy.

Therefore, MSME is divided into three types: microenterprise, small enterprise, and

medium enterprise.

The POS system is a modern cash register system commonly used in business to

complete sales transactions. There are many third-party software programs nowadays.

MSME is mostly the one using it since it is affordable with a low budget, and it comes

with a subscription, so business owners don’t have to worry about its maintenance.[21]

The dataset used in this design came from sales data from one of the MSMEs in

Indonesia in the period January–December 2023. As shown in Table 1, the dataset had

118 menu items and 29,779 transaction data. It’s important to know the variables that are

in the POS system’s report of sales data.

Results and Discussion

Business Intelligence Design

This section explains the design of the proposed BI regarding the menu engineering-

focused BI, the unified model diagram for this BI, the system architecture that will

construct this BI, and the mockup for visualization of the BI.

Menu engineering is a special technique that is known to be able to see the

performance comparison of items on the menu[5]. The menu engineering model was first

introduced by Kasavana and Smith. There are four categories determined by menu

popularity and contribution margin.[22] The formula used to classify a menu can be seen

below.

To calculate the Contribution Margin (CM)

    (1)

To calculate menu popularity or Menu Mix (MM)

  





 (2)

The data has to be transformed into a suitable table using the menu engineering

approach by calculating CM and MM%. Referring to formula (1) to calculate CM, this

table needs the difference between the Harga Jual and Harga Pokok variables. Referring

to formula (2) to calculate MM%, this table needs to count the Jumlah based on the Kode

Barang variable divided by the total of the Jumlah variable times 100%. The output table

is shown in Table 2, where this study adds ID as an alias to Kode Barang.

Table 2

The Dataset Post-Data Aggregation Using Menu Engineering Approach

Kode Barang

MM%

Jonathan Shinray Fang, Dwi Hosanna Bangkalang

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2898

8991914101056

2251.507

4.20%

00086

955.690

3.99%

8886008101053

1638.358

3.06%

000071

1720.189

2.71%

8886008101091

1977.574

2.60%

…

Modelling

One of the suggestions in the earlier research paper is to make menu engineering

segmentation using an algorithm. For a brief explanation, the dataset was first

standardized using Z-Score Normalization, then using the K-Means unsupervised

algorithm to label the data for the next classifying process in the K Nearest Neighbor

supervised algorithm.

Modelling

One of the suggestions in the earlier research paper is to make menu engineering

segmentation using an algorithm. For a brief explanation, the dataset was first

standardized using Z-Score Normalization, then using the K-Means unsupervised

algorithm to label the data for the next classifying process in the K Nearest Neighbor

supervised algorithm.

The formula for Z-Score normalization to standardize the dataset in this study is

seen in formula (3).

xb= (xa-x )/σ (3)

Where:

xb = new value

xa = old value

x  = average

σ = standard deviation

To calculate K-Means, the first step is deciding how many n clusters are needed. In

this study, where Kasanava menu engineering is applied, it is clear that four clusters are

needed. For the initial cluster centroids, this study will take four IDs. Then calculate the

distance between objects with the initial centroids, there are several methods to calculate

this, in this study uses Euclidean distance, as seen in formula (4).

d= √(〖(x_2-x_1)〗^2+〖(y_2-y_1)〗^2 ) (4)

Then, to complete it, this manual computation will have different centroids and

measure it all along again until there’s no change in the output. The final output table is

shown in Table 3.

Table 3

K-Means Labelled Dataset

MM%

Label

2251.507

4.20%

955.690

3.99%

1638.358

3.06%

1720.189

2.71%

1977.574

2.60%

…

A Proposed Menu Engineering-Based Business Intelligence Design Using K-Means Algorithm

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2899

To evaluate the K-means model, it applies the Davies-Bouldin Index (DBI). DBI is

a method to evaluate cluster performances by examining the maximum distance between

clusters and, at the same time, decreasing the distance between cluster members.[23] The

main concept of DBI is that clusters are more optimal the cluster formed when they have

a smaller score.[24] A positive DBI value that is close to zero is considered more

favorable. The performance results of the initial cluster and the DBI score are shown in

Table 4.

Table 4

Davies-Bouldin Index

n-cluster

DBI score

5.676

3.474

1.093

2.436

13.649

In Table 4, it seems that among the others, the 4 clusters have the lowest DBI score,

so it can be concluded that the 4 clusters either have a high degree of similarity or produce

a low variance within their cluster. In the beginning, the scope of this study was to design

a menu engineering BI. According to Kasavana-Smith’s method, menu engineering only

has four categories. By evaluating using DBI, the difference between each n-cluster is

shown based on the score, and it seems the best k is still obtained by using four clusters.

Table 5 displays the result from the K-Means cluster, with each cluster detailed by

cluster count, CM detail, and MM% detail.

Table 5

K-Means Cluster Detail

Cluster

MM%

Count

Average

Min

Max

Average

Min

Max

1038.559

303.1579

1645.161

0.005617

0.003151

0.013809

5083.305

3747.7

6798.942

0.009324

0.003336

0.021779

1651.488

930.6122

2409.355

0.024694

0.016867

0.042014

2178.274

1557.93

3449.682

0.006705

0.003182

0.013933

Based on Table 5, this study has generated a category label for each cluster in menu

engineering; the star category is judged by the highest maximum CM times MM%, the

dog category by the lowest CM times MM%, and the other two categories are judged by

compare, where the higher CM becomes the puzzle and the other is plowhorse. The results

are presented in Table 6.

Table 6

K-Means Cluster Categories

Cluster

Categories

Total Count

Dog

Star

Plowhorse

Puzzle

Menu Effectiveness Review

Jonathan Shinray Fang, Dwi Hosanna Bangkalang

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2900

Fig 7 shows a menu effectiveness review framework

This review is necessary to determine the changes necessary to improve menu

items. As proposed by Jones, P., & Mifli, M., the activities in this framework comprise

promotion, repositioning, retention, eliminating, and modification. However, the focus of

this study is modifying activity which involving evaluate the presentation, pricing, re-

costing, and recipe.[25] Many researchers agree that menu prices have a significant

influence on clients' purchasing decisions.

Based on this review framework, the suggestion can be concluded into this study

four categories:

Table 7

Menu Effectiveness Recommendation

Categ

ory

Recommendations

Star

55, 56, 61,

65, 68, 69,

70, 76, 78,

79, 80, 81,

83, 84, 85

a. Maintain the performance of the

restaurant.(Jones & Mifll, 2001)

b. Maintain the appropriate recipe standards

for menu quality, quantities, and

appearance.(Tom & Annaraud, 2017)

c. Periodically raise selling prices while

accounting for rising demand from

customers and rivals' prices.(Setiyawati &

Hosanna, 2020)

d. Keep an eye on rising menu raw material

costs and make necessary adjustments to

menu selling prices.(Setiyawati & Hosanna,

2020)

Plowh

orse

1, 2, 3, 4, 5,

6, 7, 8, 62,

63, 64, 66,

67, 71, 72,

73, 74, 75,

77, 82

a. Maintain the appropriate recipe standards

for menu quality, quantities, and

appearance.(Tom & Annaraud, 2017)

b. Examining methods of cost reduction,(Tom

& Annaraud, 2017) such as keeping an eye

on the amount of ingredients ordered,

optimizing processing effectiveness, and

streamlining presentation while preserving

food quality and presentational appeal.

c. Gradually raise food costs while keeping an

eye on rising demand from customers and

rivals' prices.(Setiyawati & Hosanna, 2020)

A Proposed Menu Engineering-Based Business Intelligence Design Using K-Means Algorithm

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2901

Puzzle

57, 58, 59,

60, 87, 88,

89, 90, 92,

93, 95, 96,

97, 98, 99,

100, 101,

102, 103,

104, 105,

106, 108,

109, 110,

111, 112,

113, 115,

116, 117,

118

a. To attract clients, offering deals on food

delivery services and/or advertising the

menu on social media

b. Putting the menu in a prominent location,

either on the menu display screen,(Tom &

Annaraud, 2017) or having the waiter

deliver it to customers when they place food

or drink orders to increase its popularity

c. Give the menu a new, more appealing name.

d. Review costing, and the next action that can

be made is to lower the food's selling price

by considering the state of profit margins,

the cost of goods supplied, and the prices

that competitors are charging

e. Take into account removing the menu

Dog

9, 10, 11,

12, 13, 14,

15, 16, 17,

18, 19, 20,

21, 22, 23,

24, 25, 26,

27, 28, 29,

30, 31, 32,

33, 34, 35,

36, 37, 38,

39, 40, 41,

42, 43, 44,

45, 46, 47,

48, 49, 50,

51, 52, 53,

54, 86, 91,

94, 107,

114

a. Cost reduction to larger the profit

b. Increase price to larger the profit

c. Promote to gain more popularity

d. Eliminate the menu (Tom & Annaraud,

2017); this policy serves as a substitute for

cutting significant business costs when it

can’t be pushed to other categories

e. Give the menu a new, more appealing name.

f. Put together a menu package from this

cluster with very popular foods and/or

beverages from other clusters.(Adiatma,

Andriatna, & Sudono, 2014)

For example, this study will use one of the dog category members in this menu, ID

14, who goes by the name Indomie Kuah Cakalang. By using this suggestion based on

the menu effectiveness review framework, this study recommends whether to increase

the price or remove the menu, as the cost cannot be reduced and will have no significant

impact on promotion.

Architecture and Design

At this stage, the author describes how a system works by modeling the system

using the Unified Modeling Language (UML), which is the industry standard for

designing, visualizing, and documenting a system, this helps the author to describe and

design software systems. The type of Unified Modeling Language (UML) used is a Use

Case Diagram, which is a diagram that describes functional requirements or activities in

a system and actors that are connected, and an Activity Diagram, which is a diagram that

describes the workflow or activities of a system.

Use Case Diagram

Jonathan Shinray Fang, Dwi Hosanna Bangkalang

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2902

This Use Case diagram illustrates the concept of functional requirements analysis,

in which some activities and actors are interconnected in the system.

Fig. 4. Use Case Diagram

Fig. 4 explains that in this system there is only one actor, namely the user and use

cases, which are activities that can be operated by users in this BI.

Activity Diagram

This activity diagram describes the concept of data flow or activity from the system

as shown in Fig.6.

Fig. 5. Activity Diagram

Fig. 5 explains the flow of generating menu engineering analysis. From the

beginning, when the user uploaded the sales data, the system kept it as a temporary data

frame. By aggregating the initial data to make a new dataset that applies to the rule of

menu engineering, it can be used to process the knowledge discovery by using the menu

engineering formulas. That resulted in a report that explains the insight from the sales

data.

A Proposed Menu Engineering-Based Business Intelligence Design Using K-Means Algorithm

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2903

System Architecture

The system architecture here focuses on creating website-based BI so that the

picture adapts to the development of the website, which is divided into two, namely the

front end where users interact, and the back end where all the computing and analysis

takes place.

Fig. 6. System Architecture

As shown in Fig.7, from the far right, there are recommended devices in the form

of tablets and laptops with landscape orientation. Refers to the postal system. This postal

system will generate user sales reports so that later users can enter data files on this BI

website. Continue is the final font. Where there is a browser, the technology used is a

bootstrap framework and programming language in the form of HyperText Markup

Language (HTML), Cascading Style Sheets (CSS), and JavaScript (JS). And for the

backend, it uses technology in the form of the Django framework and the Python

programming language, along with libraries that support BI computing.

Fig. 8. Aggregate Data Page Design

In Fig. 8, there is a display of the results of data aggregation that has been carried

out on the previous dataset, producing new variables that are by the formula of menu

engineering. On this page, the user also gets a brief explanation regarding their data, like

the count of menu items and the count of transactions.

Jonathan Shinray Fang, Dwi Hosanna Bangkalang

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2904

Fig. 9. K-Means Menu Clustered and Detail Page Design

In Fig. 9, this is the display of the menu-engineered categories, and when the next

button is clicked, it will show the user the details and the recommendations on each menu.

When hovering over one of the dots, it will tell the user the menu ID.

Fig. 10. K-Means Menu Clustered Category Recommendations

Fig. 10 shows a detailed menu of cluster categories with the recommendations for

each cluster category. This design also tells the ID of each menu in the category.

Conclusion

The goal of menu engineering with a website approach in business intelligence

design is to assist micro, small, and medium business owners in managing their

companies. The efficacy and efficiency of the K-Means method have been shown to grow

with its implementation. The average similarity has been assessed using the Davies-

Bouldin Index, and the results of this study indicate that the four clusters are the most

optimal, since it is in line with menu engineering that employs four categories. It is

intended that by offering helpful development suggestions and insights, businesses would

be able to thrive in this cutthroat environment. Adding the services required for business

management to business intelligence is the next action that may be taken. Suggestions for

A Proposed Menu Engineering-Based Business Intelligence Design Using K-Means Algorithm

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2905

further research include adding business intelligence features that are useful in the

engineering menu analysis process, as well as simply considering adding algorithms to

carry out analysis for more efficient results.

Jonathan Shinray Fang, Dwi Hosanna Bangkalang

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2906

Bibliography

Adiatma, Dani, Andriatna, Wendi, & Sudono, Agus. (2014). Analisis Perbaikan Menu

Unpopular Di Restoran Dapur Sunda Di Kota Jakarta. The Journal Gastronomy

Tourism, 1(1), 35–48. https://doi.org/https://doi.org/10.17509/gastur.v1i1.4654

Bahwita, Umma Sabrina. (2022). Potensi Start Up Wirausaha Muda Ekonomi Kreatif

Subsektor Kuliner Di Masa Pandemi Covid-19 di Kota Banda Aceh Ditinjau dari

Perspektif Ekonomi Islam (Studi Pada Urban Teahouse Lampineung Banda Aceh).

UIN Ar-Raniry.

de Riandra, Chealza Nuansa, & Islam, Muh Ariffudin. (2021). Perancangan identitas

visual catering & bakery cherry. BARIK-Jurnal S1 Desain Komunikasi Visual, 2(2),

43–56.

Fadly, Fadly. (2022). Upaya Pengembangan Kualitas Produk Dan Kinerja Bisnis Rumah

Makan Panjalu. Institut Agama Islam Negeri (IAIN) Palopo.

Haryadi, Rofiq Noorman, Rojali, Anda, & Fauzan, Mirja. (2021). Sosialisasi Penggunaan

Online Shop berbasis Website di UMKM Cimanggis. Jurnal Pengabdian

Masyarakat Madani (JPMM), 1(1), 10–16.

Jones, Peter, & Mifll, Mazalan. (2001). Menu Development and Analysis in UK

Restaurant Chains. Tourism and Hospitality Research, 3(1), 61–71.

https://doi.org/10.1177/146735840100300105

Laeliyah, Siti. (2017). Analisis Strategi Pengembangan Bisnis Kuliner Dalam Perspektif

Ekonomi Islam. Skripsi. Semarang: Universitas Islam Negeri Walisongo.

MM, Ir DJATI SUTOMO. (2007). Menjadi Enterpreneur Jempolan. Penerbit Republika.

Setiyawati, Nina, & Hosanna, Dwi. (2020). Penerapan Metode Two-Step Cluster Dalam

Analisis Menu Engineering Pada Usaha Kuliner the Implementation of Two-Step

Clustering Method in Menu Engineering Analysis in Culinary Business. Jurnal

Teknologi Informasi Dan Ilmu Komputer (JTIIK), 7(2), 359–366.

https://doi.org/10.25126/jtiik.202072012

Sumarto, Marco Alfan. (2023). Analisis dan Perancangan Aplikasi Point of Sale (POS)

untuk Usaha Mikro, Kecil, dan Menengah (UMKM) dengan Metode Rapid

Application Development (RAD). Jurnal Studi Komunikasi Dan Media, 27(1), 17–

34.

Sutaguna, I. Nyoman Tri, Par, S. S. T., Par, M., Winanto Nawarcono, S. E., Andayani,

Sri Umiatun, Sos, S., Noor, Laili Savitri, SE, M. M., Par, Ramdhan Kurniawan S.

S. T., & Evi Noviaty, S. E. (2023). MANAJEMEN STRATEGI BISNIS. Cendikia

Mulia Mandiri.

Tom, Mary, & Annaraud, Katerina. (2017). A fuzzy multi-criteria decision-making model

A Proposed Menu Engineering-Based Business Intelligence Design Using K-Means Algorithm

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 8, August 2024 2907

for menu engineering. IEEE International Conference on Fuzzy Systems.

https://doi.org/10.1109/FUZZ-IEEE.2017.8015612

Zebua, Rony Sandra Yofa, Hendriyani, Chandra, Sukmadewi, Rani, Thaha, Abdurrahman

Rahim, Tahir, Rusdin, Purbasari, Ratih, Novel, Nurillah Jamil Achmawati,

Dewintari, Putri, Paramita, Catarina Cori Pradnya, & Hierdawati, Trie. (2023).

BISNIS DIGITAL: Strategi Administrasi Bisnis Digital Untuk Menghadapi Masa

Depan. PT. Sonpedia Publishing Indonesia.