MARKET OVERVIEW
Market size
The advent of the internet, smartphones, and various applications has led to a significant increase in digital data, commonly referred to as Big Data Both private enterprises and governments acknowledge the immense potential of harnessing this information to create real value for consumers and enhance productivity over time While Big Data has the capacity to transform businesses and economies, it is data science that truly serves as the catalyst for this change.
● The Big Data as a Service Market size is expected to grow from USD 25.44 billion in 2023 to USD 86.76 billion by 2028, at a CAGR of 27.81% during the forecast period (2023-2028).
● As businesses increasingly adopt data-driven marketing strategies, mobile and hybrid working environments, and worldwide supply networks, cloud computing is becoming more ubiquitous.
The banking market's big data analysis is projected to grow from USD 5.83 million this year to USD 19.72 million, with a CAGR of 23.11% during the forecast period By leveraging big data analytics, banks can gain insights into customer behavior, including investment patterns and shopping trends, which aids in making informed decisions about interest rates across different regions Additionally, big data analytics enables major services to effectively store data, extract valuable business insights, and enhance scalability as data volumes increase.
Future of Big Data
With the data and predictions about Big Data above, we can predict what Big Data will be like in the future.
1 Continued growth and expansion: The Big Data market has been experiencing significant growth over the past decade, and it is likely to continue expanding in the future Organizations across various industries are recognizing the value of using data analytics and insights to drive better decision-making and operational efficiencies.
2 Increasing adoption of advanced analytics: As Big Data technologies become more mature, there will likely be an increasing trend toward adopting advanced analytics capabilities This includes machine learning, artificial intelligence, and predictive analytics that can provide more accurate and actionable insights from vast amounts of data.
An example of Bank used Big data
A number of typical businesses have applied Big Data in the banking sector not only abroad but also in Vietnam Here are some examples:
1 JPMorgan Chase: The largest bank in the United States has used Big Data to analyze customer information and provide expectations about consumer behavior, thereby enhancing risk management and improving customer experience.
2 Wells Fargo: This bank has also developed Big Data to identify customer spending patterns, thereby creating products based on actual user needs.
4 Techcombank: In Vietnam, Techcombank is one of the banks that has successfully applied Big Data They use data from different sources to analyze customer actions, creating products and services tailored to user needs.
5 VietcomBank: This bank has also applied Big Data to business activities Data from sources such as transactions, signal records, and client data are analyzed to optimize risk management and provide key financial solutions to clients.
WHAT IS BIG DATA? TYPE OF BIG DATA
Definition
Big data refers to extremely large and complex data sets that cannot be easily managed, processed or analyzed with traditional data processing applications.
Types of Big Data
Structured data is organized and formatted information that computers can easily interpret By adhering to a specific schema or data model, it enhances the efficiency of searching and processing data.
Examples of structured data formats include relational databases, XML, and JSON b Unstructured Data (with example)
Unstructured data encompasses various forms of information that lack a specific data model or organized structure This type of data is often found in free text, audio recordings, images, videos, social media posts, emails, and other disorganized formats.
Customer reviews for a product or service exemplify unstructured data, as each review varies in format and contains diverse types of information In contrast, semi-structured data includes elements that do not conform to a strict structure but still possess some organizational properties, such as JSON or XML files, which allow for easier data processing and analysis.
Semi-structured data refers to data that lacks a rigid schema or model yet maintains an organizational structure This type of data includes tags, markers, or delimiters that help distinguish various elements within it.
Management Đại học Kinh tế Quố…
The application of E KYC in banking sector
A few examples of semi-structured data sources are emails, XML and other markup languages, binary executables, TCP/IP packets, zipped files, data integrated from different sources, and web pages.
Identify data sources of Big data
- Direct data transmission: data from the Internet of Things (IoT) and connected devices are transmitted into information technology systems from devices such as smartphones and smart cars.
- Administrative data such as electronic medical records, insurance records, and banking records; …
- Data from commercial activities such as credit transactions, and online transactions on mobile devices.
- Data from sensor devices such as satellite images, road sensors, and climate.
- Data from tracking devices such as cell phones and GPS.
- Behavioral data such as online searches for products and services.
- Published available data: is information and data that is widely and publicly available such as official websites of the Governments of countries.
- Other sources: some other data sources come from customers, suppliers, or cloud data.
Data warehouse & Data lake
Big data can be stored in traditional on-premise data warehouses, but there are also flexible and cost-effective alternatives available Cloud solutions, data lakes, data pipelines, and Hadoop offer efficient options for storing and processing large datasets.
Building a data warehouse enables organizations to integrate data from various systems across departments, simplifying data access and facilitating informed decision-making However, as data volumes grow daily, challenges arise in delivering comprehensive insights To address these limitations, the data lake solution was developed, offering a more flexible approach to data management.
Eng CEAC - Nexif - Draft 2006 - to chi…
Differences between data warehouse and data lake:
Criteria Data lake Data warehouse
Type Data All data is retained regardless of source and original structure Data is kept in raw form, converted only when ready for use
Includes data extracted from transactional systems Data is cleaned and transformed
History Big data technology used in data lakes is relatively new
Data warehouses have been utilized for decades, distinguishing themselves from big data They collect various types of data, including structured, semi-structured, and unstructured formats, directly from source systems in their original form.
Structured data and arranged them in schemas as defined to build a data warehouse
Time Data lakes can retain all data This includes not only data that is currently in use but also data that may be used in the future
Additionally, data is retained at all times so that it can be turned back in time and performed analysis
During data warehouse development, significant time is spent analyzing different data sources
User Data lakes are ideal for users who want deep analysis like data
Data warehouses are highly beneficial for users due to their structured design, making them accessible for scientists who require advanced analytics tools These tools offer essential capabilities such as predictive modeling and statistical analysis, facilitating easier data interpretation and utilization.
Cost Storage costs are cheaper than data warehouses
Mission Contains all data and data types, it allows users to access pre-transformed, cleaned and structured data
Provides insights on predefined questions for predefined data types
Data lakes offer a fast processing time by enabling users to access raw data prior to transformation, cleaning, and structuring This capability allows for quicker results compared to traditional data warehouses.
Processing time is slower The data warehouse provides insights on predefined questions for identified data types Therefore, any changes to the data warehouse require additional time
Benefit Integrate different types of data to ask completely new questions
Provide reports and key performance indicators
Limit Data is kept in its raw form, transformed only when it is ready for use
Interbank shared data
In the following, I will share two more data sources that banks can collect for their business activities, including shared interbank data sources and outsourced data sources.
First, let's talk about shared interbank databases As you can see on the slides, do we have a definition of what a shared database is?
Shared databases are integral to various industries, particularly in banking, where they facilitate the management of online transactions These systems enable customers to conduct transactions remotely, eliminating the need for in-person visits to the bank Furthermore, user information is centrally stored, monitored, and automatically processed within the database system.
Evaluation of the Experience of Shared Databases Among Banks in Vietnam and Worldwide:
From there, we can assess the level of application of shared interbank databases in Vietnam through Government Project 06 in the period 2022 - 2025, with a vision to
By 2030, the project aims to enable banking units to access and utilize the National Population Database and Citizen Identification Database, along with electronic identity accounts from the Ministry of Public Security Effective data connectivity is essential for successful digitization in the banking sector, highlighting the importance of shared databases within the Vietnamese banking system.
Outsourced data
Data outsourcing is a data management approach where the data owner shares responsibility for data management with external providers This allows for specialized data management services to be utilized, enhancing efficiency and expertise in handling data.
Benefits of Data Outsourcing in Banking:
- Some benefits of data outsourcing in banking include:
- Eliminating the cost of hiring full-time processors
- Involving trained processors with accumulated knowledge
- Closing loans faster and with increased efficiency
- Having more time to focus on your core competency
- Leveraging big data analytics and modeling
Risks of Data Outsourcing in Banking:
While there are benefits to using third-party vendors, significant risks also exist Recent years have seen consumer banks suffer serious reputational and financial setbacks due to vendor errors For instance, a retail bank experienced a major incident where millions of customers were unable to access their funds because of a computer failure during a software update by an IT vendor This situation was exacerbated when the vendor lost a USB stick containing sensitive customer data, resulting in the illegal sale of that information.
Specific Examples of Vietnamese and Global Banks Outsourcing Data:
Here are some examples of banks in Vietnam and around the world that outsource data:
Vietcombank, officially known as the Joint Commercial Bank for Foreign Trade of Vietnam, enhances its market and customer insights by acquiring data from reputable market research firms like Nielsen and GfK, as well as financial service providers such as Bloomberg.
- Techcombank (Vietnam Technological and Commercial Bank): Banks can also buy data from technology companies like IBM, Oracle, or Amazon Web Services (AWS) for data analysis and system management.
The diverse use of data sources by banks in Vietnam helps improve business decisions, enhance performance, and provide better customer experiences
Importantly, having diverse data sources helps banks spread risks when using data.
CHARACTERISTICS OF BIG DATA
What is VOLUME?
Volume refers to the vast quantities of data generated daily at an unprecedented rate, with major sources including platforms like Facebook, Twitter, YouTube, and the Internet of Things (IoT) Understanding this trend is essential for grasping the scale of data production in today's digital landscape.
What is VELOCITY?
Velocity signifies the rapid rate at which various sources produce data daily, with a continuous and increasing flow Currently, there are 1.50 billion daily active users on Facebook's mobile platform, reflecting a 25% year-over-year growth This surge illustrates the growing number of users and the substantial data generated by each individual If velocity is effectively managed, valuable data insights can be derived, enabling decisions to be made based on real-time information.
What is VERACITY?
Veracity refers to the uncertainty surrounding data caused by its incompleteness and inconsistency In certain instances, the disarray of available data renders it unreliable The sheer volume of such data complicates efforts to uphold quality and accuracy.
Many business leaders fail to utilize data for decision-making due to its uncertain nature A survey revealed that 27% of participants were uncertain about the accuracy of their data Consequently, the US economy incurs an annual cost of $3.1 trillion due to poor quality data.
What is VARIETY?
Variety in data refers to the different types generated from various sources, including structured, unstructured, and semi-structured formats Historically, data generation was limited to formats like Excel, text, and databases However, today, data is produced from a multitude of sources, encompassing diverse types such as sensors, audio, images, and videos.
What is the VALUE?
Value in data pertains to its utility in decision-making processes By effectively analyzing vast amounts of big data, organizations can enhance their value Conversely, without this analysis, data remains ineffective and unbeneficial.
APPLICATION OF SOME INDUSTRY
Retail Industry
Big Data enables the Retail Industry to analyze diverse customer data, uncovering insights into market competition and customer interests This analysis helps businesses identify customer experience journeys, shopping trends, and levels of satisfaction, ultimately enhancing performance and sales efficiency.
Big Data algorithms play a crucial role in combating payment fraud by analyzing transaction data to detect issues such as credit card fraud and account takeovers This technology enables retailers to evaluate customer creditworthiness, thereby reducing the risk of non-payment on high-value purchases Additionally, retailers leverage Big Data to identify and prevent fraudulent returns, effectively safeguarding against the abuse of return policies.
Big Data plays a crucial role in loss prevention by analyzing inventory shrinkage patterns caused by theft or errors, enabling retailers to implement effective preventive measures Additionally, data analytics can detect suspicious employee behavior that may signal fraud, significantly reducing the risks of internal theft.
Supply Chain Risk Management involves assessing supplier risks through Big Data, enabling retailers to consider geopolitical factors, transportation challenges, and production delays for better sourcing decisions Additionally, Big Data plays a crucial role in logistics optimization by enhancing shipping routes and delivery schedules, thereby minimizing supply chain disruptions and associated delivery risks.
Customer satisfaction is crucial for the success of any business, as happy customers lead to loyalty Big Data empowers retailers to improve customer experiences by analyzing data to predict demand and make informed, customer-focused decisions By leveraging insights from various sources such as websites, mobile apps, social media, and sensors, retailers can personalize marketing campaigns effectively This strategic use of Big Data not only enhances customer engagement but also strengthens a retailer's competitive position in the market.
Example: Case study of Mall of America, Bloomington, Minnesota
The largest shopping center in the northern United States boasts over 500 retailers, 50 restaurants, 14 cinemas, 2 hotels, indoor theme parks, and museums With its vast complex structure, the center experiences tens of thousands of transactions daily, making it challenging to offer personalized experiences to its diverse customer base The key to enhancing customer experience lies in the utilization of Big Data.
IBM introduced the ELF Chatbot to enhance customer navigation within their extensive commercial center, enabling a deeper understanding of customer needs This innovative tool provides personalized experiences by offering tailored product recommendations, such as a list of suppliers for gas stoves that match desired specifications and competitive pricing Additionally, after a purchase, the Chatbot offers valuable advice on maintenance and complementary products to enhance the kitchen's comfort and luxury.
Big Data algorithms analyze customer behavior, purchase history, and preferences to provide personalized product recommendations, increasing cross- selling and upselling opportunities.
Big Data Analytics empowers retailers to understand customer habits, identifying high-demand products and services while determining which offerings to discontinue By analyzing data, retailers can also forecast product returns, enhancing restocking processes and customer communication Furthermore, this technology allows retailers to anticipate significant shifts in the retail landscape, enabling them to introduce new products aligned with current market trends Ultimately, Big Data facilitates accurate predictions of consumer demand, boosts business efficiency, and positions retailers to lead in the market.
Also, retailers in certain sectors (e.g.: fashion, and home improvement) use weather data to anticipate demand fluctuations due to weather conditions and minimize weather-related risks.
Big Data is revolutionizing the retail industry by allowing retailers to monitor store-level demand in real-time, ensuring that best-selling items are always in stock It streamlines product lifecycle management and complex operations, enhancing understanding of the supply chain and distribution to lower costs Additionally, Big Data helps retailers optimize asset utilization, budgeting, performance, and service quality Valuable data is generated from various assets, including factory machinery, customer-owned equipment, energy grid infrastructure, and product logs.
E-commerce
Big Data enables businesses to identify customer interests and deliver targeted advertisements Furthermore, it allows companies to forecast upcoming demand for products, understand market preferences, and effectively manage inventory and allocation of goods.
Fraud detection and prevention are critical for e-commerce platforms, which utilize transaction monitoring to analyze large volumes of data in real-time, identifying unusual patterns that may signal fraudulent activity By employing machine learning algorithms, these platforms can effectively detect anomalies and flag suspicious transactions for further investigation Additionally, user behavior analysis through Big Data analytics allows for the profiling of customer behavior, enabling the identification of deviations from typical patterns, such as an unexpected high-value purchase.
Customer behavior analysis plays a crucial role in e-commerce, particularly in understanding shopping cart abandonment By leveraging big data, businesses can identify the underlying reasons for this phenomenon and develop effective strategies to recover potentially lost sales Additionally, customer segmentation enables e-commerce platforms to categorize customers based on their behavior, demographics, and preferences, facilitating more targeted marketing and messaging efforts.
Big Data empowers businesses to enhance user experiences, exemplified by the use of artificial intelligence applications like chatbots However, marketers do not solely depend on chatbots; they aim to gather and analyze customer data to deliver optimal experiences for their clients.
Timing is crucial for marketers, as each business has a diverse customer base with unique characteristics Analyzing Big Data enables businesses to identify effective content and optimal delivery times By collecting data on followers and customers, businesses can foster ongoing connections through social media and email marketing Moreover, analysis tools allow marketers to pinpoint the best times for posting or sending emails to maximize customer engagement.
E-commerce companies utilize big data to optimize pricing in real-time, adjusting based on demand, competitor pricing, and historical sales data to maximize revenue Additionally, they leverage big data analytics for demand forecasting, which helps in maintaining optimal inventory levels and minimizing the risks of overstocking or understocking Furthermore, big data plays a crucial role in market research and trend analysis, enabling businesses to identify new product opportunities and emerging consumer preferences while gaining insights into competitors' strategies and performance in the e-commerce landscape.
Analyzing business page content is essential for reducing abandonment rates and enhancing customer click-through rates, ultimately leading to improved order conversion rates With 48% of Big Data focused on customer analysis, understanding customer behavior is crucial for encouraging repeat purchases The more detailed data you gather on visitor behavior, the more effectively Big Data can assist in boosting your conversion rates.
Digital Marketing
The Fourth Industrial Revolution changed many industries, including the economic sector The explosion of technology has increased its impact.
Digital marketing is crucial for the success of businesses, particularly as social media and online shopping become essential in people's lives The integration of Big Data into digital marketing strategies has made it an indispensable component for companies looking to thrive in today's market.
○ Payment Fraud: Big Data analytics can detect unusual payment patterns and behaviors, helping to identify and prevent payment fraud, which is a risk for e-commerce and online payment systems.
○ Account Takeovers: By monitoring login and account activity data, big data can flag suspicious login attempts or unusual user behavior, reducing the risk of account takeovers.
○ Big Data is used to identify and prevent ad fraud in digital advertising It analyzes patterns of clicks and impressions to detect fraudulent activities, protecting ad budgets from wastage.
● Increase customer acquisition and retention
Leveraging Big Data enables businesses to analyze customer behaviors and trends, fostering loyalty in today's technological landscape By implementing a robust Big Data analytics strategy, companies can effectively gather and utilize customer data This analysis provides valuable insights into essential behaviors necessary for retaining customers Understanding these insights allows businesses to meet customer expectations, which is crucial for achieving high customer retention rates.
● Solve advertiser problems and provide insights for Marketing
Big Data analytics has the potential to revolutionize businesses by enhancing customer satisfaction, refining product offerings, and optimizing marketing strategies By analyzing data from various Big Data sources, marketers can gain insights into customer purchasing behaviors, tailor advertising campaigns effectively, and boost the overall efficiency of brand marketing efforts.
● Drive innovation and product development
Big Data drives innovation and product development for businesses by gathering customer data, capturing feedback, and analyzing user needs This continuous flow of accurate information enables companies to redesign existing products and create new ones, similar to how Apple frequently launches new phones.
○ Customer Funnel Analysis: Big Data helps track the entire customer journey, identifying touchpoints, drop-offs, and areas for improvement.
○ Attribution Modeling: Attribution models determine the impact of different marketing channels and touchpoints on conversions, optimizing marketing spend.
○ Big Data can be used to scan and analyze ad content to ensure it complies with advertising standards and industry regulations, reducing the risk of regulatory fines.
Social media sentiment analysis utilizes big data tools to evaluate mentions and sentiments related to a brand, enabling effective monitoring of brand reputation This process helps in identifying potential risks to reputation and facilitates prompt responses to negative public relations challenges.
○ Crisis Management: Real-time analysis of social media data can help companies detect and manage crises or PR emergencies, reducing reputational damage.
APPLICATION OF BIG DATA IN THE BANKING SECTOR
Analyzing customer spending habits
Banks have direct access to a wealth of information and historical data regarding customers' spending habits and behaviors, including detailed insights into their annual income and the banking services they utilize This data enables banks to conduct in-depth analyses, particularly by applying information filtering techniques to account for factors such as holiday seasons and macroeconomic conditions like inflation and unemployment rates Understanding these variables is crucial for assessing fluctuations in a bank's income and expenditures, which plays a vital role in risk assessment, loan appraisal, and the expansion of service offerings or cross-selling products to customers.
2.Customer segmentation and audit records.
Customer segmentation is crucial for effective bank marketing strategies and product design By analyzing customer consumption habits and identifying preferred service types, banks can create a comprehensive database that aids in the segmentation process Utilizing big data allows banks to gain valuable insights into customer needs, habits, and spending trends, enabling them to tailor their offerings Understanding transaction-related information helps banks categorize customers into groups such as easy consumers, cautious investors, and loyal clients This knowledge empowers banks to forecast expected expenditures and income, facilitating detailed planning that aligns with both company profits and customer interests.
Improve service quality by establishing a system to collect and analyze customer feedback
Customers can provide feedback after transactions or through customer support, but they often share their opinions on social media platforms like Facebook and Zalo Big data tools can effectively analyze this information, gathering public feedback about bank brands to ensure timely and comprehensive responses to customers Additionally, these tools can help mitigate the impact of rumors that could harm business operations and customer trust When banks actively listen to and value customer opinions, making necessary improvements, it enhances brand loyalty and positively impacts their image.
Personalized marketing
To effectively target customers, banks must leverage marketing strategies informed by an understanding of individual spending habits By analyzing transaction history alongside unstructured data from social networks, banks can gain a comprehensive view of customer needs and preferences This insight allows banks to develop tailored solutions and marketing plans that enhance customer engagement For instance, utilizing email marketing, banks can inform customers about competitive short-term loan options, attractive savings deposit rates, and various incentive programs.
Detect and prevent fraud and illegal behavior
Big data enhances banking security by preventing unauthorized transactions and improving industry standards By analyzing customers' transaction histories and credit records, banks can identify unusual activities, such as large ATM withdrawals that may indicate card theft To combat this, banks implement security measures to verify transactions Utilizing data analysis and machine learning algorithms, banks can differentiate between criminal and legitimate transactions, enabling real-time detection and extraction of illegal activities while recommending prompt actions.
Risk control, legal compliance, and financial reporting transparency
Big data algorithms enhance legal compliance in accounting, auditing, and financial reporting, leading to lower management costs Additionally, big data collection and storage enable banks to swiftly analyze risks and implement response strategies Furthermore, big data facilitates coordination among various departments and integrates data processing requests into a centralized system, thereby supporting control measures, preventing data loss, and mitigating risks and fraud.
CASE STUDY BANKS USE BIG DATA
TP BANK
● TPBank is actively employing Big Data to enhance customer experiences.
TPBank's technological transformation is exemplified by its LiveBank model, an automated transaction system available 24/7 that replaces conventional branches Since its launch in 2017, TPBank has expanded to 330 LiveBank locations across the country The bank's guiding principle, articulated by Mr Hung, emphasizes the importance of thinking big, starting small, and scaling success.
● In September 2019, TP Bank became the first bank, in collaboration with FPT
IS, to implement and analyze large-scale big data systems within the banking sector.
On July 29, 2023, TP Bank formed a strategic partnership with FPT IS to improve service quality for corporate clients This collaboration focuses on exchanging technology and digital solutions to facilitate and expedite the digitization processes for various enterprises and individual businesses.
● TPBank has redefined the banking experience by providing essential financial products and services through modern technology such as AI and Big Data
TP Bank has achieved significant success with BoostML, a tool that leverages Big Data and AI to provide advanced algorithms This innovative solution allows customers to quickly and accurately evaluate data information and models.
TP Bank has successfully implemented BoostML to create an automated customer service experience through the Messenger platform, allowing for personalized and efficient support without long wait times Customers can also take control of their experience by managing functions like changing their PIN and locking their cards in emergencies.
● Since its implementation, TPBank has seen an average monthly increase in customers of 10%, doubling the productivity of agents, and reducing customer wait times by 50%.
As of September 2022, TPBank has reached over 7 million customers, with 70% belonging to the younger demographics of Gen Y and Gen Z In 2022, the bank successfully added 2 million new customers, achieving a total growth of 100% Currently, over 2.5 million TPBank customers engage in regular transactions via digital channels.
In the first five months of this year, online banking registrations surged by 87% compared to last year, with a remarkable 790% increase in account openings through the comprehensive eKYC method on the TPBank app Additionally, online transactions have skyrocketed, now representing 92% of all transactions at the bank.
TPBank has launched innovative products and services tailored to customer trends, effectively identifying and targeting their audience By leveraging technology, TPBank has successfully attracted new customers, improved brand favorability, and strengthened the loyalty of its existing clientele.
MCREDIT
The story of Mcredit, a Limited Liability Financial Company, is a tale of a consumer finance enterprise that has made significant investments in its technology platform, digital transformation, and data analysis.
Founded in 2016, Mcredit is a joint venture between the Military Commercial Joint Stock Bank and Japan's Shinsei Bank In 2021, the company demonstrated remarkable growth in both quantity and quality, overcoming challenges from the Covid-19 pandemic through efficient operations, increased market share, and strong credit quality.
From the 4th position in the consumer finance sector in 2018, by the end of
In 2021, Mcredit achieved a total asset scale of nearly 19 trillion VND, securing the 3rd position previously held by HD Saison With a market share exceeding 9%, Mcredit is closing in on the 2nd position, which has been consistently held by Homecredit from 2016 to 2022.
Mcredit aims to become the leading company in efficiency by 2026, positioning itself among the top two in scale while targeting a customer base that is nearly ten times larger than it was in 2020.
Mcredit is implementing strategic initiatives through programs that prioritize a customer-centric approach, distinguishing itself from the prevailing product-centric strategies in the market This shift places customers at the heart of Mcredit's overall strategy.
The application of Big Data and AI has contributed to addressing these three trends as follows:
In the consumer finance industry, organizations that deliver services quickly gain a significant competitive edge, as the duration from customer application to disbursement is increasingly reduced due to market competition Leveraging Big Data and AI enables these companies to streamline processes, allowing for application, assessment, approval, and disbursement to be completed in just minutes.
Consumer finance organizations that prioritize convenience and provide customer-centric services tailored to individual needs are likely to thrive By leveraging Big Data and AI, these organizations can accurately identify customer requirements, enabling them to offer the most appropriate products and incentive programs.
Data and AI in the consumer finance industry is to enhance the customer experience and streamline company operations.
By leveraging Big Data and AI, consumer finance organizations can optimize their processes and reduce operational costs, gaining a competitive advantage Enhanced risk assessments enable these organizations to lower provisioning costs, allowing for reduced interest rates and increased customer incentives while preserving profit margins Mcredit effectively utilizes Big Data to achieve these outcomes.
Applying Big Data for Loan Approval
Mcredit's loan approval process incorporates two crucial checkpoints: identity verification alongside screening customers with unfavorable credit histories and customer credit scoring.
Mcredit has effectively offered financial services without requiring physical interactions, emphasizing data analysis and customer personalization This shift has led to quicker credit risk assessments and enhanced service delivery times, ultimately attracting more customers who value these conveniences.
Credit scoring outcomes are essential for customizing products, interest rates, and credit limits to match individual customer profiles The integration of Big Data and AI has enhanced the automation of credit scoring and assessment processes, enabling proactive operations and tailored offerings for each customer.
- The actual non-performing loan ratio of Mcredit decreased from 6.5% in 2020 to 6.2% in 2021.
- Profit per employee increased from 63 million VND per person in 2019 to 214 million VND per person in 2021.
At the end of 2021, Mcredit achieved a Cost-to-Income Ratio (CIR) of 33%, which is 4.2% lower than the industry average of the top five companies This improvement contributed to an increase in the company's net profit margin, rising from 8.7% in 2020 to 13.7%.
- The Return on Equity (ROE) for the company in 2021 reached 28.8%, marking a 47% growth compared to 2020.
- Profit per employee has increased more than threefold in the past year.
Using big data to understand target customers
Big Data and AI are not only tools for assessing "risk" but also contribute to clearly identifying customer profiles, even in "real-time."
Mr Khang highlighted that Mcredit's credit scoring models, including A-score, B-score, and C-score, effectively utilize both traditional data, such as demographics and credit information, and non-traditional data sources This innovative approach allows Mcredit to predict customer churn and identify potential customers for growth, enabling proactive customer retention strategies that conserve resources on new customer acquisition while enhancing sales and managing existing relationships.
Furthermore, analyzing Big Data helps Mcredit build a "customer profile" to understand customer needs, allowing them to formulate appropriate marketing strategies and products.
Customers benefit from loan offers featuring lower interest rates, thanks to decreased operational costs for companies Additionally, the loan processing time is significantly reduced for high-quality segments, enhancing the overall customer experience and satisfaction.
Customers who engaged with Mcredit in 2021, when combined with those from the preceding four years, saw brand awareness of Mcredit increase from 39% in
2021 to 67% in 2022 (a nearly 71% increase compared to 2021) These are positive indicators of Mcredit's transformation.
In reality, Mcredit experienced rapid growth in both 2021 and 2022 In 2020, Mcredit's market share was only about 6.4% By 2021, it had risen to 9.6%, and estimates for 2022 suggest it is around 12%.
Mr Le Quoc Ninh outlined Mcredit's goals for the 2022-2026 period, focusing on market leadership efficiency through two key indicators: return on equity (ROE) and profit per employee (labor productivity) Currently, Mcredit boasts a leading ROE of approximately 41%, surpassing the highest among banks at around 30% Although this ratio may decrease due to planned capital increases, the company aims to maintain its leadership position Mcredit has also developed a clear roadmap for enhancing labor productivity, with ongoing transformations already yielding positive results.
We enhance cost efficiency and digitization to better serve a larger customer base Our workforce growth is significantly lower than operational growth, allowing us to optimize personnel and achieve early milestones in profit per employee.
Mcredit is undergoing a significant digital transformation by leveraging technology to optimize its Hybrid Cloud platform architecture and enhance cybersecurity measures This transformation includes the design of advanced data systems, analytical tools, and data management infrastructure that facilitate informed business decisions Additionally, Mcredit is focused on optimizing operations and processes through Business Process Management (BPM) and Customer Relationship Management (CRM) systems.
DWH/Datalake, and Smart Collection.
Case study of how banks and other financial institutions use big data in risk management
The adoption of Big Data and advanced analytical tools for risk mitigation is on the rise A survey by the European Banking Authority (EBA) highlights a notable increase in the utilization of these tools for risk-scoring and risk-modeling among banks and analysts.
Advanced analytics models utilizing extensive institutional data can enhance credit scoring for primary customers, especially when combined with external data Meanwhile, non-bank customers can be evaluated through API access to payment data, enabling innovative services like instant lending for non-bank clients and pre-approved loans for primary customers.
Now, let’s look at some examples of how big data has been used in banks regarding the risk management field.
Big data can make a significant breakthrough in the financial and banking industry due to the vast amount of unstructured data being used for analysis, assessing the creditworthiness of customers
To assess a borrower's creditworthiness, banks typically use financial data such as employment contracts, salary statements, and credit history from credit information centers (CIC) However, this approach has limitations, as individuals without prior borrowing or credit card experience lack a credit history Moreover, many credit card holders may not possess formal employment contracts, and their salary statements might not accurately reflect their true income Additionally, personal financial situations can change over time, rendering the information in CIC outdated for those who have previously borrowed.
Credit scoring methods leveraging Big Data and AI transform all types of data into valuable insights Behavioral data, online shopping habits, telecommunications records, and health data can all play a role in assessing creditworthiness Even customers without bank loans have monthly payments that reflect their financial management Timely payments are indicative of a customer's discipline and reliability in meeting financial obligations, providing crucial information for credit assessments.
Kreditech is a German online lending company that evaluates individuals' creditworthiness through their online data rather than traditional credit ratings Utilizing a self-learning algorithm, Kreditech analyzes big data to compute credit scores in seconds, drawing from up to 20,000 data points The company incorporates location-based information, social networking activity, online shopping behavior, and general online behavior to assess loan applicants' creditworthiness.
This approach recognizes that traditional methods of assessing an individual's creditworthiness are often inadequate Instead, the company aims to leverage all available resources to accurately evaluate a person's financial situation at any given time.
Lendit, a peer-to-peer financial lending service based in South Korea, utilizes a big data credit score model that incorporates borrowers' Facebook information and their behavior while reviewing investment instructions on its website.
A job seeker who recently transitioned to a lower-paying position at a venture company applies for a loan By examining the applicant's social media, it becomes evident that he and his family have stable professional backgrounds and reliable income sources Consequently, big data analytics indicates that the applicant's ability to repay the loan is rated higher than his income level, enabling him to secure more favorable loan terms.
A new credit scoring model for the fintech industry
● On improving risk management system:
The United Overseas Bank (UOB) Limited, the third-largest bank in South East Asia, has launched a big data-driven risk management system of their own since 2015.
By harnessing big data and advanced technological analysis, the bank significantly reduced the time required to calculate value at risk from up to 20 hours to just minutes This efficiency opens the door to the possibility of implementing real-time risk analysis in the near future.
● KEB Hana Bank leveraged big data to strengthen their cyber security
By developing a system that collects and analyzes log data using big data, the bank is able to prevent malicious code attacks and intrusions.
CitiBank is making significant investments in big data technologies through its initiative, Citi Ventures, which includes funding promising startups and forming partnerships with tech companies A key area of focus for the company is cybersecurity, reflecting its commitment to enhancing security measures in the digital landscape.
CitiBank has strategically invested in Feedzai, a data science firm that leverages real-time machine learning and predictive modeling to analyze large datasets, effectively identifying fraudulent activities and reducing financial risks for online banking services.
CitiBank's ability to detect suspicious transactions, such as unusual charges, allows for timely notifications to users This service not only benefits consumers but also aids payment providers and retailers in monitoring financial activities and identifying potential threats to their businesses.
VII CHALLENGES OF BIG DATA IN THE BANKING SECTOR
Financial resources
According to the Vietnam Report (2019), the primary challenge for organizations in Vietnam is securing substantial investment funds for technology acquisition and application Many Vietnamese companies seeking to utilize big data often rely on foreign service providers, incurring significant costs for royalties and professional consulting The infrastructure costs associated with big data platforms like Hadoop and Spark are directly linked to the storage, computing, and processing capabilities utilized by banks Additional expenses, including management, maintenance, network connectivity, and data preservation, further complicate the financial landscape While large banks have the capital for investment, they must carefully evaluate costs against profit and strategic priorities in their annual growth plans In contrast, small commercial banks remain cautious in their investments to maintain a balance in operating costs.
To optimize financial resources, banks should leverage cloud computing technology, which allows access to pre-built analytical models and hypercomputing resources on a pay-as-you-go basis Collaborating with fintech companies can help reduce data collection and cleaning costs, fostering a cohesive supply chain model that aligns more closely with customer needs This approach will enhance the service supply chain, resulting in a unified, transparent, and efficient database at every stage.
Human Resources
To effectively harness big data and create value, banks in Vietnam and the ASEAN region must prioritize high-quality human resources and competitive recruitment There is a significant talent shortage in the banking sector, particularly in skills related to statistics, data mining, and big data technologies Additionally, many bank managers lack the necessary expertise in information technology, highlighting the need for collaboration with IT departments to develop comprehensive digital strategies To optimize big data applications, it is crucial for managers to be well-prepared with robust development strategies, especially concerning human resources.
To enhance banking modernization and Big Data applications, organizations should not only recruit experienced personnel but also focus on developing internal human resources through cost-effective training in data analysis, programming, and databases Business departments must adopt various programming languages for proactive data processing and a holistic understanding of the data system Meanwhile, the information technology department should stay updated with global technology trends in database design, query processing, and data integration to ensure rapid adaptation.
The immense volume of big data presents significant challenges for storage and processing, particularly for banks whose IT infrastructure may not fully leverage its potential Key technical hurdles include data collection, editing, and interference filtering, necessitating intelligent storage design and rapid processing of queries Additionally, the demand for complex in-depth reports and effective data integration requires advanced modeling and analysis Furthermore, banks face challenges in technology selection, algorithm updates for privacy protection, and ensuring the security of large data sets, especially with the increasing reliance on cloud storage solutions.
To effectively leverage big data, it is essential to establish robust infrastructure for data collection and storage, ensuring secure access and transmission This includes implementing storage systems, servers, management software, data integration tools, and data analysis software Banks should opt for interconnected big data solutions and utilize data from existing systems to maximize the use of previously deployed infrastructure and resources.
Data
Enterprises often struggle to manage the vast array of data they encounter, particularly when distinguishing valuable information from irrelevant content Despite the growing proportion of potentially useful data, the overwhelming amount of useless data complicates this process Consequently, companies must enhance their data analysis techniques and explore innovative applications for data that may initially seem insignificant.
Vietnamese consumers exhibit fluctuating psychological and behavioral patterns, frequently influenced by short-term trends and digital communities, including social media influencers This variability presents a challenge for banks in accurately analyzing customer needs Missteps in this analysis can turn big data into a double-edged sword.
As banks increasingly adopt cloud storage and collaborate with financial technology firms, the focus on customer information security intensifies With extensive networks of trading points and international branches, banks must prioritize the establishment of control procedures, decentralization of system access, and training of staff on customer identification and data security The rise in cyber security crimes, coupled with the global complexity of threats, necessitates that banks exercise caution to safeguard data across systems, personnel, and partnerships Failure to ensure these security measures could expose banks to significant risks, impacting their reputation, media perception, and overall business operations.
While cloud computing providers present certain security benefits compared to traditional bank data centers, such as enhanced centralized data protection and monitoring, they also introduce risks associated with third-party data management Consequently, NHTM must implement rigorous standards and systems for partner and personnel recruitment and management It is crucial to prioritize the establishment of a data backup center for post-disaster recovery, enhance the overall safety system, and elevate security levels to ensure comprehensive data protection.
LEGAL FRAMEWORK
The government has recently issued Decree 13 to facilitate the digital transformation of banks; however, several issues remain unresolved It is crucial for the government to review and enhance legal documents within the banking sector, as well as to establish and improve national databases to ensure open access for banks operating under approved jurisdictions Additionally, investment in technological infrastructure and the promotion of a modern engineering environment are essential Strengthening partnerships for technology transfer from advanced countries will provide a solid foundation for the adoption of new technologies in banking Legal provisions that assist banks in customer identification and verification will instill greater confidence in deploying digital banking services and effectively managing customer disputes.