Challenges in Decoding Consumer Behavior with Data Science

Decoding the ever-evolving consumer behavior is one of the biggest challenges faced by marketers around the world. The future of consumer behavior research is put into question by the advances in data science. Today, when consumers are all the time exposed to new technologies, trends such as facial recognition, artificial intelligence, and voice technology did not advance as rapidly as predicted, marketing intelligence gained a significant share of the spotlight. This paper gives an overview of possible ways to anticipate consumer data intelligence development from the perspectives of a robust data set and deep artificial intelligence expertise for better understanding, modeling, and predicting consumer behavior. Showing that marketing cannot happen in isolation in the era of digital overexposure, it requires a deeper understanding of consumer behavior. Data scientists, analysts, and marketers around the world have to work together to increase consumer loyalty, grow revenue, and improve the predictiveness of their models and effectiveness of their marketing spend. Efficiently integrating consumer behavior data into marketing strategies can help companies improve their approach towards attracting and winning the diverse and dynamic consumer segments and retaining them. This synthesis of current research will be helpful to both researchers and practitioners that work on the use of data science to understand and predict consumer behavior, as well as those making long-range planning marketing decisions.


Introduction
Consumption continues to change with technological advancements and shifts in consumers' values and goals (Malter et al., 2020).The consumer is ultimately the key determinant of the success of an organization.The need for an in-depth and objective understanding of the consumers, therefore, in terms of what runs in their minds and hearts for the way they behave and act when they go about making complex purchase decisions cannot be over-emphasized (Moses & Clark., 2020;Sankaran, 2019).Increasing global digitalization brings huge and ever-growing amounts of data (Skiera, 2016).Decoding the ever-evolving consumer behavior is one of the biggest challenges faced by marketers around the world.The adoption of contemporary methods in consumer data analytics is slow and many businesses fail to understand their consumers as well as they want.The future of consumer behavior research is put into question by the advances in data science.This paper gives an overview of possible ways to anticipate consumer data intelligence development from the perspectives of a robust data set and deep artificial intelligence expertise for better understanding, modeling, and predicting consumer behavior.

Background
In this section, an overview of Consumer Behavior Research and Data Science is given and definitions used for the analysis in this review are introduced.

Consumer Behavior Research (CBR)
In recent years, technological changes have significantly influenced the nature of consumption as the customer journey has transitioned to include more interaction on digital platforms that complements interaction in physical stores.Besides, this shift allows us to collect more data at different stages of the customer journey, which further allows us to analyze behavior in ways that were not previously available (Malter et al., 2020;Tong et al., 2020).Not only have technological advancements changed the nature of consumption but they have also significantly influenced the methods used in consumer research by adding both new sources of data and improved analytical tools (Ding et al., 2020;Ohme et al., 2020).The adoption of contemporary methods in consumer data analytics is slow and many businesses fail to understand their consumers as well as they want.The future of CBR is put into question by the advances in data science.

Data science
Data science (DS) combines multiple fields including statistics, scientific methods, and data analysis to extract value from data, being is an umbrella term used for multiple industries, such as data analytics, big data, marketing intelligence, data mining, machine learning and artificial intelligence, and predictive analytics, and is being increasingly adopted to analyze and predict consumer behavior (Cognetik, 2020;Sankaran, 2019).
Where: Big Data.Big data is a collection of unstructured data that has very large volume, comes from variety of sources like web, business organizations etc. in different formats and comes to us with a great velocity which makes processing complex and tedious using traditional database management tools.The major demanding issues in big data processing include storage, search, distribution, transfer, analysis and visualization (Khade, 2016).In consumer behavior marketing, big data is used to analyze data points of a customer's journey from exploration to sale, powering marketers with tools and knowledge to make more informed decisions (Margalit, 2020;Saheb & Saheb, 2020).
Data Mining.Data mining and analytics have played an important role in knowledge discovery and decision making/supports in the process industry over the past several decades (Ge et al., 2017).Data mining is defined as a process used to extract usable data from a larger set of any raw data.It implies analyzing data patterns in large batches of data using one or more software.
Predictive Analytics (PA).The most widely used data set in consumer behavior, and the one we'll be referring mostly to in this article, is PA.Predictive behavior modeling can reveal many insights to support marketing strategy.
Machine Learning (ML).ML is used in DS to make predictions and also to discover patterns in the data, in situations where necessary the machine to learn from the big amounts of data, and then apply that knowledge to new pieces of data that streams into the system (Liu et al., 2018;Zolghadri & Couffin, 2018).

Artificial Intelligence (AI)
. Some researchers propose that AI is the field of study that describes the capability of ML just like humans and the ability and refers to programs, algorithms, systems and machines that demonstrate intelligence (Khanna et al., 2020;Shankar, 2018).This paper follows another way to describe AI, that depends not on its underlying technology but rather its marketing and business applications, such as automating business processes, gaining insights from data, or engaging consumers and employees (Davenport et al., 2020).
Marketing Intelligence.The term marketing intelligence refers to developing insights obtained from data for use in marketing decision-making (Eggert & Alberts, 2020).Data mining techniques can help to accomplish such a goal by extracting or detecting patterns or forecasting consumer behavior from large databases.Marketing intelligence has long been an implicit office standard, irrespective of specific big data solutions, systems or projects (Hu et al., 2019).It is already an implicit standard, because the term intelligence is often not even mentioned in science and practice, but impacts marketing practice as social engineering (Fan et al., 2015;Lies, 2019).

Methodology
To better understand challenges in decoding consumer behavior with DS, this paper presents a systematic literature review around the concepts, tools and techniques behind the increasing field of DS applied to CBR.
One way to achieve greater rigor and better levels of reliability in a literature review is to adopt a systematic approach, which allows the researcher to make a rigorous and reliable assessment of the research carried out within a specific topic (Brereton et al., 2007;Levy & Ellis, 2006).The result must be the "state of the art" and demonstrate that the research in question contributes something new to the existing body of knowledge, the methodological approach is mainly supported in three phases: input; processing, and output (Sampaio, 2007).The input phase begins with the definition and presentation of the main goal of this research: "Determine the most recent applications of DS techniques in CBR context".
After that, continues with the process of data source identification requiring the definition of rigorous string that suits the different bibliographic databases selected.Scientific articles (ar) or conference proceedings (cp) related to CBR and DS from six main academic databases were searched.These academic databases include Springer Link, Web of Science, Scopus, IEEE Explore, Google Scholar, and Science Direct.Concerning the goal of identify the publications related to research works around the application of DS in CBR, in the first, it is used the string: TITLE-ABS-KEY ("DS") AND TITLE-ABS-KEY ("CBR") AND (LIMIT-TO (DOCTYPE, "cp") OR LIMIT-TO (DOCTYPE, "ar")).So, applying exclusion and inclusion criteria cited, the total document results are 1029.In this case, as all results are about recent articles, published between 2016 and 2020, in the English language.
All the publications titles and abstracts were read manually for relevance checking.This process resulted in 968 publications being excluded.Lastly, 61 eligible publications were selected and added 3 more from the snowballing process.The analyzed publications were investigated based on the relevance to the research domain and availability.

Findings and Discussion
A business may experience thousands of digital interactions with a single user across display, search, social, and on the site or app (He et al., 2018).These interactions take place on multiple devices, such as mobile, desktop, tablet, or wearable devices.Companies can use diverse consumer-related data (Batistič & der Laken, 2019).However, with the rise of AI and ML algorithms, analyzing data points from multiple data sources to create a holistic view of users is now realistic and attainable (Bąska et al., 2019;Sá et al., 2018).

Impact of DS.
The impact of DS has been felt across a range of activities, by providing solutions for many industries that have been struggling for a long time.The most active data generators and consumers are the public sector, healthcare, manufacturing, and retail (Novikov, 2020).Specifically, the big data and business analytics market was valued at USD 138.9 billion in 2020 and is forecasted to grow up to USD 229.4 billion by 2025, at a Compound Annual Growth Rate of 10.6% during the forecast period (Market Reports World, 2020).
The paths to transform digital information into value and to allow companies to become data-driven can be schematized into four major components, presented in The ability of DS to visualize consumer behavior has enabled to predict consumer likes and dislikes and has taken the capabilities beyond mere data collection and analysis.By incorporating the right tools and processes, businesses can now efficiently utilize the insights to influence the decisions of consumers through robust communication (Finoti et al., 2019).

Taking Consumer Behavior to the Next Level with DS
Behavioral research in information systems employing quantitative methods has traditionally relied on mainly survey-based approaches to gather subjective user data.With new advances in technology such as mobile computing, wearable devices, and social media, along with computational capabilities, organizations are in a position to leverage objective data in addressing IT issues typically addressed in behavioral research (Ducange et al., 2018;Motiwalla et al., 2019).DS takes data analysis to the next level, allowing businesses to predict what users might do (Khade, 2016).
Consumer analytics play an essential role in organizations since businesses have access to vast consumer interaction data from multiple channels, including mobile, social media, stores, and e-commerce sites (Chung & Park, 2018).Intended to provide the answers to diverse consumer-related questions, consumer analytics can embrace different types of business analytics.Relying on DS and ML, these two types can provide forecasts and recommend actions that a business can take (Auder et al., 2018).

Forecasting
Arguably one of the most significant benefits of using data in consumer behavior analytics is forecasting Creates intelligent and evidencebased estimates of sales goals based upon current and past sales performance reports.

Conclusion
In the future, AI is likely to substantially change both marketing strategies and consumer behaviors (Davenpor et al., 2020).The efficient implementation of DS will enable organizations to enhance the overall consumer experience by developing robust data analytics models (Fernández-Manzano & González-Vasco, 2018).In particular, marketers must incorporate analytics into their daily decisions around lead generation campaigns, advertising, events and the myriad other ways marketing investments are allocated to affect consumer satisfaction, brand awareness, trust, loyalty, consumer perceived value and consumer retention (Mosavi et al., 2018;Motiwalla et al., 2019).
This paper intended to give an overview of possible ways to anticipate consumer data intelligence development from the perspectives of a robust data set and deep AI expertise for better understanding, modeling, and predicting consumer behavior.
(Dew et al., 2020), 2018)ng capabilities, increases in data transparency and open data sources, growth in the Internet of Things, and smartphone device usage are some of the drivers helping bridge the worlds of data, people, and things(Gupta et al., 2018).For many industries, DS has emerged as a leverage to predict trends and make informed decisions(Tkaczynski et al., 2018).The Connection Between DS and CBR Understanding how consumers think, feel, and respond to a company's offerings has always been a tricky business(Hsu, 2017).Marketing research relies on individuallevel estimates to understand the rich heterogeneity of consumers, firms, and products(Dew et al., 2020).Accessibility to large datasets enables the application of complex DS algorithms and tools to process huge amounts of bytes of unstructured information, allowing relevant feature extraction and recognizing high-level abstractions with increasing generalizability.

Table 2 .
Four main consumer data types.

Table 3 .
Key concepts of Consumer analytics

Table 4 .
Four strategic focus areas