A conceptualizing and designing scalable beyond banking ecosystemns using machine learning and data analytics methodologies

Abstract

This thesis examines the concept of Beyond Banking, framing it as an intra-institutional (C2C) banking ecosystem approach. The literature review shows that most research on Beyond Banking focuses on the intra-institutional customer data matching (RQ1) model involving third-party partners. There has been limited exploration of C2C at the bank level within Beyond Banking, where a bank acts as a marketplace facilitator connecting its own customers, who can serve as both providers and consumers within the ecosystem. This indicates that banks in German-speaking countries have a limited understanding of Beyond Banking in relation to ecosystem strategies. Furthermore, a lack of well-defined business cases and essential skills hinders the development of effective Beyond Banking services and the generation of targeted data insights through data analytics (DA) and AI/ML techniques. Leading the field are Asian big tech firms and banks, which develop banking ecosystem offerings substantial value via scalable super apps with many participants and bundled services (Bian et al., 2022).

Based on the research gaps identified, this dissertation examines three central research questions:

RQ1: How can combining intra- and inter-institutional customer data matching improve the scaling of Beyond Banking Ecosystems in the financial sector?

RQ2: How can machine learning and data analytics methodologies be utilised to develop and implement effective Beyond Banking Sales Signals, and what are the specific use cases that demonstrate their impact on sales performance in the financial sector?

RQ3: What are the main influencing factors and evaluation criteria for successfully creating a scalable Beyond Banking Ecosystem, and how can these factors be incorporated into the developed model?

The aim is to develop an evaluated design for creating scalable Beyond Banking Ecosystems (BBE) using machine learning (ML) and data analytics (DA) methodologies. The design will be created using the Design Science Research methodology, which will be developed through four evaluation cycles involving experts first in the local services a tech 1 by sect vs. n, it e cycle is assumed successfully. Before evaluating t e design, sufficient data will first be gathered from various case studies, further expert interviews, and a survey to confirm the relevance and acceptance of Beyond Ba king (Ecosystem) for 1 e perspectives of banks and customers. This data will also help identify factors a criteria it at will serve as artefacts and other elements for the design evaluation.

The results offer guidance and classify different BBE levels while identifying key facts ts for it er development. Additionally, initial use cases for data recommender were created using DA and AI/ML methods. Finally, a initial blueprint for BBE development was designed, evaluated, and accepted mainly in an artifical 1 setting, serving as a foundation for future research in real-world applications. BB 1g decision n res now have an expanded, coupled decision design for BBE 3.0, which will max and inspires, a can be used to improve decision-m king on the implementation of BBE 3.0 initiatives.