Professional Documents
Culture Documents
XYZ
INDONESIA USING DESCRIPTIVE ANALITYCS
ABSTRACT
The development of information technology has caused companies in the Information
Technology Industry sector to innovate and develop products in accordance with customer needs. PT.
XYZ is a company that innovates due to government regulations and customer needs. After a change in
the methodology of the company PT. XYZ has decreased purchases in the application so it requires the
right marketing methods for customers. Marketing Program is creating marketing content by considering
customer actions (customer action) during interaction (customer interaction) so that it can recommend
the best marketing program content to meet customer needs. To determine the right marketing program
in this study using big data then analyzing the data with random forest for classification and K-means
for clustering then triangulation of expert judgment is carried out to determine the marketing program
based on data analysis. In this study, the results of important variables are money, frequency, last
transaction, province name and job type. Of the five important variables, the most optimal are obtained
with 2 clusters, namely C1 and C2. C1 is a cluster with High prospect seen from the average purchase
through the application of PT. XYZ which is greater than C2. So, it is found that C1 is B2B (Business
to Business) and C2 is B2C (Business to Customer). To get reward or marketing program content, users
will collect points and can be exchanged for marketing program content in accordance with the points
that have been determined.
2. Knowing which variables are the most This strategic method relies on the scope of
important (Important variables) for marketing and building collaboration between
segmenting customer applications PT. marketing and sales.
XYZ. III. Research Methods
3. Knowing the optimal number of customer
segmentation PT XYZ applications. 3.1 Framework for Mind
In this study, the author refers to the concept of
4. Knowing the characteristics of the targeted segmentation which goals to group customers
segment (highly prospect). with the same characteristics. Consumer
segmentation can be done based on geographic,
5. Knowing what is the marketing program demographic, psychographic, and behavioral
that is recommended in order to increase factors (Armstrong & Kotler, 2015).
sales in the application PT.XYZ.
For this study customer segmentation will use
the factors of Recency, Amount and
Transactional Frequency. Thus the framework
II. THEORY BASIS
of thought in research. After finding the most
According to (Armstrong & Kotler, 2015) important variable (Important Variable),
marketing is a process of a company in creating segmentation optimization is performed so that
value for customers and building stronger the five most influential variables on purchases
relationships with customers to get value back (Amount and Frequency) are found so that the
from customers. In the midst of business characteristics of high prospect users are known
competition today, companies must be able to and can be done program marketing.
understand consumer needs so that they can
provide more value to consumers. The more
consumers or customers obtained and the higher
the transaction, the more it will contribute to the
IV. Analysis results and Conclution
Picture 1 Framework
After the Random Forest and clustering models
are formed, the Triangulation step method,
according to Patton (1987 Moleong), there are
3.2 Research stages two strategies, namely: (1) checking the degree
of trust in the findings of research results of
several data collection techniques and (2)
In this study, the authors study at the stage of the
checking the degree of trust of some data sources
data mining process based on CRISP-DM
with same method.
(Cross-Industry Standard Process for Data
Mining) in which the sequence is not rigid and
the results of one stage will be input for the next Content Marketing program (marketing
stage. According to (Chapman, et al., 2000) that program) is made based on the results of
the process in data mining consists of 6 stages as triangulation that is making Quest Guidance so
follows: that customers know how to get points and then
redeem points into desired rewards, desirable
rewards in the form of marketing programs that
have been made
From the picture above it can be seen that the Picture 4 Marketing program Cluster 2
outer ring of CRISP-DM (The Cross-Industry
Standard Process for Data Mining) is not broken
& continuously iterated to perfect the results to REFERENCE
be achieved in order to answer business
challenges by taking lessons from previous
iterations.. Agrawal, D. (2014). Analytics based decision
making. Journal of Indian Business Research,
(2000). CRISP-DM 1.0 Step-by-step data 6(4), 332-340. doi:10.1108/JIBR-09-2014-
mining guide. SPSS Inc. Dipetik 2018 0062
Angela Stringfellow, What is Next Best Action Kotu, V., & Desphande, B. P. (2015). Predictive
Marketing? Benefits, Challenges, Best Analytics and Data Mining. Waltham, USA:
Practices, and More. NgData, 6 Novembe Elsevier Inc.
Indrawati, P. (2015). Metode Penelitian Manajemen 2017. https://www.ngdata.com/what-is-next-best-
dan Bisnis Konvergensi Teknologi action-marketing/
Komunikasi dan Informasi. (D. Sumayyah,
Penyunt.) Bandung, Jawa Barat, Indonesia: PT Brooks, C., & Thompson, C. (2017). Preditive
Refika Aditama. Modelling in Teaching & Learning. Handbook
of learning Analytics, 61-62.
Keith J. Tuckwell, M. J. (2016). Think Marketing.
Canada: Pearson Canada Inc. Collica, R. (2018). Customer Segmentation and
Clustering Using SAS Enterprise Miner. USA: SAS
Kerin, R., & Peterson, R. (2015). Pemasaran
Strategis: Kasus dan Komentar. Jakarta: Institute Inc
Indeks.
Collin Priest, Delivering Next Best Action with
Kotler, P., & Keller, K. (2018). Marketing artificial Inteligence. Datarobot, 10 April 2018.
Management. Jakarta: Indeks. https://blog.datarobot.com/delivering-next-
best-action-with-artificial-intelligence
Kotler, P., & Lane Keller, K. (2016). Marketing
Management (Global Edition ed.). USA: Grottrup Soren, W. T. (2016). Data Mining with
Pearson Education Limited. doi:ISBN 978-0- SPSS Theory, Exercises and Solutions. Berlin,
13-385646-0 Germany: Springer International Publishing
Switzerland. doi:10.1007/978-3-319-28709-6