Business analytics: new concepts

and trends
Abstract:

In the last decade, the concept of Business analytics (BA) has gone through the revolutionary
changes. It has gained a lot on the popularity and attracted immense interest both in academic
and commercial communities. Accepting this reality the main research topic of the paper is to
investigate and present the global tendencies regarding the development of business analytics
concept. We have been motivated with the huge changes in ICT environment and have studied
how they reflect on the domain of Business analytics. The main idea was to elaborate and make
distinction between crucial characteristics of the classical and new emerging concepts of
Business analytics. In addition to the technical aspects, the broader business context of all
presented concepts and phenomena are discussed. Dealing with essential elements of the BA
concept this paper points out that the history of Business analytics has been strongly influenced
by the changes and innovations in the sphere of information communications technologies
(ICT). In particular, the paper addresses two tendencies connected to the status of Business
analytics in the digital economy. The first tendency concerns the emergence of Big Data
phenomenon and its implications to both business and public environment. The other tendency
concerns the increasing business request and need for data and analytics democratization. The
paper concludes that specifically, two forms of analytics - embedded analytics and self-service
analytics, appeared as strong enablers of the process of data and analytics democratization.
Although the considerable part of the paper tackled the issue of the ICT trends and its impact on
the business analytics, the specific part of the paper presents the broader business implications
of data and analytics deployment. Our opinion is that this part provides the significant
contribution to the BA topic. We demonstrate that there is strong evidence that one of the most
important benefits that BA brings to the company is competitive advantage. Also, it is
emphasized that the other big issue in business domain concerns the forms of BA deployment
in the organization. We strongly support the opinion that successful BA deployment requires the
clear data and analytics strategy, with elements of proposed ICT innovations and BA solutions.

Key words: business analytics, information-communication technologies (ICT), big data,

democratization of analytics, open data initiative, embedded analytics, self-service analytics

JEL: M15, M21, C80

1 Introduction

The concept of Business analytics (BA) may be regarded as a result of an interaction of a few
research disciplines and software tools: quantitative data analysis - statistical analysis,
management science techniques and methods, computer science and information
communication technology (ICT) tools. Although the development of all these disciplines reflects
on the progress of the BA concept, it is not exaggerated to say that the dynamic changes in ICT
domain have the dominant influence on the BA.

In the last ten years, we have been witnesses to a very dynamic development of Business
analytics. As it was pointed out by Davenport and Harris in the period after 2006 a revolution in
Business analytics has happened (Davenport, Harris, 2017). The authors explained this
1
revolution mainly as a set of huge changes in ICT domain, which had a powerful impact on BA.
They emphasized that in the last decade the development of Business analytics and its
applications in business processes have passed a few historical phases, which they noted as
“four eras in ten years”: from the Analytics 1.0 to the Analytics 4.0.

The phase of Analytics 1.0 refers to the period when all classical analytical tools have been
used. This period spanned from the mid of the last century until 2005. The focus of this BA
concept was on the usage of descriptive statistical methods, optimization and simulation
techniques, classical reporting capabilities and the visual data presentations. In this long period,
different techniques from statistical domain and other quantitative disciplines, as well as from
management science were extensively explored for the purpose of business analytics. The final
goal of the BA usage was to provide support for business decision making, and the people
included in this process were mainly separated: data analysts on the one side and business
decision makers on the other side. The huge change happened at the mid of the previous
decade. The period from 2005 to 2012 is characterized by the emergence of the Big data
phenomenon and the shift in the analytics toward the Big Data analytics as a specific
standalone entity. This phase which refers to the rise of Big Data is marked as Analytics 2.0.
The pioneering proponents of this BA era were Google, Yahoo, Microsoft, eBay, Facebook,
LinkedIn etc., whose business models were, and still are, based on the data gathering,
manipulating and analyzing. Their main products are data products and services. In the last few
years (2012 until nowadays) much of the analytics efforts have been focused on the combined
usage of the previous two analytical capabilities, which was annotated as an Analytics 3.0.
Often, this era is characterized as a mixture of “Big and small data”: big firms are engaged in
dealings with big data and small firms have a lot of potential dealings with classical - small data
analytics. The objective of the firms in this era is not only to use BA for decision support, but
also to create new products and services. The final Analytics 4.0 era is at its beginning.
Comparing to the previous three eras this era has some specifics: while in the eras of Analytics
1.0 to Analytics 3.0 the role of the human analyst is of crucial importance, the Analytics 4.0 is
more autonomous. Despite the need for specific professionals, such as quantitative analysts
and data scientists, which are the workhorse of the first three analytics eras, the Analytics 4.0 is
more oriented to the democratization of the analytics jobs. New job positions are emerging,
such as: citizen data scientists and business translators. Artificial intelligence, machine learning
and cognitive computing are the new technologies that promise to radically change classical BA
paradigm. Nowadays, all described BA types are living and functioning in practice side by side.

The structure of the paper is as follows. After the introduction, in the second section of the paper
the classical concept of Business Analytics is discussed, whit the focus on the definition of the
term, key features and different types of business analytics. The next section elaborates the
topic of the modern tendencies regarding the information communication technologies and their
impact on BA. In particular, the Big Data phenomenon and the process of data and analytics
democratization are discussed. The fourth section deals with business implications of data
analytics usage, covering two issues: the link between BA and company’s competitive
advantage and main challenges of Business analytics deployment. The fifth section presents
the future trends of business analytics. Finally, in the sixth section concluding remarks on the
main findings are made.

2 Classical concept of Business Analytics: definition of the term and key features

Classical concept of Business analytics encompasses different tools, techniques and processes
with the purpose of analyzing business data and extracting some new patterns, trends and/or
insights that provide additional business values. One definition of BA explains it as “a process of
transforming data into actions through analysis and insights in the context of organizational
decision making and problem solving” (Evans, 2017, p.30). According to the Gartner definition
„Business analytics is comprised of solutions used to build analysis models and simulations to
create scenarios, understand realities and predict future states. Business analytics includes
2
data mining, predictive analytics, applied analytics and statistics, and is delivered as an
application suitable for a business user” (Gartner, IT Glossary, 2018).

Classical approach to business analytics has been around for a long period. It differentiates
between three distinct types of analytics: descriptive, predictive and prescriptive analytics
(Schniederjans, Schniederjans & Starkey, 2015, pp.18-20). The relationship between these
types is not competitive, but primarily complement. Each of these analytics is focused on
specific business problems and provides specific solutions. The focus of the descriptive
analytics is on the past events - it tries to give the answer to the following question: What has
happened in the past? This type of analytics is referred to as a post-festum (post-mortem)
analytics and is based mainly on the usage of descriptive statistical techniques. Today, the
majority of all business analytics still belongs to this analytics type.

Predictive analytics is oriented to the future and on the question: What will happen in the future?
This type of analytics combines the historical data with different quantitative techniques,
algorithms and rules to estimate the probability or the likelihood of the future situations and
business outcomes (Siegel, 2016). Often this analytics is referred to as a form of an advanced
analytics with the usage of the techniques, such as: trend and regression analysis, forecasting,
multivariate statistical techniques, data mining techniques - pattern matching, estimation and
prediction (Gartner IT Glossary, 2018). The final goal of the predictive analytics is to reveal
patterns in data that should serve as a base for prediction of customers’ behavior, as well as the
features of products, services and markets

Finally, prescriptive analytics is concentrated on the problems of optimization and simulation

looking for the answer on the following question: What can we do to improve the business
outcomes? This analytics goes further comparing to the predictive analytics, as it aims to
provide not only prediction, but also an adequate decision options and the relevant implications
of each decision. It includes data, mathematical and computational models and business rules,
which are derived from the existing business processes. Here, data are internal and external,
i.e. they are coming from inside organization, as well as from the outside business environment.
Also, this analytics uses hybrid data - structured and unstructured data. Further on, the used
mathematical and computational models are more complex and belong to the following scientific
disciplines: applied statistics, operational research (simulations, optimization techniques, and
game theory), machine learning, artificial intelligence, decision science theory, natural language
processing, image processing, computer vision and metaheuristics.

The final component of prescriptive analytics comprises business rules, which are relevant and
derived from the business processes. More precisely, they are extracted from the business
strategy, policies and preferences, and present the objective constraints and boundaries, best
practical examples and so on.

The main features and characteristics of the three classical analytics types could be
summarized by the following short schemes:

Descriptive analytics:
Data + Model (mainly descriptive statistics) = Description of the past (1)

Predictive analytics:
Data + Model (predictive statistics) = Prediction oriented to the future (2)

Prescriptive analytics:
Data (structured and unstructured) + Model (complex models from various domains) + Rules =
Prediction + Decision + Recommendation of the action + Advice (3)

Nowadays descriptive and predictive analytics are dominant in practice. Prescriptive analytics
does not appear in a full force, yet it is still in its infancy phase and is considered as a
3
technology of the future analytics. As Basu, A. (2013) remarked prescriptive analytics is one of
revolutionary technologies that will be the main business topic in the upcoming years. Further
on, he noted and discussed five crucial attributes of the prescriptive analytics (Basu, 2013, p. 8):

1. Hybrid data;
2. Integrated predictions & prescriptions;
3. Prescriptions & side effects;
4. Adaptive algorithms;
5. Feedback mechanism.

One of the key features of Prescriptive analytics technology is ability to process hybrid data, i.e.
combination of structured, semi-structured and unstructured data. The evolution of analytics
regarding the data type may be found in an extended text and graphic sources suggested by
Gartner (Basu, A., 2013).

Besides the traditional quantitative techniques and technologies, which are suitable for
structured data processing, hybrid data demands essentially new technologies to be processed.
That is signal and image processing, natural language processing, computer vision and more.

Apart from this basic BA classification, some additional and/or modified types of analytics can
be found in the literature. One additional BA framework is composed of four analytics types.
Besides the classical three analytics – descriptive, predictive and prescriptive analytics, it also
includes the diagnostic analytics or inquisitive analytics. These two terms are used with an
equal meaning (Schniederjans, et al 2015, Gartner IT Glossary, 2018, Vlamis, 2015, Kuipers,
2014, Cornerstone Glossary, 2018). Diagnostic analytics is closely related to the descriptive
analytics and is mainly concerned with the past events and partly with the present situation.
While the descriptive analytics is trying to give answer on the question: What happened in the
past, the diagnostic analytics is dealing with the following question: Why something happened in
the past? Usually the answer is providing by testing the business hypothesis, which includes the
usage of statistical analysis, factor analysis, and so on.

Recently one more term has been widely used. It is an advanced analytics. It has a broader
meaning then the particular BA types. While the traditional analytical tools are mainly related to
the descriptive statistical analysis of historical business data, tools for an advanced analytics are
oriented to the future and comprise predictive analytics, data mining, location or spatial data
analysis, big data analytics, to name a few of analytical tools that belong to the advanced
analytics. Gartner definition of advanced analytics points out that it includes sophisticated
quantitative methods from the domain of an advanced statistics, data mining, machine learning,
simulation and optimization to produce patterns, trends and insights that traditional approaches
are not capable to produce (Gartner BI and Analytics Report, 2016).

3 How modern ICTs reflect on business analytics?

The importance of information communication technologies as an integral part of the concept of

business analytics has already been emphasized in the previous text. Although many of ICT
innovations, directly or indirectly, reflect on the business analytics, further on in this paper only
two tendencies from ICT domain that have substantial impact on business analytics, are
discussed: the phenomenon of Big data revolution and the process of democratization of public
and business data and analytics.

3.1 The Big data phenomenon and streaming analytics

The concept of the Big data has attracted a large attention in the global ICT society and
academic community in the last decade. The rapid – almost an explosive increase in data
volume (sometimes called a tsunami of data) is without precedent in the history of data
4
gathering and storage. According to the IBM estimates the annual growth rate of data volume
was 40% in the recent few years and 90% of all data produced in the long period of the human
civilization has been produced in the last 2 years (IBM Software Group, White paper, 2015, p.5).
It is, also, an amazing evidence that more than 2,500 petabytes of data are created every day
(SAS Institute, Inc., 2017a). Further on, it is estimated by [Link], that by 2020, every
individual, who is online, will create every second almost 1,7 Mb of new data, which will result of
44 Zb (Zettabyte = 1021 byte) of data by that time.

This explosive data growth is mainly the result of the ICT tendency to the global computer
connections through the networks of people (a number of social networks), things (Internet of
Things – IoT) and processes. Also, some other technologies, such as Global Positioning
Systems and Geographical Information Systems, as well as the technology of Radio Frequency
Identification – RFID, have contributed to the enormous increase in data volume in all sphere of
work and life (Dietrich, Heller, & Yang, 2015),

There is no unique definition of this phenomenon, but the dominant definition of Big data put in
focus the three Vs: Volume, Variety and Velocity of data (Zikopoulos, Eaton, de Roos, Deutch,
& Lapis, 2012, pp. 25-36). In addition to the substantial enlargement in data volume, the other
characteristic – Variety of data reflects the various data forms and data sources from which data
comes from. Comparing to the traditional structured data (electronic data tables, spreadsheet
data editors, relational databases, etc.) the volume of semi-structured (log files, XML, e-mail,
etc.) and unstructured data (video, images, audio data, data on social networks and so on) has
been increased more rapidly. According to the IBM findings more than 80% of all data volume
which is currently generated belongs to the semi-structured and unstructured data (Basu, 2013,
p. 8). The third characteristic – Velocity refers to the dynamic nature of Big data processing.
While the previous computing paradigm was based on the processing of static data, the new
computing paradigm based on Big data technology allows the processing of dynamic data
streams (Ghosh, 2016).

The Big data phenomenon brings significant novelties and changes in the business sector and
public domain. More on this topic may be found in the paper written by Chroneos-Krasavac,
Soldic-Aleksic and Petkovic (2016). Staying on the ground of business analytics the most
serious impact of Big data on business analytics reflects on the new version of business
analytics which was born. That is a streaming analytics, i.e. analytics on „data-in-motion“. So,
nowadays we can compare the classical business analytics concept, based on the „stock“ of
data - data-at-rest (i.e. data stored in. data bases and data warehouses), with the Big data
analytics concept, which is based on the flow of data - „data-in-motion“ (data generated by
sensors, RFID and various devices - mobile, as well as the flow of data on social networks). As
the important feature of the streaming analytics is its ability to provide data processing in real
time, this analytics is also known as a real-time big data analytics.

As it was explained in the introduction part of the paper, in the periods of Analytics 2.0 and
Analytics 3.0 eras the focus was on a Big data concept. Main characteristics of Analytics 3.0.
can be summarized as follows (Davenport & Dyché, 2013, pp. 27-30):

 Different data sources and types are entering the analytical process;
 Variety of IT architectures (clustered servers of Hadoop, traditional data base) are used;
 Faster big data technologies and data processing will mature and being employed;
 Analytics models are being integrated and embedded into business applications;
 A new IT professions and roles are emerged, such as: data analysts, data scientists,
data engineers and data translators. Also, in addition to Chief Data Officer and Chief
Information Officer a new role of Chief Analytics Officers appeared;
 While the classical analytics is focused mainly on descriptive and predictive analytical
models, modern analytics tend to take advantages of prescriptive analytics.

5
The usage of big data concept and real-time analytics makes room for the substantial changes
in business processes and public services. It is estimated by SAS that 89% organizations are
looking at the Big Data as an important business opportunity, but only 3% of organizations
consider their data management systems as adequate for big data processing (SAS Institute,
Inc., 2017). For business organizations, this means that if they want to have better
performance and be in the leading position in the industry they have to face serious challenges,
otherwise they will end up losing opportunities and consequently lagging behind the
competitors (Chen, Preston & Swink, 2015). Therefore, one crucial question may be posed:
Does Big data and analytics bring better business performance? An extensive analysis of this
topic and the answer on the question can be found in the paper Big Data: the Management
Revolution, written by Mc Afee, [Link]. (2012). Authors led a team which was focused on the
investigation of the usage of data and analytics in business decision making and the
consequent business performance. They tested the following hypothesis: data-driven
companies are better performers comparing to the companies that do not rely on the data in
business running? Their conclusion was that data-driven organizations have better business
performance than other organizations. Further on they specified: „companies in the top third of
their industry in the use of data-driven decision making were, on average, 5% more productive
and 6% more profitable than their competitors“ (Mc Afee et al, 2012, p.6).

3.2 Democratization of Data and Business Analytics

The usage of Big data across various business functions is a great challenge for managers at all
business levels and layers. To take advantage of Big data requires the specific set of skills
which are incorporated into a new profession – data scientist. It is estimated that nowadays only
the United States experiences the shortage of 190.000 data scientists (Srujana, Sharma &
Amitava, 2016, p.43). Besides that, in many companies great part of useful business data are
concentrated in the ICT departments and only ICT staff and a number of executive managers
have an access to data and analytics. This is severe limitation for management to get a full
business benefits across all the levels of the enterprise. In this respect Srujana et al. (2016, p.
43) pointed out “democratization of data and analytics is the next promising frontier for business
success and data economy”. The term data democratization relates to the possibility of the end
user to access data in any digital format. It means making data and analytical tools available
and accessible to people who need them at all organizational layers. That way organizations
could improve business decision-making and make it dynamic and decentralized. There is no
unique solution regarding the level of centralization and decentralization of data and analytics
which is adequate for different organizations. The tendency of democratization of data and
analytics doesn’t have the same meaning for all businesses. It opens the space for
organizations to reconsider and recalibrate the right mix of centralization and decentralization
suitable for their own needs. The real scale of data democratization in some particular company
heavily depends on the corporate culture and management decisions. Although the topic of data
democratization is primarily the managerial issue, it is worth to mention that some technical
advances, such as advances in virtualization, data federation software and cloud data storage,
can contribute to the data sharing in organization.

The practice of democratization of non-business data and analytics has already been accepted
in many public organizations, primarily governmental institutions. Some of the well-known
examples of public data discovery platforms are Find The Data and Socrata websites. Socrata
is a public cloud-based platform that provides government data to all public users in the United
States, allowing them to use, discover and analyze data related to the cities, counties, states
and federal level (Tyler Technologies, Inc. 2018). Besides the free access to the public data,
Socrata allows consumers to use analytical tools to make insights from the data.

In this context, we refer to the Open data initiatives that have been taking ground in many
countries, regions and cities. The idea on open data initiative is very similar to the established
practice of open source software, open hardware and the similar (Kitchin, 2014). Governments
6
of many cities, states and countries are fostering the open data initiative for public data and
have established the appropriate websites (Kassen, 2017). The first open data portal was
established in May 2009, by the US government administration.

In addition, from the very beginning the idea was adopted by many international organizations
such as World Bank, United Nations, International Monetary Fund and European Commission -
EU Open Data Portal and Public Data Portal (Gray, 2014). The general framework and support
for the open data initiative in Serbia is provided by the UNDP project “Open data – Open
possibilities”. The project started in 2015 and resulted in the open data portal of Republic of
Serbia activated at the beginning of October 2017 (Open Data Portal of Republic of Serbia,
2018). The progress of the particular countries in the field of open data activities and their
achievements may be followed at the international score lists, such as: Global Open Data Index
and Open Data Barometer. Serbia is ranked at 41 in 2018 Global Open Data Index among the
94 countries. The Index calculation is based on the level of data openness of the following
indicators: Government Budget, Government Spending, National Laws, Draft Legislation,
National Statistics, Administrative Boundaries, Procurement, Air Quality, Company Register,
National Maps, Election Results, Weather Forecast, Locations, Water Quality and Land
Ownership. More details on the Index structure and the particular country’s position may be
found at Global open data Index web site (Global Open Data Index, 2018).

Open data portals have powerful impact at both macro and micro-economic levels (Berends,
Carrara, & Radu, 2017). The main economic benefits of open data are reflecting on: improved
efficiency and quality of public administration, greater transparency of public services, new
business models and innovative business services, as well as, various significant social
contributions. It is estimated by Europe Commission that open data initiatives will contribute to
the growth of EU data economy up to EUR 739 billion by 2020 that accounts for 4% of the EU
GDP. The Economic Value of Open data for EU member states may be summarized as follows:
direct market size of € 325 billion for the period 2016-2020, 100.000 jobs in Open Data in 2020
and € 1,7 billion cost savings for public sector in 2020 (Open Data Barometer, 2016, pp.3-4).
The other study revealed the open data potential in the following areas – Consumer finance,
Consumer products, Health care, Oil and gas, Education, Electricity and Transportation. It
reveals that the economic potential of open data in these domains is approximately 3 - 5 trillion
$ per year (Manyika, Chui, Farrell, Van Kuiken, Groves, & Almasi Doshi, 2013, p. 9). Apart from
direct economic benefits it is believed that open data portals should encourage the creative
business initiatives and foster knowledge economy development. Also, it is important to note
that Open data entrepreneurial ecosystem could not exist without support, i.e. it must be
permanently encouraged and cultivated (Walker, Simperl, 2018).

As a valuable enablers of data and analytics democratization an open source platforms (such as
Hadoop) and programming languages (for example R) have appeared. On the other side, on the
BA’s market there are lot of BA producers that provide distributed models for analytics, as well
as easy-to-use and flexible interfaces to users’ legacy systems. More on this topic can be found
in Gartner annual report on Magic Quadrant for business Intelligence and Analytics Platform
(Gartner BI and Analytics Report, 2017). Additionally, it is evident that serious steps have been
made to integrate the embedded analytics into business applications. Therefore, the embedded
analytics is considered to be a future of Business Intelligence (Eckerson, 2016, pp. 9-13).

The wave of democratization of data and business analytics has brought some new roles in the
business world. Here we point out on citizen data scientists and business translators. Citizen
data scientist is neither an ICT specialist, nor the expert from the field of statistics or analytics.
Rather it is a person who is skillful in data handling and manipulation, as well as in creating
models with analytical software. Thanks to a new breed of BA software solutions the citizen data
scientists are taking more powerful and growing role in completing BA tasks without requiring a
data science degree (Gartner BI and Analytics Report, 2016). Business translators are persons
from the operation business levels in organization who possess business knowledge and the
general understanding of ICT. Their task is to make good connection between ICT professionals
7
and business people. According to the Forrester report acceleration in the trend towards
democratization of data usage and data analysis will continue. Their study shows that in 2015
almost 51% of business decision makers who used data and analysis in decision-making
process have no problem in accessing data and their analysis without the help of ICT staff. The
prediction is that this trend will continue and very soon reach the figure of 70% (Mc Cormick,
2016, p. 2).

3.3 Embedded Analytics

The wave of democratization of BA concept brings new forms of analytics inclusion in business
applications: embedded analytics and self-service analytics. As it was earlier pointed out, the
main purpose of BA is to elicit some knowledge and insight from data. In this regard the concept
of BA may be considered as an integral part of the wider concept of Business intelligence (BI).
The subject relating to the relationship between Business intelligence and Business analytics is
examined by Heinze, J. (Heinze, 2016).

While the traditional BI concept was mainly report and dashboard oriented with possibilities for
reactive actions, embedded analytics brings the new perspective: not only reports, queries and
dashboards, but also, the analytical models and outputs are embedded into applications. This
way the concept of BI is being transformed from reactive to proactive set of business tools.

The history of development of Business Intelligence tools, techniques and concept can be
observed as an evolution process rather than the revolution one. It is presented at the Figure 1
(Eckerson, 2016).

Figure 1 The evolution of BI – and its future (Source: Eckerson, 2016, p. 4.)

It is evident that BI concept has been developing in two directions: the first is concentrated on
the reporting functionalities, while the other focuses on the analysis. These two functionalities
have been aimed to the different classes of users: the first report-centric BI functionality has
been oriented toward the casual users, while the other one – analysis-centric to data
professionals and more powerful users. The final phase of BI evolution is mixture of these two
capabilities. It generates an embedded reports and operationalized models. These two outputs

8
are supposed to be actionable, i.e. to produce a prompt and automatic action in business
environment.

According to the Gartner definition of embedded analytics it concerns to the usage of reporting
and analytic programming features in transactional business applications. These features can
be stored as functionalities inside the application or outside it (residing in some other system).
However, the main point is that they must be easily accessible for users, without need to
change system (Gartner IT Glossary, 2018).

What kind of tools and functionalities organizations embed into their applications? Generally, it
could be a set of the following functionalities:

 Charts and components (static report and/or report’s parts, tables, queries, maps etc.);
 Dynamic reports and dashboards, that provide an interactivity for the users;
 Transaction execution (apart from analytical capabilities some BI tools can provide also
the transaction execution);
 Predictive analytics, that encompasses various advanced analytical models and
techniques;
 Self-service analytics – aimed to the power users for data extraction, classification,
visualization and making specific reports and/or dashboards.

Having in mind the technological characteristics of the embedded BI it may be concluded that it
is not a new phenomenon. On contrary, it has been around for a long period. The following table
presents the main features of this technology in the last three decades.

Table 1 : Embedded BI – time perspective

Period Platform Data Source User features

1990s Desktop Files, RDBMS Static Reports
Dynamic-interactive Reports,
2000s Web OLAP, XML
Dashboards, OLAP
Big data, Hadoop, NoSQL,
2010s Cloud Cloud apps, streams, Predictive and self-service analytics
server logs

Source: Authors modified the illustration presented in Eckerson Group Report, (Eckerson, 2016, p.10)

Generally speaking the embedded analytics use different analytic features that are incorporated
in transactional business applications. These analytical features may be part of application or be
stored outside of applications, but it is important that user can approach them relatively easily,
without switching different systems.

3.4 Self-service analytics

Embedded analytics paved the way for the self-service analytics usage. Namely, a new set of
tools has emerged that help ordinary business users to prepare descriptive analysis and visual
exploration of the data at their disposal. This idea is not new. It demonstrates that data analysis
is not the job to be exclusively done by business analysts, but by an entire company work force.
The proponents of the idea believe that data analysis jobs should be removed from the ICT
departments to the business operation levels. So, it is expected business professionals will be
enabled to perform at least simply queries, data exploration and visualization and report
generating. In line with this idea is the emergence of flexible and easy-to-use software tools and
functionalities to handle and visualize data. Therefore, more intuitive, business centric analytical
tools with flexible interfaces are taking the ground and promising to be available to the ordinary
business users. That is the process of democratization of analytics.
9
As the new trend of the rise of self-service analytics (SSA) is getting momentum, the greatest
challenge for the organizations is the existence of the adequate data governance policy which
could provide high-quality data that is easily accessible (Brunet, 2018). Some studies show that
55% of organizations with a very high level of adoption of self-service analytics have
implemented serious data governance strategy and rules (Krensky, 2015, p.3). Only this way
the self-service analytics can provide full potential for casual and power business users.

4 Business implications of data analytics usage

In the previous sections, the classical concept of BA is introduced and then the modern ICT
trends and innovations, reflecting on the BA, are discussed. This section is expanding the scope
of discussion in the broader business context. It tackles the issue of some business implications
of Data analytics (DA) deployment.

4.1 The link between BA and company’s competitive advantage

Many research studies dealing with the concept of BA, as well as the business practice, have
shown that the usage of BA is widespread across almost all company business functions.
However, the most dominant influence of BA, in particular Advanced and Predictive Analytics
(APA), is recorded in the sphere of marketing, sales, executive management, information
technology and finance (Dresner Advisory Services, 2015). Among many statistical and other
quantitative techniques, the most popular is the usage of the following techniques: regression
techniques (linear, nonlinear and logistic regression), different clustering and classification
techniques, descriptive statistical procedures, principal component analysis, support vector
machine – SVM models, neural networks – NN models, decision trees models (C4.5, CHAID,
random forests), language processing and sentiment analysis (Dresner Advisory Services,
2015). The usage of all these techniques provides the companies with the substantial potential
and possibilities for business success. According to some marketing studies, the main benefits
that BA brings to the companies are: competitive advantage, which is seen as the most
important benefit for business users (noted by 57% respondents), new revenue opportunities,
increased profitability, improved customer service and cost savings (Ventana Research Report,
2015, p.7).

The impact of BA on company’s business running, in particular on the competitive advantage,

was first presented in the well-known paper Competing on Analytics, prepared and written by
Davenport, Cohen and Jacobson (2005), more than one decade ago. The paper was based on
the research conducted with the purpose to investigate the relationship between the companies’
usage of sophisticated analytics and their business performance. The research had been
sponsored by the great ICT companies - SAS and Intel. The authors of the paper shed the light
on the new specificity regarding the competition – competition based on the extensive use of
analytics, data and fact-based decision making. They argued that this type of competition is not
something new, yet it has already existed for the period of few decades. This was particularly
true for the finance industry, mainly financial investment and trading business, where
organizations have been competing using analytics for a long period. This practice has been
adopted by the other industries, such as consumer profiling and finance, retailing,
telecommunications, marketing, tourism, supply chain management, insurance industry,
entertainment and so on. Furthermore, analytics based competition has been spreading within
organizations – from department units to enterprise-wide level. Describing the companies, which
are analytically oriented in conducting business processes, the authors of the research
emphasized the most prominent attributes of these companies:

 The strong support of analytical-based decision making by senior executives;

 An extensive use of statistical methods, predictive analytics, optimization, simulation and
other complex analytical methods;
10
 Widespread usage of the analytics not only in one business unit, but in multiple and cross-
section units and business functions.

Based on the conducted survey the authors remarked that analytical competition is not simply
„yes“ or „no“ attribute. On the contrary, analytically oriented firms may pass through several
phases in their analytical business orientation. The authors defined these phases as follows
(Davenport et al, 2005, p. 4):

 Phase 1: Major Barriers;

 Phase 2: Local Activity;
 Phase 3:Vision Not Yet Realized;
 Phase 4: Almost There;
 Phase 5: Analytical Competitors.

One third of the organizations in the survey were positioned in the phase 5, only 7% in phase 1
and about 20% in each other phases. The final conclusion derived from this research indicated
that opportunities for analytical competition may be found in every industry. Consequently, the
individual firm that did not recognize and embrace these opportunities is in risky position to lag
behind the competitors regarding the crucial business performance.

4.2 Main challenges of Business analytics deployment

As it was pointed out in the previous text the flood of data, generated by various devices, is
transforming the world around us and has a potential to change business data landscape in
general. In that context a crucial practical issue for organization is the following: How to
organize data flow and make it accessible to the potential users who will deliver business value?
The critical part of the answer on the previous question is the implementation of the adequate
data infrastructure and efficient data analytics strategy. Data analytics could take a role of a
paramount business differentiator that brings the new insight-driven business models. These
models are supposed to be data-focused models. In that respect, we recall on the famous
online companies, such as, Google, Facebook, Yahoo, LinkedIn, Microsoft, etc., which the most
are striking examples of organizations whose business processes are completely data driven. It
could be observed that, in general, organizations empowered with this kind of business models
have substantial potential to outperform the organizations without data analytics experience in
business running. How much the organization can take advantage of this potential depends of
its capacity to go through the process of digital transformation. In short, the process of digital
transformation leads to the data and analytics driven organization which is making smarter
decisions, has potential to act not only reactively, but rather more proactively, creates more
accurate business forecast, and becomes more predictable and profitable.

How to set out on the road to become data driven organization? The first big issue for
organization is how it treats the available data. Many organizations are treating data and
analytics as passive and static business resources. On contrary, data and analytics must be
treated as dynamic resources, capable to deliver real-time information to all line-of-business
(LOB) managers and users. The next big issue for organization is: what is its current position on
the “business analytics journey” (Shashikiran, 2018). Does it belong to the stage of descriptive
analytics, or to the stages of more matured analytics - predictive, prescriptive or more
sophisticated - an advanced analytics? This starting point reflects on the further activities and
decisions which should be resolved, such as: the choice of the data and analytics infrastructure,
creation and deployment of the renewed (or completely new innovative) data-centric services
and products, expansion of the good practice across the organization, recruiting, training and
allocation of the human resources for data analytics assignments, etc.

To deal with the previous issues organization needs to make a well-defined data strategy,
concerning the data and information quality, data security and governance, data integration and
11
preparation, data exploration, etc. (DalleMule & Davenport, 2017), as well as an analytics
strategy. Both of them should be in line with the business strategy, or more precisely, should be
an integral part of business strategy.

The road that leads to data driven organization is usually full of bumps – technical and non-
technical. The technical obstacles are connected to the data storage, data quality (Batini &
Scannapieco, 2016), the integration of data from different data sources, hardware and software
platforms, human resources, etc. On the other hand, one of crucial non-technical impediments
is the lack of data-driven mindset in the organization. In that respect, the most serious efforts
should be put on the change of the cultural mindset in the organization (Edmonds, 2014). This
change is supposed to be directed toward the more open data-driven patterns of decision
making in all sphere of business running, which opens the door for the process of data and
analytics democratization. The most common enablers of the process of data and analytics
democratization are open data platforms, open source software platforms, as well as data
virtualization and data federation software. Yet, there are substantial challenges to make this
concept a reality. Namely, the process of data and analytics democratization requires a strong
commitment of managers at all business levels. In particular, the top business management is
supposed to provide the insurance that the data and analytics strategy would be put in practice.
The final goal of this strategy may be the creation of the collective intelligence in organization,
which means the free information sharing throughout the organization.

5 The future trends of business analytics

Nevertheless at which scope the business organizations accept the BA concept, the majority of
them are in position to deal with the modern ICT innovations and consequently prepare for the
future. What they can expect on the field of business analytics in the future? The future of
business analytics is shaped by the following trends:

 The trend of exploring Big data structures will continue and will be the focal point for
both academic research and commercial application. The whole world is facing a
tsunami of data coming from different sources, usually at extremely high speed and in
various formats;
 An algorithmic economy is getting on the power - looking for an adequate response to
Big data phenomenon, organizations and other business subjects are forced to turn to
the new algorithmic solutions derived from the field of machine learning (ML), data
mining (DM), cognitive computing, natural language processing (NLP) and deep
learning (DL). To be successful in these endeavors organizations must be flexible in
integrating different data and analytics architecture in seamlessly connected
infrastructure;
 The widespread usage of mobile devices foster the explosive growth of mobile
applications – so, an application economy enters the stable and more matured period of
development;
 Visual analytics is on the rise. It becomes a lingua franca for various data professionals
and ordinary business users. Therefore, many ICT vendors are preparing products and
tools for data visualization;
 Cloud analytics has taken the ground and would be an acceptable solution for the
growing number of organizations;
 On the hardware side, the inclusion of Graphics processing units (GPUs) is getting on
momentum, replacing CPUs that have been used for decades (Weldon, 2017);
 In addition, one of crucial elements of the future of data analytics is a focus on the
speed at which organizations could generate business value from the data sources.
Real-time analytics is in demand for fast moving data – data-in-motion, as well for data-
at-rest, to deliver business insights and responses;
 Streaming analytics (data manipulation and analysis on the fly), with appropriate tools,
is entering the BA market space;
12
  Finally, the big challenge in the next years will be the usage of Internet-of-things (IoT)
technology, which will have an essential impact on the data gathering, storage and
processing, as well as, on the data analysis at all organizational levels.

6 Conclusion

In the last decade, the concept of Business analytics (BA) has gained a lot on the popularity and
attracted immense interest both in academic and professional - commercial communities. It
includes different tools, techniques and processes which are focusing on analyzing business
data and extracting some new patterns, trends and/or insights, that provide additional business
values.

Dealing with essential elements of the BA concept this paper points out that the history of
Business analytics has been strongly influenced by the changes and innovations in the sphere
of information communications technologies. The last decade was the most turbulent period,
which has brought a set of huge changes in ICT domain with a powerful impact on BA. In
particular, the paper addresses two tendencies connected to the status of Business analytics in
the digital economy. The first tendency concerns the emergence of Big Data phenomenon and
its implications to both business and public environment. The other tendency concerns the
increasing business request and need for data and analytics democratization. Data and
analytics democratization can be defined as data and analytics resources at the disposal to the
employees in accordance with their needs, at all organizational levels and layers. The paper
concludes that specifically, two forms of analytics - embedded analytics and self-service
analytics, appeared as strong enablers of the process of data and analytics democratization.
Today this process is more or less in force.

Although the considerable part of the paper tackled the issue of the ICT trends and its impact on
the business analytics, also the broader business implications of data and analytics deployment
has been discussed. Regarding all the above, the connection between BA and company’s
competitive advantage was firstly considered. There is strong evidence that one of the most
important benefits that BA brings to the company is competitive advantage. The other big issue
concerns the forms of BA deployment in the organization. We strongly support the opinion that
successful BA deployment requires the clear data and analytics strategy, with elements of
proposed ICT innovations and BA solutions. An integral part of this strategy should be a plan (a
roadmap or a guideline) which could lead organization through the process of digital
transformation. The final goal of digital transformation is the creation of data driven organization
which is characterized with the widespread usage of data and analytics for making smarter
decisions. We emphasized that one of the critical impediment in the process of digital
transformation may be the lack of data-driven mindset and culture in the organization. To
overcome this problem organization needs a strong support of managers at all business levels.
In particular, the top business management is supposed to be consequential in creating and
maintaining the atmosphere of BA acceptance throughout the organization.

Finally, we have presented the future BA trends. These trends reveal the essence of the new
developing era of BA - Analytics 4.0, which is now at its beginning. Yet, the main algorithms and
tools on which the Analytics 4.0 is based on, such as algorithms of machine learning, deep
learning and cognitive computing, are far from the infancy stage. They are already in use and
will gain momentum in the next period.

Acknowledgements

Authors gratefully acknowledge the financial support from the Ministry of Education, Science
and Technology of the Republic of Serbia, Grant No. 179005.

13
References

[1] Basu, A. (2013). Executive edge: Five pillars of prescriptive analytics success.
Analytics, March/April 2013, 8-13, INFORMS, Retrieved from:
[Link]
success/
[2] Batini, C., Scannapieco, M. (2016). Data and Information Quality - Dimensions,
Principles and Techniques, Springer International Publishing Switzerland, DOI
10.1007/978-3-319-24106-7
[3] Berends, J., Carrara, W., and Radu, C. (2017). Analytical Report n9: The Economic
Benefits of Open Data. European Data Portal. Retrieved from:
[Link]
_report_n9_economic_benefits_of_open_data.pdf
[4] Brunet, P. (2018). The rise of self-service analytics. Analytics, January/February 2018,
34-40. INFORMS, Retrieved from: [Link]
analytics/
[5] Chen, D.Q., Preston, D. S. & Swink, M. (2015). How the Use of Big Data Analytics
Affects Value Creation in Supply Chain Management. Journal of Management
Information Systems, 32(4), 4-39, DOI: 10.1080/07421222.2015.1138364
[6] Chroneos Krasavac, B., Soldić-Aleksić, J., Petković, G. (2016). The Big Data
Phenomenon – the Business and Public Impact. Industrija, 44(2), 117-144, DOI:
10.5937/industrija44-10036
[7] Cornerstone Glossary (2018). Diagnostic Analytics. Retrieved from:
[Link]
[8] DalleMule,L., Davenport, T. H. (2017). What’s Your Data Strategy? Harvard Business
Review, May-June 2017, Retrieved from: [Link]
strategy?referral=03759&cm_vc=rr_item_page.bottom
[9] Davenport, T., Cohen, D., and Jacobson, A. (2005). Competing on Analytics. Babson
Executive Education - Working Knowledge Research Center. Retrieved from:
[Link]
[10] Davenport, H.T., Dyché, J. (2013). Big Data in Big Companies: Executive Summary.
International Institute for Analytics, Sponsored by SAS Institute Inc. Retrieved from:
[Link] research1/big-data-big-companies-
[Link]
[11] Davenport, H. T., Harris, J., and Abney, D. (2017). Competing on Analytics: The New
Science of Winning. Brighton: Harvard Business Review Press
[12] Dietrich, D. Heller, B. Yang, B. (2015). Data Science and Big Data Analytics:
Discovering, Analysing, Visualizing and Presenting Data. EMC Education Service.
Indianapolis, Indiana: John Wiley & Sons, Inc.
[13] Dresner Advisory Services. (2015). Advanced and Predictive Analytics (Market Study
2015). Wisdom of Crowds® Series, Licensed to Dell, Sponsored by Dell, LLC.
Retrieved from: [Link]
[14] Eckerson, W. W. (2016). Embedded Analytics: The Evolution of BI – and its Future.
Eckerson Group. Retrieved from:
[Link]
[Link]
[15] Edmonds, S.C. (2014). The Culture Engine: A Framework for Driving Results, Inspiring
Your Employees, and Transforming Your Workplace. New Jersey: John Wiley & Sons,
Inc.
[16] Evans, R. J. (2017), Business Analytics - Methods, Models and Decisions. 2nd edition.
London, England: Pearson Education Limited
[17] Gartner BI and Analytics Report, (2016), Magic Quadrant for Advanced Analytics
Platforms, February 2016
[18] Gartner BI and Analytics Report (2017). Magic Quadrant for Business Intelligence and
Analytics Platform. Retrieved from: [Link]
magic-quadrant-2017/
14
[19] Gartner, IT Glossary, 2018, Retrieved from:
[Link]
[20] Ghosh, J. (2016). Big Data Analytics: A Field of Opportunities for Information Systems
and Technology Researchers. Journal of Global Information Technology Management,
19(4), 217-222, DOI: 10. 1080/1097198X.2016.1249667
[21] Gray, J., (2014). Towards a Genealogy of Open Data. Paper presented at the General
Conference of the European Consortium for Political Research in Glasgow, 3-6th
September 2014. Available at Social Science Research Network (SSRN). DOI:
10.2139/ssrn.2605828.
[22] Heinze, J. (2016). Business Intelligence vs. Business Analytics: What’s The Difference?
July 2016. Retrieved from: [Link]
business-
analytics
[23] IBM Software (2015). Understanding big data so you can act with confidence. White
paper ITPP in association with IBM,January 30, 2015, Retrieved from:
[Link]
[24] INTEL, (2017), Five Steps to Delivering the Data-Driven Business, white paper,
Enterprise Analytics Data Center,
[25] Kassen, M. (2017). Understanding transparency of government from a Nordic
perspective: open government and open data movement as a multidimensional
collaborative phenomenon in Sweden. Journal of Global Information Technology
Management, 20(4), .236-275, DOI: 10.1080/1097198X.2017.1388696
[26] Kitchin, Rob (2014). The Data Revolution - Big Data, Open Data, Infractures & Their
Consequences. London: SAGE Publications Ltd, DOI: 10.4135/9781473909472
[27] Krensky, P. (2015). The Landscape of Self-Service Analytics. Aberdeen Group.
Retrieved from: [Link]
[28] Kuipers, E. (2014). Inquisitive Analytics - Mapad: the new way of thinking. Retrieved
from: [Link]
[29] Manyika, J., Chui, M., Farrell, D., Van Kuiken, S., Groves, P., Almasi Doshi, E. (2013).
Open data: Unlocking innovation and performance with liquid information. McKinsey
Global Institute. Retrieved from:
[Link]
unlocking-innovation-and-performance-with-liquid-information
[30] Mc Afee, A., and Brynjolfsson, E. (2012). Big Data: the Management Revolution.
Harvard Business Review, October 2012. Retrieved from: [Link]
data-the-management-revolution
[31] Mc Cormick, J. (2016). Predictions 2017: Artificial Intelligence Will Drive the Insights
Revolution. Forrester Research. Retrieved from:
[Link]
artificial-intelligence-will-grow-300-in-2017/#1efe27da5509
[32] Open Data Barometer (2016) Retrieved from:
[Link]
[33] Global Open Data Index (2017) Place Overview – table view. Retrieved from:
[Link]
[34] Open Data Portal of Republic of Serbia (2018). Retrieved from: [Link]
[35] European Open Data Portal (2018) [Link]
[36] SAS Institute, Inc. (2017). Becoming a data-driven organization, white paper, IFL
Science
[37] SAS Institute, Inc. (2017). Roles and Responsibilities in Data Management for Big
Data, presented by Information Management
[38] Schniederjans, M.J., Schniederjans, D.G. Starkey, C.M (2015). Business Analytics -
Principles, Concepts, and Applications with SAS: What, Why, and How. New Jersey,
USA: Pearson Education, Inc.
[39] Siegel, E. (2016). Predictive Analytics - The power to predict who will click, buy, lie, or
die, Hoboken, New Jersey: John Wiley & Sons, Inc.

15
[40] Shashikiran, P.B., (2018). The Amazing Analytics Journey. Analytics, Jul/August 2018,
50-54, [Link]/10.1287/LYTX.2018.04.12
[41] Sigma’s DIPPTM framework (2017). Retrieved from:
[Link] [Link]
[42] Srujana, H.M., Sanjay S. Sharma and Amitava, D. (2016). Democratization of analytics:
New frontier of data economy. Analytics, March/April 2016, 42-50, Retrieved from:
[Link]
economy/
[43] Tyler technologies, Inc. (2018). The Future of Connected Governmen., Retrieved from:
[Link]
[44] Ventana Research (2015). Next-generation Predictive Analytics: Using Forward-
Looking Insights to gain Competitive advantage. Research Report Executive Summary.
Retrieved from: [Link]
Pred.-[Link]
[45] Vlamis, T. (2015). The Four Realms of Analytics. Retrieved from:
[Link]
[46] Walker, J. & Simperl, E. (2018). Analytical Report 10: Open Data and Entrepreneurship.
European Data Portal, February 2018, Retrieved from:
[Link]
analytical_report_10_open_data_and_entrepreneurship.pdf
[47] Weldon, D. (2017). 6 trends shaping the future of data analytics. Retrieved from:
[Link]
shaping-the-future-of-data-analytics#slide-1
[48] Zikopoulos, P.C., Eaton, C., de Roos, D., Deutch, T., Lapis, G., (2012). Understanding
Big Data, Analytics for Enterprise Class Hadoop and Streaming Data. New York:
McGraw-Hill

16