Application of clickstream analytics for decision support in marketing – a recommendation for Vietnamese small and medium-sized enterprises
Tran Thi Song Minh/ MICA 2018 Proceedings
International Conference on Marketing in the Connected Age (MICA-2018), October 6th, 2018
Danang City, Vietnam
Application of Clickstream Analytics for Decision Support in
Marketing – A Recommendation for Vietnamese Small and
Medium-Sized Enterprises
Tran Thi Song Minha*
aFaculty of Economic Informatics, National Economics University, 207 Giai Phong Street, Hanoi, Vietnam
A B S T R A C T
New wave of technology in the connected age is fundamentally changing the way enterprises compete,
operate, and do marketing. Today’s empowered customers and prospects generate voluminous and diverse
data through activity across channels. Clickstream analytics, an inter-disciplinary field associated with big
data and web mining, is a powerful tool to help enterprises gain the most insight into customer trends and
behaviors from clickstream data. Due to its high potential for analyzing customer behavior, clickstream data
analytics has received high attention by marketing researchers and practitioners alike. However, despite
overwhelming recognition of the need for clickstream data analytics, the application of clickstream data
analytics for most Vietnamese small and medium-sized enterprises remains in its infancy as there is a lack of
the proper foundation to collect and analyze clickstream data, and to integrate it into enterprise’s value chain.
Therefore, this paper provided an in-depth and manager-friendly coverage on clickstream analytics
application for decision support in marketing and at the same time, it concluded with managerial
recommendations to Vietnam government and SME owners in the matter of digital innovation strategy and
clickstream analytics application in small and medium-sized enterprises.
Keywords: Clickstream analytics; connected age; decision support; fourth industrial revolution; marketing;
small and medium-sized enterprises.
1. Introduction
1.1. Research goal and background
The overall goal of the paper is to make a strong and reasoned recommendation to Vietnamese small-medium
sized enterprises (SMEs) on clickstream analytics. It is a promising information technology application for
decision support in Marketing in the connected age. Specific objectives include providing an overview of the
application of clickstream analytics in the context of the new wave of technology in the connected age,
introducing a system approach to adopting clickstream analytics solutions in enterprise context, and making
practical recommendations for Vietnam government and SME owners.
Geoffrey Moore said, “Without big data analytics, companies are blind and deaf, wandering out onto the web
like deer on a freeway.” [1]. According to Peter Sondergaard, “Information is the oil of the 21st century, and
analytics is the combustion engine” [1]. These great quotes are incredible insights into the world of big data
analytics, the benefits of metrics, and much more. They are messages about strategic value of big data (a
* Corresponding author. E-mail address: [email protected]
160
Tran Thi Song Minh/ MICA 2018 Proceedings
valuable asset) and analytics (a specific and purpose-focused analysis). Together, big data and analytics
fundamentally change the way enterprises compete and operate.
With the various disruptive technologies revolving around increased level of connectivity, consumers are
most informed than ever before [13]. According to the third annual Digital Consumer Study [30], around 50% of
consumers have performed shopping-related tasks on their mobile phones in the past three months. Consumers
state they are using digital technology to search, browse and purchase, sometimes exclusively on one website.
When taking through a website, web user/customer will generate a large amount of the so called "clickstream
data" or "digital footprint". The question is how to gain a holistic view of users’/customers’ online journey? The
answer is clickstream analytics, an inter-disciplinary field associated with big data and web mining. One of the
main benefits of clickstream data is the derivation of customized user experience [19]. It is a powerful tool for
businesses to leverage on. Online business can gain an edge to their competitors by having insights into
customers’ preferences and behaviors [28].
1.2. Business challenges
Although insight-driven marketing is already a major area under the modern marketing communication
umbrella, clickstream analytics are much lesser known to laymen outside of the marketing industry. With the
boom of start-ups and high proportion of family-run small and medium-sized enterprises (SMEs) in Vietnam, an
in-depth and manager-friendly coverage on clickstream analytics could boost business performance and provide
significant value to the public. This is the key motivation behind this paper: “Application of clickstream
analytics for decision support in marketing in the connected age - A recommendation for Vietnamese small and
medium-sized enterprises”.
1.3. Research methodologies
In this research, we have applied a system approach to business problem solving with information technology
to ensure that the recommendation for application of clickstream analytics is reasonable, reliable and feasible.
The system approach is widely used in problem-solving in different contexts, including science and technology.
The following steps are required for this approach: Defining the problem or opportunity, developing alternative
solutions, selecting a solution, designing the solution, implementing the solution, and reviewing the solution.
Being a qualitative research-based paper, an IS-related literature review (Information System) was selected as
the most appropriate methodology for a review and synthesis of existing knowledge. This is a systematic
methodological data processing approach comprised of three major stages: 1) inputs (literature gathering and
screening), 2) processing (following Bloom’s Taxonomy), and 3) outputs (writing the literature review). The
process of "Literature gathering & screening/ Literature processing/ Report producing" is applied repeatedly to
ensure that the relevant knowledge was gathered [14].
Throughout the literature review process, we follow a consistent and coherent approach to application of
the non-routine cognitive skills: abstract reasoning (the ability to make and manipulate models – Models in
Fig. 1, Fig. 2, Fig. 3,Fig. 4), system thinking (the ability to model the components of the system, to connect
the inputs and outputs among those components into a sensible whole that reflects the structure and
dynamics of the phenomenon observed – The Clickstream analytics value chain in Fig. 3), collaboration (the
ability to develop ideas and plans with others, ability to give and receive critical feedback – In-depth
interviews with experts from SMEs for current status of clickstream analytics application), and
experimentation (the ability to create and test promising new alternatives, consistent with available resources
– A strong and reasoned recommendation to Vietnamese small-medium sized enterprises with Web presence
for clickstream analytics application).
The following four sections will provide an introduction to the new wave of technology in the connected age
and its implications for marketing (section 2), an overview of the application of clickstream analytics for
decision support in marketing, including real-world case studies and research (section 3), a system approach to
the implementation of clickstream analytics in the enterprise context (section 4); and finally some managerial
recommendations for Vietnam government and SME owners (section 5).
2. New wave of technology in the connected age and its implications for marketing
2.1. New wave of technology in the connected age
Through the ages, human beings have witnessed many phases of extraordinary technological feats,
161
Tran Thi Song Minh/ MICA 2018 Proceedings
categorizing into four phases of industrial revolutions: the first revolution with the steam engine as the greatest
invention (the original industrial revolution), the second revolution with electricity and the technique of mass
production, and the third revolution with the wide use of digital computers and the appearance of the Internet
connected the world. Now, the world is experiencing a new storm of the current and latest revolution, with an
intensive and extensive change in size, scope, and complexity never seen before. This is the fourth industrial
revolution (4IR). This is the evolution to cyber-physical systems with industrial Internet of Things (IoT) and
decentralized intelligence in manufacturing, production, logistics and the industry.
Interestingly the concept of marketing has evolved through the exact same phases (Fig. 1.). The first phase is
known as Marketing 1.0. It is a product-driven marketing where all activities were centered on the products, and
mass production with the lowest possible cost was the thrust of the marketing strategy (product-centric). The
second phase is known as Marketing 2.0, started early the third industrial revolution. It is the marketing practice
in which customers become the first priority (consumer-centric). The third phase is known as Marketing 3.0,
started to begin of the connected age, and is ongoing. It is not focused merely on products and or consumers but
human-centric, or value-driven. The fourth phase, started to begin of the fourth industrial revolution (early
2010s), is known as Marketing 4.0. This is a holistic approach to marketing that combines online and offline
exposure to consumers. It exploits the efficiency and effectiveness of interconnectivity of online data and
services while balance it with the human touch on customer engagement [13]. Fig. 1. depicts the evolving of
marketing tandem with the four industrial eras.
Fig. 1. Marketing evolving in tandem with the four industrial eras.
Source: Author's compilation from [13, 29]
In the 4IR, top innovations that have had the most significant economic impact include: Cloud Technology,
the Internet of Things, mobile Internet, Social Media, Big Data & advanced Analytics, Artificial Intelligence,
Reality/ Virtual Reality, Automation of Knowledge Work, Advanced Robotics, 3-D Printing, Blockchain
Technology, and many others [3, 16]. Cloud Computing has gone mainstream for many enterprises, IoT is
changing how both industrial and consumer-oriented businesses do business, and Big Data & Analytics have
become the core platform.
2.2. Implications of innovative technologies to marketing
Having a clearer understanding of innovative technologies in the connected age, we can return to cover their
impacts in shaping marketing industry in the past decade and the one to come. Technology becomes essential
and integral to the operations of highly effective organizations. They have not only changed the consumer
behavior, but also the marketing and business model.
Firstly, consumer behavior is evolving with the rise of the empowered customer. Technology is a powerful
tool in influencing and changing consumer behavior, and digital tools have become integrated into the customer's
shopping experience. With the booming IoT trend, people are moving most day-to-day activities to their
mobiles, tablets, and computers. This includes entertainment, socializing, and shopping. The difference is that
162
Tran Thi Song Minh/ MICA 2018 Proceedings
most online platforms nowadays contain functions that allow two-way communications, between consumers and
businesses, even on non-social media platform.
Secondly, new technologies have changed basic marketing activities. They help businesses perform vital
marketing functions in the face of the rapid changes of today’s environment. Many digital marketing techniques,
solutions, tools and models are developed for enterprises, such as marketing automation, social media marketing,
mobile marketing, inbound marketing, outbound marketing, viral marketing, influencer marketing, mobile
marketing, etc.
Thirdly, new technologies drive new business models. There are many new businesses founded on the
community in the digital age, like Google, Facebook, Uber, Airbnb, Italki, Tesla, etc., They focus on
understanding human digital relationship and harness the insights. Technology is the key to satisfied customers
and successful business. Data can speak on behalf of people about what they want and need [22].
This paper will continue to explain the process of clickstream analytics from a business manager perspective.
We will focus on application of clickstream analytics in marketing.
3. Clickstream analytics and decision support in marketing
3.1. Why clickstream analytics for marketing decision support?
Clickstream analytics refers to the management and analysis of pattern in clickstream and associated data
collected as a result of user interaction with web resources. Clickstream analytics is not technologically radical,
but it is only in the digital transformation that the discipline has been known to a wider group of researchers and
practitioners and become increasingly more popular in the online retail space. Clickstream analytics can help to
transform businesses into agile, insight-driven organizations.
Clickstream data is powerful and relevant for managers. Web users reveal what products they are interested
in, at what stage the purchase process is aborted, what questions in FAQs is clicked most. Through A/B-Testing
clickstream data can reveal which kind of product presentation leads to most purchases, how the website should
be designed.
Fig. 2. Clickstream analytics as a solution for marketing decision support.
Source: Author
In business environment, the idea to develop an IT solution/ information system is kickstarted by one or more
than one of the following factors: 1) a business problem, 2) an opportunity presents itself, and 3) a change in
environment, competition, or regulation that forces the company to adapt to survive. With this approach,
applying clickstream analytics into the value chain of the business is a given, and a smart choice for businesses
going through digital transformation (Fig. 2.). Firstly, businesses are facing problems in terms of lack of
163
Tran Thi Song Minh/ MICA 2018 Proceedings
business intelligence on customers' behavior and preference, and also with customers being more well-informed.
Secondly, many of the new disruptive technologies are bringing in a golden opportunity to businesses to compete
in the global supply chain. Thirdly, the application of innovative technologies, particularly those based on big
data and advanced analytics are key survival and success conditions to businesses. Fig. 2. describes the key
considerations in choosing clickstream analytics as a solution for marketing decision support.
In the next section we explore the value-adding process of clickstream analytics.
3.2. Clickstream analytics value chain
In terms of getting to an outcome from big data & analytics, it is exactly the so called "five element data
value chain", beginning with "data", through "information", "analytics", "insight", and finally to "action". It is
also referred as "the data/information/analytics/insights value chain" [18]. Fig. 3. depicts the value chain of the
clickstream analytics from system perspective, that helps extract web usage pattern/ insights from clickstream
data. These insights can be used for multiple purposes, e.g. to increase the customer value, improve the website,
or improve the data preprocessing. Simplified, the value chain of clickstream analytics is an operational feedback
loop that sustains itself. There are five steps/ activities, each with a unique objective/ outcome and adds value to
the value chain's output (Fig. 3.).
Fig. 3. Clickstream analytics value chain.
Source: Authors compilation from [18, 23, 25]
Step 1: Data collection. In this step, various user interactions with business’s system are recorded (raw data).
However, the load of raw data recorded is massive and messy, and the raw data themselves are not formatted to
be mined and being not be directly used. There is a need for data preparation before clickstream analytics
performance.
Step 2: Data preparation. The preprocessing in the second step encompasses collecting, merging, cleaning,
and structuring relevant data. For example, when analyzing the demand for a new brand of household product,
the manager need to be retrieved relevant search and purchase trend from web log file and transform them into
mining-able format (prepared data). The outcome of the preprocessing is information.
Step 3: Analytics performance. In this step, the prepared data will be pulled and processed with analytics
software to give relevant analytics (information that has been sorted through, using a range of algorithms and
164
Tran Thi Song Minh/ MICA 2018 Proceedings
programs, so that aggregated trends or results are visible, e.g. usage pattern, customer profiles, page profiles,
visit profiles, customer value).
Step 4: Insight presentation. It is about presentation of key insights. Once the analytics/ data-driven metrics
are obtained from the analytics software, human involvement is essential in drawing out the actual insights (key
learning from the analytics, identified in terms of meaningful business conclusions that can be drawn).
Sometimes it takes several metrics and other information combined to tell a story.
Step 5: Decision making. Management can base on the presented insights (in form of intelligent and
meaningful recommendations) and consider the available resources to take actions to correct or exploit the
results of all the work done in the data/information/analytics/insights value chain so far, that are fitting to the
business’ overall current performance, and forwarding strategy. The actions taken by the business ultimately
leads back to user interactions. This closes the value-adding cycle and the data will continuously be used to
shape future business decisions.
3.3. Application of clickstream analytics for decision support in marketing
For the uninitiated, clickstream data when captured, aggregated and processed can provide a lot of insightful
information that can benefit businesses. A disclaimer note is needed that clickstream analytics is available only if
you have any form of online presence. In association with big data analytics, clickstream analytics is a valuable
tool for businesses to be used to drive sales by optimizing every aspect of the user experience on their websites
from the first mouse click to the last. As sorted accordingly by [5], several of the top direct benefits are:
Click-path optimization. Optimization can pin-point which path has the lowest “conversion rate”, meaning
customer stop or exit before finishing the purchase. This would signal to you that you need to make changes to
certain product page paths.
Market basket analysis. Basket analysis gives you a clearer understanding of aggregate consumption
behavior. A customer can reach your product page via different path (direct, suggested item, external ads-
placement, etc.). This valuable information allows marketer to identify and focus resources on the most efficient
path to the product.
Website resource allocation. For slightly bigger businesses with multiple modules on their websites, web
development and management can be human-capital intensive. It is a challenge to allocate human resources
efficiently. Clickstream data analysis shows which part of the site is more heavily visited and need constant
renovation and maintenance.
Customer segmentation. By categorizing customers by their behaviors, marketers have a way of segmenting
the buyers and their relevant parameters (purchasing power, preferences, etc.). This gives marketers valuable and
actionable insights.
This is a non-exhaustive list of benefits that clickstream analytics bring to the table. It should be enough,
however, to justify serious considerations from any business with online presence to explore their options in
adopting it.
The above benefits listed may sound foreign to the traditional business managers, especially in the
Vietnamese market. We can take another perspective at how these data fit into businesses’ three decision making
levels.
Strategic level. Senior level decisions are high-level questions regarding what product to focus on, the
development direction, and market awareness. Seasonal traffic to enterprise’s online stores and purchase patterns
help management to decide the category of products in demand and in the right cycles. The overall data profile
of the users also let us know which features are more desired by users for future product development.
Tactical level. Middle level decisions are production and competition-based questions such as most efficient
means of productions, best pricing strategy, advertising and promotion. Analysis of product choice, which is
picked over other alternatives, provides insights for pricing analysis on price to feature preference.
Operational level. Decisions at this level are the day to day matters like tracking, documentation of
operations, dealing with new operational issues. An example is how the career social network LinkedIn packages
users’ interaction. To get customers to visit the site, LinkedIn take data when other users search up the target
customer’s name or visit his profile. LinkedIn would send an email inform you like “23 people visited your
profile this week”.
3.4. Real-world case studies and research of clickstream analytics in marketing
Due to its high potential for analyzing consumer behavior, clickstream data has received high attention by
marketing researchers and practitioners alike. In the following, some real-world case studies and promising
research areas that apply clickstream data in the marketing context are highlighted. To get a holistic view,
165
Tran Thi Song Minh/ MICA 2018 Proceedings
examples are given for each of the 4 “Ps” of the marketing mix – product, place, price, and promotion.
There are numerous examples of firms leveraging clickstream data to improve their existing products or
inform the process of developing new products, thus increasing success rates. The Financial Times (FT), for
example, analyzes its readers’ clickstreams to measure engagement with topics. Supported by this information,
FT decides which news stories to cover [2]. Similarly, streaming service Netflix is using the behavioral data
collected from its users on their website and mobile apps to guide the creation of its original content, leading to
hugely successful movies and series such as House of Cards [6].
While place used to stand for brick and mortar shops, in the online realm it signifies a company’s online store
or mobile app. Here, the usage of clickstream analytics is immediately apparent, as nowadays, websites usually
personalize what content we see based on our past clickstreams. Recently, researchers from the MIT have gone a
step further by personalizing how the content on a website is presented. Using clickstream data, the researchers
derived how each user processes information – in a detailed versus holistic manner or in visual versus verbal
manner. Based on this prediction, the presentation of website content was automatically adapted to the personal
information processing style. This system, called website morphing, has been successfully tested in the field with
a large global telecommunications company, which was able to lift sales by 20% [24].
Personalizing prices online is a delicate topic as customers usually feel treated unfairly if they have to pay a
higher price for an identical product than another customer. However, marketing managers can circumvent this
perception of unfair treatment by using individualized price promotions and coupons. Microsoft feeds users’
clickstream information into a predictive system that creates individualized e-mail advertisements. The results
are conversion rates up to 70% higher than for conventional e-mail campaigns [11].
Another aspect of marketing that has benefitted from clickstream data immensely are promotions. Based on
users’ prior behavior online, predictive models can adapt who sees an advertisement and what is displayed in
that ad. This type of behavioral targeting has been shown to double advertising effectiveness compared to
untargeted advertisements [12]. Consequently, nowadays, behavioral targeting is widely used in the industry and
offered by all major advertising agencies.
Beyond the 4 Ps, there are numerous other applications of clickstream data in the realm of marketing. As
many firms nowadays try to establish relationships with their customers, they have also turned to clickstream
analytics to make their customer relationship management efforts more effective and efficient. For example, data
from clickstreams can be used to analyze the recency, frequency, and monetary value (RFM) of a user’s past
online shop visits in order to estimate the strength of the relationship with her as well as her customer lifetime
value (CLV) [9].
In the next section, we examine the process of implementation of clickstream analytics in enterprise context.
4. Implementation of clickstream analytics solutions in enterprises
4.1. Strategies for implementation of clickstream analytics
Referring to the process chart (Fig. 3. in 3.2.) on the step-by-step process of data analytics, managers can
have a clearer look at what their options are in engaging into web analytics field. Our aim is to make it easy and
user-friendly for managers to understand what resources they would need to acquire for each option, in terms of
both human and asset resources. With reference to the criteria by Larry Chase [8], we see three main options
businesses can consider. There are also other considerations to list the costs and benefits of each path. Fig. 4.
depicts the three options for acquisition of a clickstream analytics solution.
Option 1: Complete in-house web analytics team. The company hires both the back-end IT team and the
marketing analytics team. The back-end team works with the existing web development team and their job is the
setting up and maintenance of the software analytics tools. The analysts of the marketing analytics team also
need to have a strong understanding of how to navigate the analytics tool as users to extract relevant insights.
Pre-requisite knowledge, experiences, or training courses are vital for potential staff in these roles. The company
also needs to pay subscription fees for the analytic software. The advantages of in-house option are the full
control of the team and workflow and higher efficiency from internal communication versus external
communication. The disadvantages are administrative and salary expenses from hiring additional teams, and the
need to source for talented and capable individuals to build the teams.
166
Tran Thi Song Minh/ MICA 2018 Proceedings
Fig. 4. The three options for acquisition of a clickstream analytics solution.
Source: Author
Option 2: In-house analysts, out-sourced back-end. The difference from option 1 is that you do not hire the
back-end IT team. The company would then only need to hire the marketing analytics team who would
frequently communicate with the out-sourced back-end services to maintain the information flow. This option
has most of the benefits from option 1, while the business still maintain control over the main analytics process
for efficient research and can choose a reputable company to provide back-end server services guarantee
reliability of the system and data. The disadvantages are similar disadvantages to option 1 but to a lesser extent
due to outsourcing the back-end IT staff.
Option 3: Complete out-source. This is a viable option for businesses with lean operations. The company
would be fully out-sourcing the job to a marketing analytics agency. The agency would obtain the data on behalf
of the company, perform the analysis, and make routine presentations to the management for a fixed fee. The
company would then only need a traditional marketing team and do not have to hire extra data analysts. With
complete out-source option, analytics package ensures high quality analysis and is a simple solution for SMEs
and fixed service fees are easier to manager and budget for. However, there are some disadvantages, namely
management is entirely reliant on the marketing agency to provide analysis and business has less control over
timeline and quality of analysis reports.
4.2. Evaluating and choosing the right web analytics software solution
Regardless of which option a business goes with, from in-house to fully outsourcing, careful selection of the
web analytics service suite can largely affect the success of a business' online presence. Managers need to have a
clear understanding of their own business structure, characteristics, and thus, operational needs. The right
software solution is one that allows the businesses to measure what is important to the business, fulfill the
business needs, and works with the business situation. By carefully evaluating and choosing the best fit, a
marketer can obtain reliable and comprehensive insights into user behaviors and make strategic decisions.
So how may a manager go about deciding on a web analytics tool, whether on their own or at the consult of a
marketing agency? There is a three-step process that helps a business evaluate and choose the right clickstream
analytics solution.
Step 1: Align website/ clickstream analytics application with business goal. It is about executive commitment
for web analytics. The business need to understand how the analytics tool can help you. By first identifying the
key goals the manager is looking at like customer retention, or conversion rate and making sure they fit the main
goals of the business, the business can be sure that acquiring an analytics tool is the right move.
Step 2: Define business need and situation. Business needs and situations vary so much that listing and
prioritizing them are complex task. For a clear understanding, we grouped factors to be considered into two main
categories: business-specific and technology-specific. Business-specific aspects are KPI need, demand for
integration of web analytics data with other data sources, the size and total analytics budget of the business,
among others. On the other hand, type of processing (real-time, batch, or inline?), security & privacy
requirement, needs for vendor’s support are technology-specific considerations.
Step 3: Evaluate and select the right web analytics software solution. Having a broad picture of business’
needs and situations, it is then key to rank and prioritize the top needs to be addressed by the web analytics tool.
According to [4, 10], there is a long list of factors to be considered when choosing a solution. We can classify
the relevant factors into two groups: functional and non-functional. Functional factors are reporting capability,
compatibility issues, and scalability, among others. In opposite side, types of installation mode, total cost,
167
Tran Thi Song Minh/ MICA 2018 Proceedings
vendor’s reliability, respecting legal requirements relating to data collection and data protection, capability to
process data in real-time, availability of documentation in customer’s language and support are considered as
non-functional factors.
5. Recommendation for Vietnamese SME owners
5.1. Current status of information technology application in Vietnamese SMEs
In Vietnam, there is a huge surge in the number of SMEs throughout the past three decades with a large
number of family businesses. The SMEs are essentially the main driver of the economy. According to a report
from the Vietnam Chamber of Commerce and Industry (VCCI), nearly 98% of Vietnamese enterprises are
SMEs, but they generate up to 51% of total employment and contribute 40% to the national GDP [27].
Digital technology has had a tremendous impact on Vietnamese enterprises. The question is: Are Vietnamese
SMEs ready to tap digital economic potential? The answer is they are not! [20]. There are still many problems
with the implementation of IT for this type of enterprises in terms of organization, management, and technology.
In the following, we will present key findings from research that closely examines Vietnamese SMEs.
First, there is an issue with implementing IT into the overarching business strategy. Most managers have a
less than ideal level of interest and attention on utilizing technology [15, 21]. Most SMEs do not have a plan for
using IT in business, and do not have dedicated planning teams for IT strategies. Hence, there is an observable
lack of considering IT solutions in order to achieve business goals [15].
Second, there is lack of specialized human resources. Decision making is mainly performed by enterprises’
owners or leaders, but as shown above, most of them lack the strategic vision of IT application and not fully
aware of the business value of new technologies, including social networks as well as internal collaboration
tools. Even more critical is the lack of employees and managers with the knowledge to handle the complexities
of data analytics and with the ability to simplify the output for daily use. Vietnam Democratic & Legal Magazine
provides further statistics on this: Among business owners in Vietnam, up to 56% do not have any form of
tertiary education. Only 40% are university graduates or higher or have some form of specialized training or
polytechnic background.
Third, SMEs do not fully take advantage of the opportunities the Internet provides. Vietnam has great
potential in Internet and mobile connectivity, but most businesses are not yet ready to tap on this potential.
According to VCCI data, Vietnam has a 95% Internet penetration rate for SMEs, but 60% cannot effectively
apply it to achieve business goals [27]. Vietnam ranks fifth in terms of number of people connecting online in
the Asia-Pacific region with up to 55% of Vietnamese owning smart phones and some 52 million people
connecting online [27]. With the millions of connected devices, there is huge potential in the market. However,
according to Kevin O'Kane [26], managing director of small business operations at Google Asia Pacific, most
Vietnamese SMEs are unable and not ready to tap into internet-based services’ potential, such as e-Commerce.
Fourth, there is a lack of business process standardization. Business processes of SMEs are often poor and
inefficient. There is a lack of coherence with the global standards for operations and supply chain with IT. Up to
50% SMEs do not have a functional documentation process for their standard operating procedure.
With this fundamental understanding established, our focus is of course on new technology marketing.
According to Dave Chaffey and Mark Patron [7], for businesses with online presence, the top hindering factors
to adopting data analytics for goals such as conversion rate optimization (how many people who visits your
website actually make a purchase) are lack of resources, budget, and strategy. These three key factors are similar
to the settings of local SMEs in Vietnam that we have established.
Taken together, the overarching challenges that SMEs in Vietnam face in successfully leveraging IT are a
lack of 1) top management support, 2) specialized employees, 3) familiarity with online services, and 4)
standardized processes.
As a consequence, most Vietnamese SMEs are lacking the proper foundation to collect and analyze big data
– especially clickstream data – and to integrate it into their value chain. Therefore, they are unable to reap the
various kinds of benefits of clickstream analytics as described in the previous chapters.
Managers’ perception of data analytics is low, with no clear plan for investment into these technologies. Most
marketing analytics teams in SMEs are minor and small part of the traditional marketing team. In-depth
interviews with experts from SMEs from the service sector suggest that most of their companies have not looked
at clickstream analytics as a data solution, nor do they have any concept of it. In terms of the
data/information/analytics/insights value chain, most companies stop after extracting the information from
transactional data. This limits the use of information to purely operational management level. Very few
businesses are using these data for business intelligence that can be presented to support the decision-making
168
Tran Thi Song Minh/ MICA 2018 Proceedings
process at both managerial levels: the tactical and the strategic.
The current situation of IT and data analytics shows that Vietnamese SMEs are not ready to take advantage
of the emerging opportunities of new wave of technology to improve their competitiveness and to join the global
supply chain. Therefore, we will conclude with managerial recommendations to Vietnam government and SME
owners.
5.2. Recommendations for Vietnam government and SMEs owners
The digital transformation has opened the doors to all types of organizations, including small businesses and
startups. According to Kotler [13], smaller, younger, and locally based businesses will have a chance to compete
against bigger, older, and global businesses. This is particularly applicable to Vietnam SMEs, as Vietnam is
rated as one of the countries with the highest potential to build a digital knowledge-based economy with a strong
start-up culture (#7 in the world) [26]. According to Google Customer Barometer, more than half the population
is connected to the Internet and and 43% of consumers’ purchase were first started with online exposure. By
correctly applying emerging innovative technologies, especially the Internet, and mobile connectivity platform,
SMEs can perform a successful e-Business and e-Commerce.
From a marketing perspective, small businesses can be even more competitive if they can exploit the
innovative technologies to connect with communities of customers and partners for co-creation and with
competitors for co-opetition. FPT Group president Truong Gia Binh also said that the 4IR brings along new
ideas and technologies that start-ups can exploit to break into new niche markets [17]. Due to its immense
potential, digital innovation should be the number one strategy priority for SMEs. This is essential to their
survival, success, and growth.
In recent years, the government has pushed out many SMEs development initiatives. They are degrees,
resolutions, and decisions, that improve business environment, enhance SMEs competitiveness, and strengthen
cooperation between SMEs and large businesses, promoting PPP. Among them Resolution 19/NĐ-CP (2014,
2015, 2016, and 2017) on improving business environment and national competitiveness, Decision 58/2013/QĐ-
TTg on establishment and operation of the credit guarantee funds for SMEs, and the Prime Minister Decision
844/QĐ-TTg dated 18 May 2016 on promulgating the scheme to support the national innovative startup
ecosystem through 2025 are the typical representatives.
Even with the policies, to successfully push for digital innovation strategy, there are some critical factors that
should be ensured:
From government side. We need a strong digital infrastructure by focusing on the quality of Internet
connection for both landlines and mobile connection. We need to focus on building the existing 4G network to
the standard of first world countries, and at the same time research on 5G technology. This will help us get ahead
of the IoT trend. A government provided Data Centers services will enable all SMEs to have access to a basic
level of cloud services. This allows SMEs to take advantage of these innovations without having to heavily
invest in their own system. The supportive role of the government, VCCI in legal, management, capital, and IT
infrastructure issues will enable SMEs to break into the global market.
From the SMEs side. To compete effectively in the connected age, SMEs owners should change their
perception of IT application in business and management, and improve the planning of emerging innovative
technologies. Digital innovation should be the priority with the mantra ‘think big, start small, and scale up fast’.
Everyone should work together to push for the so called Vietnamese Digital Knowledge System that the
government pushed out in Decision 677/ QĐ-TTg 18/5/2017, with the website https://itrithuc.vn/. This will be a
complete ecosystem for SMEs to utilize digital innovations.
In terms of application of clickstream analytics, there are some managerial recommendations for SMEs
owners. In deciding your course of action altogether, it is essential to establish that
(1) it is not all about the data or how much of it the enterprise has. Instead, it is about the value one extracts
from the data and how. The most important product of a clickstream analytics process is business intelligence
that analyzes patterns, trends, and relationships for marketing strategic decision making. To achieve high quality
business intelligence enterprises need a professionally implemented clickstream analytics system. Above all else,
the determining factor is how managers understand, interpret and utilize the business intelligence to support their
decision-making. Solution guideline for this is managers need to understand precisely how clickstream analytics
can provide value for their business, enterprises need to think carefully about whether they can create a
genuinely workable clickstream analytics system and how the system relates to their overall business strategy as
well as their marketing management goals;
(2) during the development process there should be an effective collaboration between the back-end IT team
and the marketing team. In information system development and use, besides technology asset and human asset,
169
Tran Thi Song Minh/ MICA 2018 Proceedings
partnering relationship for joint IT-business decision making is considered as a relationship asset. Businesses can
only have a high-quality analytics system if there is a strong collaboration between the technical staff and the
business users. Solution guideline for this is for business to involve both IT team and marketing team when
implementing a clickstream analytics system. They go hand in hand throughout the development process, from
system analysis to system design, system operation, and system maintenance;
(3) the biggest challenge is the integration and acceptance of analytical tools in the enterprise. Choosing the
right clickstream analytics software is of utmost important. Furthermore, how to integrate a clickstream analytics
system as a new application smoothly into business practice, and how to overcome user’s resistance are the two
biggest questions. Solution guideline for this is that managers should put in place a change management plan for
dealing with impacted groups within the enterprise as well as with customers and other business partners in the
enterprise’s network of value creation. It means that before embarking on clickstream analytics initiatives,
managers will need to identify carefully the organizational changes required to make them work. To overcome
user’s resistance, your enterprise should seek to understand system’s target users during the development process
as soon as possible and put in place a user’s training plan.
(4) there is no one-size-fit-all solution to pick up and implement web analytics solutions. On one hand, there
are various types of enterprises with Web presence in term of sector, size, and others. The management
requirements may be different from enterprise to enterprise. On the other hand, there are dozens of great
solutions with various functionalities. Choosing an enterprise web analytics partner is hard. Solution guideline
for this is that managers need to do a careful and reasoned analysis of selection factors and criteria from different
dimensions (organization, management, and technology) when evaluating and selecting web analytics tools.
We suggest a three-pronged summary condensed from our above discussion.
First prong: Consider the size of your business. The marketing team should be proportional to the size of the
company. A small company does not warrant a need for an in-house marketing analytics team. Small businesses
should go for full package services outsourced to marketing agency (Refer to Option 3 in 4.1). This gives a good
compromise between resources spent and reward from the analytics process.
Second prong: The second consideration is whether your core business is running sustainably with minimal
operational issues. A business should first put its focus on delivering the proposed goods and services.
Establishing stable inventory, logistics, and sales teams should be the focus of the business before attempting to
upscale revenue via webs analytics. As a result, if a business has sufficient resources and size but still need to
focus on building its core operations, a semi-outsourced solution would be the best fit (Refer to Option 2 in 4.1).
Third prong: The last consideration is based on the proportion between online and offline sales. Most SMEs
in Vietnam operate within a city area or surrounding cities with both online and brick-and-mortar stores. Their
sales are mostly driven by physical store sales. A handful provides delivery services across the country and rely
on extensive online presence. When your online sale is a key driver of total revenue, there is an abundance of
web data available, and the potential gain from utilizing analytics tool is equally immense. Therefore, it would
be ideal to follow Option 1 and build an in-house team.
For the business following either option 1 and 2, it is essential to refer to 4.2. To familiarize with the process
of selecting the best-fit analytics solution for the company.
6. Conclusion
This paper aimed at providing a holistic view of clickstream analytics application for decision support in
marketing in the connected age, targeting Vietnamese SMEs. The general introduction to new wave of
technology and its impact on marketing helps managers recognize the importance of digital transformation for
marketers to guide customers throughout their journey from awareness and ultimately to advocacy. The in-depth
and manager-friendly clickstream analytics is also covered to provide marketing managers and leaders crucial
knowledge and better understanding about key issues of application of clickstream analytics for marketing: from
theory to practice. Most importantly, managerial recommendations to Vietnam government and SMEs based on
a comprehensive examination of the current information technology status in Vietnamese SMEs have been
discussed to assist Vietnamese SMEs in implementing the digital innovation strategies and clickstream analytics
solutions.
References
[1] A4Everyone (2016), 10 great data & analytics quotes, retrieved on April 4th 2018, from
170
Tran Thi Song Minh/ MICA 2018 Proceedings
<https://blog.a4everyone.com/2016/07/26/10-great-data-analytics-quotes/>.
[2] Amazon AWS (2018), AWS Case Study: Financial Times, retrieved on May 16th 2018, from
<https://aws.amazon.com/solutions/case-studies/financial-times/?nc1=h_ls>.
[3] Blackburn, S., Freeland, M., & Gärtner, D. (2017), Digital Australia: Seizing opportunities from the Fourth
[4] Brown, G. (2013), Revealed: The secret to selecting the right web analytics solution, retrieved on March
web-analytics-solution>.
[5] Buckley, J. (2017), Clickstream Data Analysis: A Powerful Tool for Your Business, retrieved on February
10th 2018, from <https://www.qubole.com/blog/clickstream-data-analysis/>.
[6] Carr, D. (2013), Giving Viewers What They Want, retrieved on May 16th 2018, from
popularity.html>.
[7] Chaffey, D., & Patron, M. (2012), From web analytics to digital marketing optimization: Increasing the
commercial
value
of
digital
analytics,
retrieved
on
March
10th
2018,
from
<https://link.springer.com/article/10.1057/dddmp.2012.20>.
[8] Chase, L., Top 10 Things to Consider When Choosing a Web Analytics Tool, retrieved on March 10th 2018,
from <http://www.wdfm.com/webanalytics.htm>.
[9] Chen, Y., Kuo, M., Wu, S., & Tang, K. (2009). Discovering recency, frequency, and monetary (RFM)
sequential patterns from customers’ purchasing data. Electronic Commerce Research and Applications, 8
(5), 241-251.
[10] Clickinsight (2017), How to Select a Web Site Analytics Tool, retrieved on March 10th 2018, from
<https://www.clickinsight.ca/res/articles/select-web-site-analytics-tool>.
[11] Davenport, T. H., Mule, L. D., & Lucker, J. (2011). Know What Your Customers Want Before They Do.
Harvard Business Review, 89 (12), 84-92.
[12] eMarketer (2010), Behavioral Targeting Doubles Ad Effectiveness, retrieved on May 17th 2018, from
<https://www.emarketer.com/Article/Behavioral-Targeting-Doubles-Ad-Effectiveness/1007599>.
[13] Kotler, P., Kartajaya, H., & Setiawan, I. (2017). Marketing 4.0: Moving from Traditional to Digital, Wiley.
[14] Levy, Y., Ellis, T. J., (2006). A Systems Approach to Conduct an Effective Literature Review. Informing
Science Journal, Vol. 9, 2006.
[15] NDH (2013), 55% DNNVV ít chú trọng tới CNTT, retrieved on May 11th 2018, from <http://ndh.vn/55-
dnnvv-it-chu-trong-toi-cntt-20130921073645760p4c147.news>.
[16] PWC
(2017),
Technology
Trends,
retrieved
on
May
14th
2018,
from
<https://www.strategyand.pwc.com/trend/2017-technology-trends>.
[17] StartupViet (2017), Ông Trương Gia Bình: 'Startup có lợi thế trong cách mạng 4.0', retrieved on April 5th
cach-mang-4-0-3602070.html>.
[18] Stein, A. (2012), Big Data And Analytics, Other Perspectives Worth A Look – Part 4, retrieved on March
6th 2018, from <http://steinvox.com/blog/big-data-and-analytics-other-perspectives-worth-a-look-part-4/>.
[19] Svilar, M., Chakraborty, A., & Kanioura, A. (2017), Big data analytics in marketing, retrieved on April
analytics-in-marketing>.
[20] TheLEADER (2018), Diễn đàn kinh tế thế giới: "Việt Nam chưa sẵn sàng cho cách mạng công nghiệp
4.0", retrieved on May 10th 2018, from <http://theleader.vn/dien-dan-kinh-te-the-gioi-viet-nam-chua-san-
sang-cho-cach-mang-cong-nghiep-40-20180301125152893.htm>.
[21] Thoi bao kinh doanh (2016), Ứng dụng công nghệ thông tin: Doanh nghiệp Việt loay hoay tìm mô hình,
doanh-nghiep-viet-loay-hoay-tim-mo-hinh-1009131.html>.
[22] Traffic4U (2017), Building meaningful connections in a connected age, retrieved on March 9th 2018, from
<https://www.traffic4u.nl/blog/building-meaningful-connections-a-connected-age>.
[23] Turban, E., Sharda, R., & Delen, D. (2011). Decision Support and Business Intelligence Systems (9th ed.).
New Jersey: Pearson Prentice Hall.
[24] Urban, G. L., Hauser, J. R., Liberali, G., Braun, M., & Sultan, F. (2009). Morph the Web to Build Empathy,
Trust and Sales. MIT Sloan Management Review, 50 (4).
[25] Varnagar, C. R., Madhak, N. N., Kodinariya, T. M., & Rathod, J. N., (2013). Web usage mining: A review
on process, methods and techniques. Information Communication and Embedded Systems (ICICES), 2013
International Conference, 40-46. India: IEEE.
171
Tran Thi Song Minh/ MICA 2018 Proceedings
[26] VDD (2016), Đón làn sóng công nghệ số: doanh nghiệp nhỏ và vừa Việt Nam đã sẵn sàng?, retrieved on
March 4th 2018, from <https://www.vdd.com.vn/single-post/2016/06/16/60-doanh-nghi%E1%BB%87p-kh%C3%B3-
[27] Việt Nam News (2016), SMEs struggle to use digital tech, retrieved on March 6th 2018, from
tech.html#06LJ7ooHCd8f3gIM.97>.
[28] Wilmer, D. (2016), Applying Big Data Analytics to Clickstream Data, retrieved on March 5th 2018, from
<https://www.talend.com/blog/2016/10/20/applying-big-data-analytics-to-clickstream-data/>.
[29] WorldEconomicForum (2016), What is the fourth industrial revolution?, retrieved on May 8th 2018, from
<https://www.weforum.org/agenda/2016/01/what-is-the-fourth-industrial-revolution>.
[30] Yasav, S. (2015). The impact of digital technology on consumer purchase behaviour. The Journal of
Financial Perspectives: FinTech, 3(3). EY Global Financial Services Institute.
172
13 trang
10220
Free
Bạn đang xem tài liệu "Application of clickstream analytics for decision support in marketing – a recommendation for Vietnamese small and medium-sized enterprises", để tải tài liệu gốc về máy hãy click vào nút Download ở trên
File đính kèm:
- application_of_clickstream_analytics_for_decision_support_in.pdf
