Digital Marketing has become a crucial strategic tool for commercial banks to gain a competitive edge and enhance customer satisfaction levels. So, what is digital marketing? It involves using big data mining to precisely target, segment, market, manage, and maintain commercial bank customers. For commercial banks, a significant vehicle for digital marketing is the development and application of customer tags. Retail customer tags are fundamentally a customer-centric profiling system, which is based on data mining technology, relationship network analysis, customer behavior tracking technology, and text analysis technology to deeply analyze customer demographic characteristics, assets, liabilities, transactions, and other attribute information.
Based on interviews with front-line marketers, literature review, and expert discussions, this article introduces the definition and connotation of commercial bank retail customer tags, eight attribute dimensions, four tag development methods, and customer tag application analysis based on the “5H” framework. With this, front-line marketers in commercial banks can accurately identify potential customers, consolidate and enhance customer loyalty through the sale of wealth products, and truly achieve a win-win with customers.
The Essence of Digital Marketing
Digital marketing is a concept frequently mentioned in the context of big data mining in commercial banks. Generally, digital marketing encompasses two elements: customer relationship management and big data mining. Essentially, digital marketing refers to commercial banks adopting a “customer-centric” approach to carry out and implement customer acquisition, retention, maintenance, and retention CRM activities based on BDM, thereby enhancing the bank’s refined management level of customers.
Specifically, the digital marketing of commercial banks is based on an enterprise-level data warehouse, using certain technologies and methods (such as statistical algorithms, machine learning algorithms, etc.), to extract hidden, valuable information and knowledge from data, identify patterns in the data, and thus explain known facts, predict future customer behavior and business development patterns, effectively assisting marketers in customer relationship management and business improvement.
For commercial banks, an important purpose of digital marketing is to achieve precise marketing. First, through quantitative techniques such as statistics and machine learning, accurately identify potential customers, customer acquisition, customer maintenance, and consolidation. For example, commercial banks can identify which customers have a higher response rate to wealth marketing activities, which customers have greater asset potential, which customers are easier to upgrade, which customers are on the verge of churn, which customers have lower credit levels, and which customers are more likely to commit fraud. Second, digital marketing is individual-level marketing; it can monitor customer attributes and characteristics, identify personalized customer needs, and provide customized services, thereby helping to improve customer satisfaction and loyalty, and helping commercial banks gain a leading competitive advantage.
Digital Marketing and Customer Tagging System
The fundamental purpose of digital marketing is to achieve precise marketing, and the customer tagging system built based on customer assets, liabilities, and transaction information is an important tool for identifying customer characteristics, discovering potential customers, and conducting precise marketing and management. Therefore, for commercial banks, the retail customer tagging system is an important vehicle and tool for digital marketing.
We understand “customer tags” to mean that commercial banks, based on big data mining (such as statistical algorithms, machine learning algorithms, relationship network analysis, LBS analysis, text analysis, etc.), process and calculate customer attributes, characteristics, and information in the enterprise-level data warehouse and related external information sources of commercial banks to obtain customer tag information, such as “homeowners,” “car owners,” “parents,” “international consumers,” “high-end communities,” “high-end malls,” etc.
Based on retail customer tags, front-line marketers can discover and identify high-value potential customers, outline customer profiles, enhance existing customers, prevent potential customer churn, and monitor risk contagion in customer relationship networks, thus becoming an important tool to assist customer managers in customer relationship management.
Tag Development Motivation
Front-line marketers lack tools for customer development. In the past, when front-line customer managers expanded customers and marketed products, they often relied on referrals from existing customers or screened based on some simple variables (such as customer financial assets, customer levels, etc.). Due to the limited ways for front-line personnel to obtain effective customer lists, it affected the effectiveness and efficiency of bank customer development, product marketing, and business improvement.
The identification methods for high-potential customers are limited. Generally, commercial banks classify existing customers based on customer financial assets or transaction settlement volumes. The identification methods for high-value potential customers are very limited. Some customer attributes or tags, such as whether they have a house, whether they have a car, whether they live in high-end communities, whether they have been abroad, whether they are VIP customers of other banks, whether they have a tendency and hobby for collecting artworks, etc., can highlight the customer’s asset strength. We often find that a customer’s financial asset level is relatively low, but through analyzing the customer’s credit card consumption summary, it is shown that the customer often visits high-end places, often goes abroad, and has a strong actual consumption ability. These customers are the so-called “high potential-low asset” customers, which means that customers have not regarded the bank as their “main bank.” Therefore, if we analyze the relevant consumption summaries and tag customers with “high potential” tags, it will help front-line marketers find precise marketing clues and achieve the enhancement of financial assets of related high-quality customers in the bank.
The traditional model development result variable’s discrete use requirement. How to apply the many marketing models that are usually developed to the ground and create actual value for front-line marketers is an urgent problem that needs to be solved. By discretizing the model results (such as asset enhancement probability) and converting them into five or seven levels of tag results, front-line marketers can directly use the relevant results. For example, for the churn warning model, the customer churn probability value is discretized into tags, that is, high churn probability, medium-high, medium, medium-low, and low five levels, thereby enhancing the experience of front-line tag usage.
Commercial banks need to build a 360-degree customer portrait. Commercial banks currently lack a more complete and highly valuable customer portrait system. By integrating customer information such as consumption remarks, customer addresses, marketing modeling, and basic data from the data warehouse, a 360-degree view of the customer’s portrait system (such as car and house tags, mobile phone tags, lifestyle tags, activity tags, investment tags, etc.) can be outlined, thereby enhancing the front-line personnel’s precise service and management enhancement capabilities.
Tag Connotation and Attribute Dimensions
Tag Connotation
Based on interviews with front-line bank personnel, literature summarization, and expert experience summarization, our understanding of retail customer tags is that they are based on retail customer asset information, liability information, and transaction settlement information. According to the customer’s demographic characteristics, important attribute variables, asset changes, account transactions, consumption remarks, customer address information, modeling prediction information, etc., comprehensively use data mining analysis, relationship network analysis, customer behavior trajectory analysis, text mining analysis, and other technical methods to build a customer-centric portrait system. Typical tags in the customer portrait system include homeowners, car owners, parents, salaried workers, high-end communities, high-end offices, high-end communications, VIP customers of other banks, overseas financial needs, recent high-end mall customers, recent high-end fitness customers, Apple Pay tags, mining model churn warning classification, mining model enhancement potential classification, and comprehensive income contribution, etc. Taking a domestic joint-stock commercial bank as an example, the bank’s retail customer development tags, and based on the frequency of tag use, display the tag usage situation through “word cloud analysis” technology. The larger the font of a tag in the word cloud, the higher the application frequency of the tag during the statistical period.
Tag Attribute Dimensions
According to customer attributes and business needs, it is recommended to start designing and developing a customer tag system from the following eight dimensions: customer transaction attributes, natural attributes, product attributes, asset attributes, liability attributes, risk attributes, behavior attributes, and value attributes. We organize the customer tag attribute dimensions in the order of “T-N-P-A-D-R-B-V” based on the consideration of the bank’s understanding of customers “from shallow to deep.” The basic logic is: facing an external customer, the bank may first see the customer’s transaction transfer information with other customers within the bank; secondly, the bank can consider collecting the customer’s natural attributes externally; subsequently, the bank can compare and analyze the customer’s external collection information with the existing customer natural attributes within the bank, carry out customer segmentation, and consider recommending products that similar customers have already used; products include customer financial asset products (such as savings, wealth management, funds, national bonds, insurance, etc.), as well as customer liability products (such as consumer loans, mortgage loans, small and micro loans, etc.); especially when involving lending business, there may be risks, and the bank needs to identify and prevent them; the above content has involved the customer’s financial behavior attributes, in addition, the customer’s non-financial needs, related services, and products should also be included in the customer behavior attribute observation category; finally, all the above content will be used for the bank to evaluate, identify, and apply the customer’s value.
Customer Transaction Attribute Dimension. The customer transaction attribute dimension depicts the transaction situation of retail customers in commercial banks, including fund transfers, transfers out, payrolls, retention, logins, etc., including activeness, channel preferences, large consumption, large transfers out, fund outflow types, fund periodic transfers out, fund retention ratio, monthly payroll amount segments, mobile banking login times, online banking transaction amounts, and other tags.
Customer Natural Attribute Dimension. The customer natural attribute dimension depicts the demographic characteristics
