CHAPTER I INTRODUCTION 1.1 Description of the Problem Nowadays Indonesia as “industry country” that is so many company there. And in Indonesia, there are so many companies that offer retail sales of products owned by personal retail and other companies. One example of the personal is Indotoko. Here why Indotokocalled as personal, because the retail is standalone, not in every area is available, and not endorse with the other companies. Some examples of retail that endorse with other companies is Indomaret, Alfamart, 7eleven, Circle K, etc. because the retail almost available in every area that easily to find. Products are sold to virtually complete, ranging from baby supplies, personal equipment, household goods, and many more who may not mention here. Initially the company was just grew and grew up in Java. Because the magnitude of the needs of the community in consuming goods and services, then Indotokocompany started expanding. First time established, Indotokois just small retail. Then after the owner succeeds with the business, he was enlarging the retail become bigger and larger than before after some years established. Indotoko has been expanding
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CHAPTER I
INTRODUCTION
1.1 Description of the Problem
Nowadays Indonesia as “industry country” that is so many company there.
And in Indonesia, there are so many companies that offer retail sales of products
owned by personal retail and other companies. One example of the personal is
Indotoko. Here why Indotokocalled as personal, because the retail is standalone,
not in every area is available, and not endorse with the other companies. Some
examples of retail that endorse with other companies is Indomaret, Alfamart,
7eleven, Circle K, etc. because the retail almost available in every area that easily
to find. Products are sold to virtually complete, ranging from baby supplies,
personal equipment, household goods, and many more who may not mention here.
Initially the company was just grew and grew up in Java. Because the
magnitude of the needs of the community in consuming goods and services, then
Indotokocompany started expanding. First time established, Indotokois just small
retail. Then after the owner succeeds with the business, he was enlarging the retail
become bigger and larger than before after some years established. Indotoko has
been expanding into two another branches in Yogyakarta recently, but today
comeback again to be one and the only one retail, but it expanded to be larger.
The results of this expansion were to get a positive response from the community
who can be seen from the height of the existing Indotoko visitors in each area.
Then also, the loyalty is one of much important things that must we get from the
customers to make sure that the customers do not choose the other options of the
retail, we can call it fixed customers. In this era company has been growing up
even just Retail Company.
In this research we analyze about member card. While providing
discounts to the customer, these cards allow the retailer to develop a better
understanding of individuals' purchasing habits by associating customers with
transactions. The uses of this information vary, but may include informing
product placement decisions, designing personalized marketing campaigns, and
determining the timing and extent of product promotions among others. (Raeder,
Chawla)
1.2 Problem Formulation
1. How is the associative relations that happened between items in Indotoko?
2. How is the problems solutions of the retail card member in Indotokobased on
analysis of AR-MBA?
1.3 Research Objectives
The objectives of this research are:
1.To know the associative relations that happened between items in
Indotoko.
2. To know the solutions of the retail card member in Indotokobased on
analysis of AR-MBA.
CHAPTER II
LITERATURE REVIEW
2.1 Deductive Study
2.1.1 AR-MBA
Association in data mining is the work to determine which attributes will be
obtained simultaneously. In the business world the term is commonly known as
affinity analysis. The task of the asociation rule is to find a rule that does not
cover to measure the relationship between two or more attributes.
Association rule is a form if the "previous incident" and "consequences".
(IF antecedent, THEN consequent). Along with the calculation of support and
confidence of rules. The pattern of association to be one of the most interesting
functionality in extracting data (Kumar and Wahidabanu, 2007).
Association Rule is a data mining technique to find the associative rules
between combinations of items. Examples of Association Rule of purchases in a
purchase analysis is able to know how likely a customer buys the same coffee
with sugar. With this knowledge owner can adjust the placement of the goods or
designing a marketing campaign using a combination of discount coupons for
certain items.
One example of application of Association Rule is Market Basket
Analysis. Association Rule became known for its application to analyze the
contents of the purchase shopping cart, so the Association Rule is also commonly
referred to as Market Basket Analysis. Association Rule also known as one of data
mining techniques that became the basis of a variety of other data mining
techniques.
Each consumer buys a set of different items, in different amounts, and in a
different time. Market Basket Analysis using information of something that
purchased by consumers to provide a signal or information that is who they are
and why they made the purchase? Market Basket Analysis provides an
understanding of the merchandise by telling us which products are possible to be
purchased simultaneously and which product is approved to be promoted. This
information can be used in:
1. More profitable advertising and promotion. Market Basket Analysis using advertising
and promotion in order to understand better how shoppers respond to and
communicate over the products offered, for the purpose of the retailer "How do I
change this sale? What else is sold and what was advertised ".
2. More precise targeting in return ROI (Return on Investment). Market Basket Analysis
is used to optimize campaigns and promotions to increase sales and margins by
targeting more precise.
3. Loyalty card promotions with longitudinal analysis. Longitudinal Market Basket
Analysis enables users to buy the characters retailer customer behavior over time.
Retailers use loyalty cards to capture the lifecycle of data so that they can analyze the
purchasing behavior of customers such as shopping. For example a toy retailer
explained that he did not make sense to sell a game engine (with a slight margin)
except for customers who also buy the game software and accessories (high margin).
They use the Market Basket Analysis of loyalty card data to determine their overall
margin on sales of video games and promotions to make the memory of the customer
and affect a buyer to purchase games and accessories from them and not from other
retailers.
4. Determine the layout of the new store (new store layouts) or attract more traffic to the
store, set of products which will be placed in a special place. Market Basket Analysis
also uses the space to improve traffic planning and visual merchandising to boost sales.
5. Identify when the problem in pairs / coupon (issue coupons). To increase sales or
spending items into inventory.
There are a lot of definitions of Market Basket Analysis that use to known. Such
as Market Basket Analysis a mathematical technique used by marketing
professionals to express the similarity between the individual products or product
groups. Market Basket Analysis with respect to a set of problems related to flying
businesses to find out from the point of sale transaction data. Market Basket
Analysis is a general term for the methodology of the study of the composition of
the basket of groceries purchased by households during a time of shopping.
Market Basket Analysis is a collection of a combination of products purchased
together. Market Basket Analysis trend analysis of an item bought by the same
customer at the same time. Market Basket Data is data that describes the
transaction underlying the three different entities, namely Customers, Orders /
purchases, and Items (goods)
Introduction of the customer at any time make it possible to instantly
recognizable, such as the frequency of purchases made by the customer. Three
levels of market basket data is important to understand the request quickly. These
measurements give an idea for a business. In some cases, there are some repeat
buyers, so that the proportion of the purchase of every customer close to 1. This
suggestion is used a company to increase sales per customers. Or the amount of
each purchase of products close to 1, suggestion can be an opportunity for cross-
selling during the purchase process. Important whether or not an associative rule
can be determined by two parameters, support (the support) is the percentage of
the combination item page and confidence in the database (the certainty) that the
strength of the relationship between items in the associative rule.
So, our research held in Indotokoat Jalan Kaliurang km. 5, Sleman,
Yogyakarta. There are 100 pieces of receipt of goods purchases that already
collected. Conducting Pre Data Processing, Data Processing, Establishment of
Association Rule. For application master, we use Microsoft Excel, Microsoft
Visio, and Rapid Miner software.
2.1.2 Card Member
According to (Miguel, Camanho, Joao Falcao) The establishment of loyalty
relationships with customers became a main strategic goal for this company. The
development of the company’s information system and the implementation of a
loyalty program have enabled collecting data on each customer profile (e.g.
customer name, address, date of birth, gender, number of people in the household,
the telephone number and the number of one identification document) and on their
transactions (date, time, store, products and prices). This programm is supported
essentially by a loyalty card, and currently approximately 80% of the total number
of transactions is done by customers using the loyalty card. At present, the
company customers are segmented in two ways. One of them consists on grouping
customers based on their shopping habits. This segmentation model is a simplified
version of the RFM model proposed by Bult and Wansbeek [20], and is called
internally: “frequency and monetary value” (FM) model. According to the values
of these two variables, the company specifies 8 groups of customers. Each client
integrates one of these groups, according to the average number of purchases done
in a 8 week period and the average amount of money spent per purchase. The
changes in the percentage of customers belonging to each group are used to sinal
actions required in customer relationship management. For example, if the number
of customers in the clusters with more visits to the store decreases, the company is
alerted to launch marketing campaigns in order to motivate customers to go to the
stores more often. The other method of segmentation is based on customer
necessities and preferences. In this case, customers are grouped into 7 segments
according to the mix of categories and products they purchase. Each segment is
distinguished by the high relative weight of the purchases of certain products
category compared with others segments. This type of segmentation is used to
optimize the price and range of products sold in certain stores. Mining Customer
Loyalty Card Programs 9 Concerning the promotional strategy of the company,
there are mainly 3 types of policies:
1. discounts on specific products advertised in the store shelves and leaflets,
that are applicable to all customer with a loyalty card;
2. discounts on purchases done on selected days (percentual discount or
absolute
discount on the total value of purchases). These are applicable to
customers that present at the point of sale (PoS) the discount coupon sent
by mail;
3. discounts for specific products on selected days. These can be sent by mail
or issued at PoS.
The first two types of promotions do not differentiate between customers of
different segments. The third type, instead of using the segmentation models
previously described, uses a model based on the historical purchases of the
product included in the promotion. The discounts are only issued to the most
frequent buyers of the product, or to those customers who do not normally buy
the product, to encourage new buyers.
The analysis reported in this paper is based on transactional data of customers with a
loyalty card. The database used includes the records from the last trimester of
2009. Each transaction includes: the client identity number, the date and time of
the transaction, the product transacted and the price of the product. In addition to
the transactions information, the company provided demographic information for
each customer: residence postcode, city, date of birth, gender, number of persons
in the household. The preparation of the database for the exploratory analysis
involved the integration of the data from different sources, and the elimination of
the outliers. Customers whose average amount of money spent per purchase or
the average number of purchases per month is out of the range of the mean plus
three standard deviations were excluded from the analysis. As we are interested in
the design of promotions for households it was also decided to remove from the
database all customers whose average amount of money spent per purchase was
greater than e500. These represented 0.75% of the customers included in the
original database. Usually, purchases exceeding this value are done by small
retailers that resell the products in competing stores, so these customers are not
intended to be included in promotional programmes. After the selection process,
the database contained 2.142.439 customers.
2.2 Inductive Study
The field of market basket analysis, the search for meaningful
associations in customer purchase data, is one of the oldest areas of data mining.
The typical solution involves the mining and analysis of association rules, which
take the form of statements such as people that buy diapers are likely to buy
beer." It is well-known, however, that typical transaction datasets can support
hundreds or thousands of obvious association rules for each interesting rule, and
filtering through the rules is a non-trivial task. (Raeder, Chawla)
Association rule mining (ARM) is used for identification of association
between a large set of data items. Due to large quantity of data stored in
databases, several industries are becoming concerned in mining association rules
from their databases (Gupta, Mantora 2014). This analysis may be carried out on
all the retail stores data of customer transactions. These results will guide them to
plan marketing or advertising approach. Market basket analysis will also help
managers to propose new way of arrangement in card member
CHAPTER III
RESEARCH METHOD
3.1 Object of the Research
The company is a kind of retail, named Indotokoand the owner is Mr. Bambang.
Indotokohas stood since year 2009 in Jl. Karangwaru Lor, Yogyakarta. We use card
member analysis and minimum collection 101 receipts to know the department.
Company Profile
Name : Indotoko
Owner : Bambang Sudjatmiko
Address and phone : Jl.Karangwaru Lor
Post Code : 55514
Established : Year 2009
3.2 Collecting Data Method
In this research, the collecting data method used is:
a. Observation
Researcher do the observation directly in the Indotokoto know how the layout of the retail.
b. Interview
Researcher do interview to the owner to know the profile of the company and to permit
for allowance of collecting receipt.
c. Receipt collecting
Researcher collecting the receipt of customer purchasing to know what department is often
bought by customer in Maju Swalayan. We collects 101 receipts for getting the fixed
and best result.
3.3 Types of Data
a. Primary
Collecting data method can get directly with interviewing and collecting 100 receipts of
customers purchasing in Indotoko. Method that used by researcher is AR-MBA
(Association Rule and Market Basket Analysis). From observation we set the Indotokoas
the observation place and directly research there. From interviewing the owner, we get
the profile of the retail and allowance to collecting receipts of customers purchasing.
b. Secondary
In secondary based on International Journal of Information and Computation
Technology (Gupta and Mantora 2014) Support = 20%, Confidence = 80%
Association rules are considered useful if they satisfy both a type equation here minimum
support threshold and a minimum confidence threshold that can be set by users or domain
consultants.
3.4 Flowchart
Herewith the flowchart process based on our research in the Maju Swalayan:
Figure 1 Flowchart Process
Data Collection
Implementation of Observation
Conclusion & Recommendation
Information:
Practicum Flow:
The description of practicum explains as follows:
1. Firstly, we do identification and learning deeply about the case that we want
identify.
2. Secondly, after we make problem statement, we do technical practice which
retailer that we want observe.
3. Thirdly, if the observation place already fixed, we can start to collect the data
transaction (bill) with minimum 100 bills.
4. Fourthly, as much as >100 bills that we got, we do data cleaning process. In this
process, we eliminate the bills which bill didn’t met the requirement like error
transaction, transaction that only show same department (1 department) and if the
transaction just for 1 item.
5. After we obtained 100 bills that already slip through requirement, we can execute
the data using xl-miner.
6. The processing data we do frequently based on the practicum module.
7. From the data that already processed, we can do analysis which department that
has strong relation.
8. After we know about which department that related each other, we can give
recommendation to the retailer.
9. We also can apply the model for make a member card as the study case that our
group got.
4.1 Initial condition of the research object
In our retail which is Indotoko, there is no member card at the store. According to
our discussion with the staff of the indotoko there are several problems why they didn’t
make the member card. One of the reason why they didn’t conduct the member card is
because indotoko is the one and only retail which is in region Yogyakarta and also the
placement of the retail isn’t strategic enough which they just can reach the customer
around of the region of the retail which is Karangwaru Lor. The another reason is because
there is no previous research about the customer which is related to the card member, so
they didn’t make the member card until now.
4.2 Rapid Miner Output
4.2.1 Data of Transactions
Table 4.1 Data of Transactions
no Type
1 pucuk harum, ultra uht, nyam2
2 sprite btl, marlboro, u mild
3 enaak, djarum spr filter, lucky strike
4 diapet, good day, GG filter
5 paramex, bendera, taro, so nice
6 sania, rose tepung, mi sedap, nutrisari, lifebuoy, dji sam soe
7 campina, dunhill
8 dara mi pipih, panda kc atom
9 ultra uht, dji sam soe
10 sampoerna, chunky bar, bendera
11 sari roti, soba mi, beng2, roma, malkis, top
12 sapu lantai, lap
13 pocari, larissa
14 white koffi, djawa bakery, GG signature
15 tebs, LA Filter
16 nestle btl, aqua, the kotak, pantene
no Type
17 malkis, calpico
18 white koffi, ladaku, masako
19 indomilk, bigbabol
20 so nice, marina comp powder, nano nano
21 malkis, sari roti
22 coffemix, class mild
23 hemart, gula jawa
24 kopiko, club, plastic
25 GG surya, bigbabol
26 kiwi shoe, sikat cuci, ichi ocha, okky koko drink
tango, monde, oreo, biskitop). And from the data of the receipts we found that in
department 6 most customers buy (djawa bakery, believing tawar, and prambaru
sandwich) which ratio 3:2:2 and from department 7 we found that (roma 5, malkis 3,
nabati 3). From the result, researcher make some deals which is in every purchase of
(djawa bakery, believing tawar, and prambaru sandwich, roma , malkis, nabati) that
reach until Rp. 100.000,00 will get 1 point. If the customer can collect until 5 point,
the customer will get 5% discount of purchasing using this member card.
CHAPTER V
CONCLUSION AND RECOMMENDATION
5.1 Conclusion
Based on rapid miner result of Association Rule calculation explain about the correlation
between (product) department 6 with department 7, because association rule told by
number, support is 0.05 and the confidence is 42 (in percent). So the solution is which is
in every purchase of (djawa bakery, believing tawar, and prambaru sandwich, roma ,
malkis, nabati) that reach until Rp. 100.000,00 will get 1 point. If the customer can collect
until 5 point, the customer will get 5% discount of purchasing using this member card.
5.2 Recommendation
5.2.1 For Retailer
We think based on our observation the retailer make a new card or member card
for easier the costumer to purchasing daily needs.
5.2.2 For Next Researcher
And for next researcher, the researcher already make a member card to make easier the next researcher to make another requirement to make a new member card.
Reference
Raeder, Chawla. (no year). “Market Basket Analysis with Networks”. University of Notre
Dame. USA
Gupta, Savi and Mamtora, Roopal (2014). ”A Survey on Association Rule Mining in
Market Basket Analysis”.ITM University, Gurgaon, INDIA
Annie, Kumar (2012) “Market Basket Analysis for a Supermarket based on Frequent
Itemset Mining”. Department of Computer Science, Government Arts College
Trichy, India
Miguel, Camanho, Joao Falcao “Mining Customer Loyalty Card Programs: The
Improvement of Service Levels Enabled byInnovative Segmentation and
Promotions Design.“ Faculdade de Engenharia da Universidade do Porto, Rua Dr.