HBP Product ID: ST138

t
UST138

os
JOON NAK CHOI
OHCHAN KWON
JOSEPH FERNANDEZ

rP
HSBC: Leveraging Data Analytics and AI to Enhance
Customer Life Cycle Management

yo
Our strategy centres on four key pillars: focus on our areas of strengths, digitise at scale to adapt
our operating model for the future, energise our organisation for growth, and support the
transition to a net zero global economy.
HSBC Holdings plc., Strategic Report 20221

To thrive in the AI-powered digital age, banks will need an AI-and-analytics capability stack that
op
delivers intelligent, personalized solutions and distinctive experiences at scale in real time.

McKinsey & Company, Global Banking Practice2

Having attended three consecutive meetings earlier in the morning, Amy Hui, Head of Personal
Banking and Customer Lifecycle Management, WPB HK, at HSBC in Hong Kong, was glad to have the
opportunity to return to the relative quiet of her office. She checked the newly released e-statement of her
tC

HSBC account while walking to her seat. Almost instinctively, she also noticed a new email from Jiahao
Teo, HSBC's Managing Director, Global Head of Data, Analytics & CRM, GPB & Wealth, North Asia
WPB. Even when not attending an in-person meeting, through her digital limbs – her laptop and mobile
phone – Amy remained in contact with her colleagues. Amy realized that a similar dynamic existed with
the millions of HSBC customers dispersed throughout Hong Kong. With the benefit of the data analytics
No

1 HSBC Holdings plc., Strategic Report 2022, https://www.hsbc.com/-/files/hsbc/investors/hsbc-results/2022/annual/pdfs/hsbc-

holdings-plc/230221-strategic-report-2022.pdf, accessed 3 November 2023.
2 Renny Thomas et al., “Building the AI Bank of the Future” McKinsey & Company: Global Banking Practice,” May 2021,

https://www.mckinsey.com/~/media/mckinsey/industries/financial%20services/our%20insights/building%20the%20ai%20bank
%20of%20the%20future/building-the-ai-bank-of-the-future.pdf, accessed October 2023.
Joseph Fernandez prepared this case under the supervision of Professor Joon Nak Choi and Professor Ohchan Kwon solely as a
basis for class discussion. The authors have disguised certain data to protect confidentiality. Cases are written in the past tense;
this is not meant to imply that all practices, organizations, people, places or facts mentioned in the case no longer occur, exist, or
apply. Cases are not intended to serve as endorsements, sources of primary data, or illustration of effective or ineffective
Do

handling of a business situation.

For inquiries regarding ordering and permission to reproduce the case and its materials, please write to [email protected] or visit
cbcs.ust.hk.

© 2024 by The Hong Kong University of Science and Technology. This publication shall not be digitized, photocopied or
otherwise reproduced, posted, or transmitted without the permission of the Hong Kong University of Science and Technology.

Last edited: 11 March 2024

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

team’s oversight and billions of dollars in investment, HSBC had built up an impressive data collection and

t
analytics machinery over the past several years. Jiahao’s email to Amy was direct and pointed. What were

os
the pain points she was experiencing in implementing HSBC’s relatively new Customer Life Cycle
Management (CLCM) strategy? In what ways could HSBC’s significant data-related resources be leveraged
to support her work?

While Jiahao’s query was straightforward, Amy realized that the appropriate response was much less
so. HSBC had an established digital infrastructure that operated under internal policies and external

rP
regulations. Ultimately, its data resources had to be utilized to create more meaningful, personal interactions
with each customer throughout their journey with HSBC, from the opening of a new account as a teenager
through to their retirement years. Amy responded to Jiahao’s email with a simple note, informing him that
she needed to consult with her team and would get back to him. She then copied and pasted Jiahao’s query
in a Group Chat message to her team with the added comment, ‘Let’s discuss.’

HSBC: A Dominant Presence in Hong Kong

yo
Hong Kong had a longstanding reputation as a global financial center. According to the Hong Kong
Monetary Authority (HKMA), seventy of the one hundred largest banks in the world had a presence in
Hong Kong, with twenty-nine multinational banks maintaining regional headquarters in Hong Kong. 3
These various financial institutions maintained over 1200 branches throughout the territory.4 Among all
these institutions, four accounted for the majority of both deposits and credit facilities including loans: The
Hongkong and Shanghai Banking Corporation Limited (HSBC), Bank of China (Hong Kong) Limited,
op
Standard Chartered PLC and Hang Seng Bank (a principal member of the HSBC Group) [see Exhibit 1].5
HSBC was the largest of these institutions in Hong Kong, however, across a range of criteria; out of 7.5
million Hong Kong residents, HSBC had a customer base of over 5.5 million that ranged from children to
their grandparents and professional investors to young couples saving to purchase their first home. With its
headquarters in Hong Kong, HSBC HK Ltd. generated a majority of its profits in Asia, with Hong Kong
being its single most profitable market globally [see Exhibit 2].6
tC

From Digitalization to Digital Transformation

Recognizing the importance of further transforming its digital ecosystems, HSBC invested an estimated
US$2.3 billion in 2015-2017 to expand its digital banking platforms and incorporate AI, facial recognition,
chatbots and more.7 An important initiative that further strengthened its digital capabilities was to break
No

3 Hong Kong Monetary Authority, Hong Kong, https://www.hkma.gov.hk/eng/key-functions/banking/, accessed 30 October 30

2023
4 Hong Kong Financial Services and Treasury Bureau, Hong Kong, The Facts, Financial Services, July 2020,

https://www.gov.hk/en/about/abouthk/factsheets/docs/financial_services.pdf, accessed 5 November 2023

5 Benjamin Quinlan and Eashan Trehan, “Branching Off: The Outlook for Hong Kong’s Virtual Banks,” Quinlan & Associates,

March 2021, https://www.gov.hk/en/about/abouthk/factsheets/docs/financial_services.pdf, accessed April 2021.

6 The Hongkong and Shanghai Banking Corporation Limited is a subsidiary of HSBC Holdings plc, which has several headquarters

around the world, including London and Hong Kong. For financial results, see HSBC, Annual Results 2022,
Do

https://www.hsbc.com/investors/results-and-announcements/all-reporting/annual-results-2022-quick-read, accessed 3
November 2023.
7 Sam Van Horbeek and Annie Chan, “Virtual Banking and Digital Banking: Comparing Digital Disruptions Across the World”

Wavestone, 3 December 2018, https://www.wavestone.com/en/insight/virtual-banking-open-banking-comparing-digital-

disruptions-world/, accessed 4 November 2023.

ST138 2
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
down internal data silos, which then allowed HSBC to develop more robust digital footprints of its
customers across a range of banking services.8

os
With millions of customers in Hong Kong, HSBC had a ready supply of enormous amounts of data to
work with. Machine Learning had been adopted internally since 2017 with the ongoing development of
hundreds of data models to extract various forms of granular data. The data was obtained from a variety of
daily interactions with its millions of customers, including in-person visits to branches, calls to HSBC
Contact Centres, ATM and card transactions, account activities, internet banking and the interactions

rP
between HSBC and the approximately two million customers accessing the HSBC HK App each month.
The latter two forms of engagement also provided a valuable digital footprint of customers’ behavior, such
as browsing history, clicks, pages viewed, and products or services accessed. With no shortage of data, the
challenge for Jiahao was how to best curate and leverage the data.

Like any other company, HSBC had to consider what its strategy might be with respect to the usage of
data and what competitive advantage it might wish to realize. Accordingly, it might then determine what

yo
the best sources of internal and external data might be, put the infrastructure in place to manage the data
obtained and then determine the best means to communicate the insights gained from the data [see Exhibit
3].9 One challenge with data was the its quality and reliability, and this was especially true with external
data sources [see Exhibit 4].10 Nevertheless, for HSBC, external data was essential and comprised one of
the four data domains that it established. In addition to External Data, the data domains included
Organizational Data, Customer Data and Service/Account Data. The data was then channelled through its
Data Analytical Platform, which was strategically deployed on the cloud to leverage native AI capabilities
op
and computational resources in order to generate the reports and feed the dashboards that provided the
visual market insights [see Exhibit 5].11

HSBC had an additional impetus to accelerate its digital transformation when the HKMA granted
licenses to eight, digital-only virtual banks in Hong Kong in 2019. While this was consistent with a global
trend at the time, the granting of those licenses was perceived as potentially disruptive to the banking sector
tC

in Hong Kong.

Late in 2019, when many of the virtual banks in Hong Kong were still establishing a foothold in the
Hong Kong market, HSBC launched the first version of its instant account opening journey12 The HSBC
HK app was further enhanced with the addition of more than two hundred value-added features in 2022,
one of the most popular being Budget, a financial management and planning tool for individual customers
that was introduced in early 2022 after only nine months of development. The ultimate objective of the app
was to transform the ways and means in which customers interacted with the bank, with the app accessible
No

24 hours per day, 7 days a week (24/7), and a resource at customers’ fingertips to quickly obtain answers
on personal banking matters or input on financial planning issues. HSBC’s app proved to be highly
successful, with more than 2 million app users every month (one out of three people in Hong Kong), with

8 HSBC, Communication with Case Writers, November 2023.

9 Keystone Strategy as referenced by Iansiti, Marco, and Karim R. Lakhani. Competing in the Age of AI: Strategy and Leadership
When Algorithms and Networks Run the World. Boston: Harvard Business Review Press, 2020.
10 Forrester Consulting, “Digital is Driving the next Generation of Data Marketplaces,” commissioned study on behalf of Pitney
Do

Bowes, 2017, https://www.pitneybowes.com/content/dam/pitneybowes/us/en/white-papers/pitney-bowes-forrester-data-us.pdf,

accessed 8 November 2023
11 HSBC, Communication with Case Writers, November 2023.
12 “HSBC Launches 24/7 Instant Mobile Account Opening Service,” press release HKMA website, 5 December 2019,
https://www.hkma.gov.hk/media/eng/doc/other-information/ac-opening/20200113e1.pdf, accessed 4 November 2023.

ST138 3
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

an eighteen percent Year-on-Year (YoY) growth in users. The app’s popularity enabled HSBC to glean

t
significant insights into its customers, such as their preferred mobile banking features and most common

os
mobile banking activities, and then translate those insights into more personalized services [see Exhibit 6].
HSBC’s app was part of a digital transformation that had continued on a global scale, with each year seeing
an increase in digital spend and a subsequent increase in users accessing HSBC’s banking services via
digital means [see Exhibits 7A & 7B].13

Customer Life Cycle Management

rP
While HSBC retained a dominant market position in Hong Kong, it nevertheless experienced
unprecedented and potentially disruptive changes in the ways in which it interacted with its customers. The
Covid-19 pandemic spurred more remote interactions, and many of the technology driven changes
continued to accelerate even after the pandemic given smarter phones, faster 5G mobile networks, and most
importantly – a digital native population accustomed to instantaneous access to information, on-demand
entertainment, viral social media influences and multi-faceted ways to connect locally, regionally, and even

yo
globally. Furthermore, innovative fintech companies continued to roll out convenient services, putting
pressure on HSBC to keep pace. The flip side of a digital native population in Hong Kong was an ageing
population with approximately twenty percent of the Hong Kong population aged sixty-five and over [see
Exhibit 8]. Given these dynamics in the Hong Kong market, it would be prudent for HSBC to adapt the
ways in which it interacted with its customers throughout their life journey with the bank.

Responding to these trends, HSBC launched a comprehensive Customer Life Cycle Management
op
(CLCM) strategy in 2022 [see Exhibit 9].14 CLCM involved four distinct stages: Acquisition, Activation,
Relationship Deepening and Retention. On the one hand, it involved a more personalized engagement with
its customers; on the other, it aimed to leverage its considerable digital resources as well as data analytic
and data science capabilities, and to adapt the ways in which it interacted with its customers at each stage
of their life journey with the bank. A core element of the CLCM strategy was the HSBC HK App, which
had proved to be popular with all age groups.
tC

Acquisition, the first phase of CLCM, aimed at acquiring customers with a quick (5-minutes for local
HK ID holders; 10-minutes for international passport holders) 15 account opening journey and instant
formulation of a customer’s profile upon completion. Activation, the second phase, involved a more
extensive, automated and personalised 90-day on-boarding engagement during which, through targeted
messages, new customers were guided in activating and using their different bank accounts, as well as their
credit and debit cards while also being educated about HSBC’s mobile banking apps and other value-added
No

services, such as Budget. Relationship Deepening, the third phase, was aimed at solidifying the customer
relationship by fully leveraging data analytics in supporting the customer to establish and reach their
financial goals [see Exhibit 10]. 16 Retention, the final phase of CLCM, sought to improve customer
satisfaction and to proactively detect issues that could lead to decreasing account activity, dormancy or
even account closure. Greatly enhancing customer service delivery was an award-winning digital tool
developed by Jiahao’s team called Automated Quality Management. It was recognized by HKMA for its

HSBC Holdings plc 4Q22 Results – Presentation to Investors and Analysts, https://www.hsbc.com/-/files/hsbc/investors/hsbc-
Do

13

results/2022/annual/pdfs/hsbc-holdings-plc/230228-annual-results-2022-presentation-to-investors-and-analysts.pdf, accessed 3
November 2023
14 HSBC, Communication with Case Writers, November 2023
15 HSBC HK, Mobile Account Opening, https://www.hsbc.com.hk/ways-to-bank/mobile-apps/banking/account-opening/, accessed

February 11, 2024

16 HSBC, Communication with Case Writers, November 2023

ST138 4
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
ground-breaking use of AI for financial services, and a unique voice-to-text engine capable of supporting
both Cantonese and English conversations by utilizing a Natural Language Understanding solution. This

os
24/7 quality assurance system also captured troublesome customer interactions and suggested human
interventions when needed to resolve customer issues, together with being a training resource to further
improve future customer interactions.

While CLCM was developed to serve all customer segments, there was a strategic customer segment
that Amy was keen to target: Millennials, which HSBC defined as those between the ages of eighteen and

rP
thirty-four. Millennials represented the future of the bank and were usually HSBC One customers who
enjoyed a barrier-free, comprehensive banking service. With the benefit of continuous relationship
deepening, they could eventually become Premier customers once their accumulated assets with HSBC
reached HK$1 million, providing ongoing and increasing revenue streams to HSBC. With the virtual banks
poised to compete with HSBC for Millennials, recruiting and retaining them became a key strategic
initiative for the bank.

yo
Targeting Millennials

For HSBC, millennials included university students just embarking on their journey toward financial
independence, young professionals ready to experiment with their first investments, and more mature
millennials ready to get married or starting their own businesses with the benefit of the skills gained from
their prior work experience. 17 While not necessarily a homogenous group, they shared several
characteristics. They were more likely to be tech savvy, with the youngest being true digital natives; eighty
op
percent of Hong Kong millennials preferred to manage many aspects of their lives via an app, while seventy-
eight percent concurred that a considerable amount of their time was lived online or on their phones.18 Even
though over 80% of millennials in Hong Kong were HSBC retail banking customers, they tended not to be
loyal to established brands, and were more likely to switch financial service providers.19 So, one practical
reason to target millennials was that HSBC could fully leverage its extensive data, analytics and AI
resources with this ‘mobile first’ market segment while implementing its CLCM strategy in a way that
increased the ‘stickiness’ of its existing millennial customers.
tC

Given millennials’ importance to HSBC’s long-term viability, an essential question that Amy and her
team had was to what extent might HSBC’s CLCM strategy need to be modified or enhanced to best serve
millennials in Hong Kong. What financial guidance, for example, would best serve millennials during the
Activation phase? Since they were known to be less ‘sticky,’ in what ways could HSBC better engage with
millennials for the longer-term? The more she considered these questions, Amy realized that while they
could determine the age, occupation, and income level of their millennial customers, she and her team
No

needed to think carefully about the resources that Jiahao had access to that might help to answer some of
the questions her team had and help them to develop and deliver the products and services that best met the
banking needs of millennials.

Segmenting the Market

Do

17 Ibid.
18 Brian Hui and Andrew Gilder, Tomorrow’s Customers: Reinventing Banking for Young Millennials, 2021,
https://mvocontent.s3.us-east-
2.amazonaws.com/uploads/post/document/13128/Reinventing_banking_for_young_millennials_ey_hsbc.pdf?utm_source=substa
ck&utm_medium=email , accessed 30 October 2023
19 Ibid.

ST138 5
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
Market segmentation, in which customers were separated into more homogeneous groups with similar
needs and preferences, was typically the first step of a broader strategic exercise. Market segmentation

os
enabled companies to understand customers at a more granular level, enabling them to select the best
segments to target, and position themselves better within each of these targeted segments; see Exhibit 11A
for a summary. Amy’s first challenge would be to improve how HSBC segmented millennials by leveraging
the bank’s treasure trove of data.

Traditionally, market segmentation proceeded in a top-down manner with key demographic,

rP
sociographic, or geographic filtering based on age, income, education level or geographic location. Banks
often focused on income or wealth or sought to segment markets based on specific products or services
they wished to market, such as a credit card or an insurance product. See Exhibit 11B for an example of
market segmentation based on age and income.20 In general, top-down market segmentation provided a
high-level overview of a market that could easily be understood by humans. Yet, within any such market
segment, considerable variance remained in individual customers’ buying behavior or preferences.

yo
Given HSBC’s significant data infrastructure and significant data capture of customer behavior in Hong
Kong, a bottom-up market segmentation of millennials was also possible. Bottom-up market segmentation
provided much more granular data, often obtained directly from end-customers who provided insights into
their preferences and behaviors [see Exhibit 11C].21 With sufficient volumes of data, HSBC could then
organize the data and extract the most homogeneous customer segments based on the data collected. Unlike
top-down segmentation, which could easily be performed using a filtering tool (e.g., relational databases or
even Microsoft Excel), bottom-up segmentation was conducted algorithmically by applying a clustering
op
algorithm. This approach could produce fine-grained results, with greater predictive power than top-down
segmentation because two individuals sharing the same demographics (e.g., male, middle-aged, employed)
and geographic locations (e.g, living on Hong Kong Island) could have different needs, habits, and
consumption patterns; bottom-up segmentation, on the other hand, would notice hidden behavioral
differences.22 However, bottom-up segmentation had a significant downside. Since it was produced by
machine learning (i.e., clustering) algorithms that identified prominent patterns in data without considering
tC

how interpretable it would be to humans, bottom-up segmentation tended to be complicated and difficult
for humans to make sense of. This would not be a problem, however, if bottom-up segmentation were to
be used mainly by machines (i.e., artificial intelligence) rather than human employees.

Amy’s team could choose from several different clustering algorithms to conduct bottom-up market
segmentation. By far the most popular clustering algorithm was k-means clustering, which was appropriate
for grouping individuals according to continuous variables (i.e., variables that include any number within
some range, such as 1 to 100). However, k-means clustering by itself could not incorporate discrete
No

variables (i.e., 0’s versus 1’s) or categorical variables (e.g., multiple choice responses). Other clustering
algorithms (e.g., hierarchical clustering) were less commonly used but potentially produced more robust
results. Amy’s team needed to avoid a common tendency among practitioners: to use segmentation

20 Miklós Dietz, Jay Datesh, Attila Kincses and Carlos Gonzalez Valverde, Charting Retail Banking Revenues by Revenues,
Panorama Financial Institutions and Insights Consulting (PFIC), McKinsey & Co., 21 December 21 2020,
https://www.mckinsey.com/industries/financial-services/our-insights/charting-retail-banking-revenues-by-generation, accessed
16 November 2023.
Do

21 Joon Nak Choi (2023), “Management Consulting: Segmentation,” PowerPoint presentation, 29 November, 2023, HKUST

Business School, Hong Kong.

22 Michael Wade, “Psychographics: the behavioural analysis that helped Cambridge Analytica know voters’ minds”. The

Conversation, 20 March 2018, https://theconversation.com/psychographics-the-behavioural-analysis-that-helped-cambridge-

analytica-know-voters-minds-93675, accessed 11 November 2023.

ST138 6
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

algorithms as “black boxes” without really understanding how they worked. Many practitioners across

t
various industries, especially market research, tended to misuse segmentation algorithms such as k-means

os
clustering because they did not think carefully about what they were really trying to do, what input data
might be appropriate or inappropriate, or the trade-offs between different algorithms to obtain robust and
valid results.23

Ultimately, HSBC’s goal was to segment diverse millennials into more homogeneous segments (i.e.,
subgroups) sharing similar needs and preferences. Once this was done, HSBC could target the right

rP
segments with the right products and offers to personalize their banking experience. HSBC had traditionally
utilized humans using advanced statistical analyses (i.e., machine learning) to match segments with
products and offers. However, Amy’s team now had a new tool at their disposal: artificial intelligence (AI)
that could do this automatically with minimal human input.

Artificial Intelligence (AI)

yo
While AI captivated many after the launch of ChatGPT in late 2022, banks had already been using AI
long before then. HSBC had been an early adopter of AI, launching its chatbot in 2018 and a cloud-based
financial crime detection tool in 2023. However, Amy saw many opportunities for improvement.

AI Use Across Banking Value Chains

Banks such as HSBC recognized how AI could improve many of their operations. In general, banking
op
operations involved handling large volumes of data in a time-sensitive manner. On one hand, this played to
AI’s core strength, its practically unlimited bandwidth; it could handle billions of simultaneous requests
nearly instantaneously. On the other hand, this addressed one of AI’s key weaknesses, the need to be trained
upon large amounts of data.

Management consultancy Deloitte emphasized three areas in which banks could best leverage AI:
tC

 Personalization and customer recommendation: customized insights enabled personalized

communication with customers and provided them with advice, offers and services.

 Productivity: labor-intensive work, like compliance reporting, new customer on-boarding, and
documentation, accounted for 20% of a banks’ back-office work, but 85% of the cost. AI had the
potential to automate these processes at little cost.
No

 Fraud detection and compliance: large-scale anomaly detection enabled AI to identify fraud, one
of the biggest challenges for banks given their massive transaction volumes.

23 For instance, many practitioners used k-means clustering with categorical data (i.e., mutually exclusive answer choices in
Do

multiple choice survey questions), converting categories into dichotomous variables (most often using a technique called “one hot
encoding”). The resulting dichotomous variables could be input into k-means clustering, making it mathematically possible to run
this algorithm. However, this approach violated the statistical assumptions underlying k-means clustering, resulting in poor results.
See the following discussion on Stack Exchange: https://datascience.stackexchange.com/questions/104299/one-hot-encoded-
variables-dominating-clustering.

ST138 7
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
These observations matched the priorities of Jiahao, who believed that the biggest opportunities for his
bank were to use AI to improve its customers’ experience via hyper-personalization, reduce operational

os
expenditures via automation, and reduce fraud.24

Exhibit 12A summarizes how banks and other financial institutions might leverage AI at various stages
of their value chains. Exhibit 12B provides some notable examples of major players in retail finance
implementing AI. Senior banking executives saw much potential for value creation across each of these
stages.

rP
AI use across this value chain was expected to increase in the coming years. By some estimates, the
potential value of AI for banking was as high as US$1 trillion.25 This included the potential to increase
revenues by creating more personalized banking experiences for their customers; lowering costs while
increasing employee satisfaction by reducing mundane, repetitive tasks, decreasing errors and better
allocating human and other resources; realizing new opportunities by processing higher volumes of
customer and market data that could lead to deeper market insights.26

yo
HSBC and Chatbots

Focusing on customer-facing opportunities, chatbots were one of the most common ways that AI was
used in the banking sector as of 2023. The earliest chatbots simply repeated hard-coded responses to human
inquiries. 27 With the rise of generative AI, however, chatbots became more sophisticated. AI-powered
chatbots gained the ability to learn from previous conversations, being "trained” on proper and improper
op
behaviors by humans. When the chatbot made a mistake, a human could step in and notify the chatbot of
the error. The chatbot would then correct its behaviors accordingly. Gaurav Singh, founder of Verloop,
which is a leading conversational AI operating system for business and a WhatsApp partner, identified five
keys to training chatbots:

 Recognizing the most frequently asked questions: paying attention to repeatedly asked questions
tC

and training the chatbot on the correct answers.

 Recognizing variations of these frequently asked questions.

 Understanding the customer’s intent to generate the appropriate responses.

 Evaluating underperforming answers: human monitoring of chatbot-customer interactions to

No

identify which answers are underperforming so that the chatbot can be retrained.

 Not relying solely on text: images, emojis, and punctuation provide details beyond what could be
learned from words only.

24 HSBC, Communication with Case Writers, November 2023.

25 Suparna Biswas, Brant Carson, Violet Chung, Shwaitang Singh and Renny Thomas, “AI Bank of the Future: Can Banks Meet
the AI Challenge,” McKinsey & Company, Global Banking and Securities, September 2020,
Do

https://www.mckinsey.com/industries/financial-services/our-insights/ai-bank-of-the-future-can-banks-meet-the-ai-challenge,
accessed 31 October 2023
26 Ibid.
27 Caravan Coop, “The first ‘successful’ chatterbot from 1966”, Medium.com, 18 January 2017, https://medium.com/caravan-

blog/the-first-successful-chatterbot-from-1966-17088c4c4669, accessed 22 November 2023.

ST138 8
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
See Exhibit 13 for more details.28

os
HSBC deployed its first AI-powered chatbot, also named Amy, for commercial banking in late 2018.
Soon afterwards, in 2019, it deployed a chatbot titled ‘Chat with us’ for retail banking customers. The
expectation was that a new generation of digital natives would prefer chatbots over more traditional
customer interaction channels, such as human representatives at a physical banking location. 29 Nathan
Pearson, Global Head of Conversational Banking for Wealth and Personal Banking, saw “…an explosion

rP
in the use of instant messaging in markets around the world” as customers made fewer phone calls and
switched to instant messaging. He viewed ‘Chat with us’ as strategically important because it enabled retail
customers to contact HSBC through the same modality to answer simple queries ranging from how to pay
credit card bills to instructions on how to set up mobile banking. For a customer, chatbots offered several
advantages over traditional ways of contacting HSBC. A customer would not have to journey to their
nearest branch and then wait in a queue for a customer service representative to answer a simple question.
For HSBC, chatbots enabled more efficient resource allocation. By empowering customers to service

yo
themselves through a chatbot, HSBC eliminated many time-consuming tasks for humans. The bank,
however, recognized that chatbots clearly could not answer every single question, and made sure human
agents were available to answer more complex or difficult questions. The key was that customers just
wanted their inquiries to be resolved efficiently and effectively. According to Pearson, HSBC’s customers
did not “…need our bank to be all cool and trendy, or too wacky and quirky, [they] just want[ed] to get the
task completed and get on with [their day].”30
op
HSBC’s chatbots quickly gained widespread adoption. By 2020, its chatbots were able to answer the
simplest questions immediately, handling up to 51% of incoming chat volume.31 Kevin Martin, HSBC’s
Former Chief Operating Officer of Wealth and Personal Banking, stated that chatbots handled 10 million
chat conversations in 2019 and expected them to handle 10 million per month by 2024, including most
routine activities, such as checking balances; however, human agents were expected to continue handling
more complex transactions at branch locations.32
tC

While several chatbots like Amy and ‘Chat with us’ have been deployed by various financial
institutions and continue to operate successfully, not everyone has been satisfied with them. While in the
U.S., about one in three persons had interacted with a chatbot in 2022, with the number expected to increase
in subsequent years, according to the U.S. Consumer Financial Protection Bureau (CFPB), many of those
interactions resulted in frustrated encounters that diminished trust in the financial institution’s services and
could potentially be in violation of the law.
No

Beyond the Chatbot: Artificial Intelligence and Customer Experience

28 Gaurav Singh, “AI Training: 5 Easy Ways to Train a Chatbot,” Verloop.io, 15 June 2022, https://verloop.io/blog/train-a-chatbot/,
accessed 22 November 2023.
29 Asian Banking and Finance, “Hong Kong Banks Bust Out the Chatbots”, https://asianbankingandfinance.net/banking-

technology/in-focus/hong-kong-banks-bust-out-chatbots, accessed 22 November 2023

30 HSBC, “Why it’s good to chat”, 18 September 2020, https://www.hsbc.com/news-and-views/news/hsbc-news-archive/why-it-
Do

is-good-to-chat, accessed 28 November 2023.

31 Charlie Nunn, “HSBC Wealth and Personal Banking,” June 2020, https://www.hsbc.com/-/files/hsbc/investors/investing-in-

hsbc/investor-events-and-presentations/2020/200611-hsbc-wpb-at-goldman-sachs.pdf, accessed 28 November 2023.

32 Kevin Martin, “How banking will change after COVID-19,” HSBC, 26 November 2020, https://www.hsbc.com/news-and-

views/views/hsbc-views/how-banking-will-change-after-covid-19, accessed 28 November 2023.

ST138 9
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
Going beyond the chatbot, the opportunities AI created for banks was widely recognized by the banking
sector in Hong Kong, with the vast majority of banks expected to increase their AI related investments over

os
the next five years. 33 One particular area where there was strong potential for banks to leverage AI
applications and increase business value was in their relationships with customers. In particular, AI
provided new and innovative ways to engage with customers and personalize services at scale, as well as
providing more varied and enhanced omnichannel experiences.34 The challenge, though, was that bank
customers varied greatly in terms of age, level of education, financial goals and other variables. To leverage
AI properly, therefore, it was necessary to understand a customer’s needs at any particular time, their

rP
behaviors, their life stage and their preferences [see Exhibit 14].35

Notwithstanding the excitement regarding these opportunities, (the human) Amy, Jiahao, and other
senior executives at HSBC recognized that AI raised serious privacy, security, and other data governance
concerns. AI heightened typical concerns regarding cybersecurity and data privacy, with the greater
complexity of AI-enabled information technology systems potentially opening the door for malicious actors
to access private customer data. These issues were particularly challenging since data privacy laws in many

yo
places had recently been strengthened with the European Union’s adoption of Global Data Protection
Regulation (GDPR) in May 2018, being one of the most comprehensive data privacy laws protecting
individuals.

AI also raised new, unique concerns for bank executives. For instance, generative AI models could
learn from interactions with humans, raising the possibility that they would learn company secrets from
humans. Those secrets would then be available on publicly-accessible generative AI platforms and easily
op
obtained by unauthorized parties. For instance, Samsung employees shared company secrets with OpenAI's
ChatGPT platform on three separate occasions. One employee entered non-functional source code into
ChatGPT for it to find and fix the bug. Another uploaded functional code for optimization and a third
uploaded a recording of a meeting to generate meeting minutes. After these incidents, Samsung banned
ChatGPT usage. 36 Based on such incidents, Italy briefly banned generative AI and many high-profile
individuals expressed their concerns and even lobbied to halt further AI developments. High-level meetings
tC

were held in the U.K., chaired by the U.K. Prime Minister, to discuss relevant matters and the U.S. President
issued an Executive Order with several new requirements for future AI developments, including a
requirement for standards, tools and tests for AI solutions and the establishment of authentication standards
to address deception and fraud with AI generated content.

In August 2019, HKMA conducted a survey of the banking industry in Hong Kong. The survey results
confirmed that Data Quality/Privacy was the top concern [see Exhibit 15].37 Based on the results of that
survey, in November 2019, HKMA established guiding principles on consumer protection as it relates to
No

33 Ibid.
34 Suparna Biswas, Brant Carson, Violet Chung, Shwaitang Singh and Renny Thomas, “AI Bank of the Future: Can Banks Meet
the AI Challenge,” McKinsey & Company, Global Banking and Securities, September 2020. (Please see my note on Footnote 25.)
35 Violet Chung, Malcolm Gomes, Sailee Ranee, Shwaitang Singh and Renny Thomas, “Reimagining Customer Engagement for

the AI Bank of the Future,” McKinsey & Company, Global Banking and Securities, 13 October 2020,
https://www.mckinsey.com/industries/financial-services/our-insights/reimagining-customer-engagement-for-the-ai-bank-of-the-
future, accessed 31 October 2023.
Do

36 Lindsey Wilkinson, “Samsung employees leaked corporate data in ChatGPT: report”, CIODive, 7 April 2023,

https://www.ciodive.com/news/Samsung-Electronics-ChatGPT-leak-data-privacy/647137/, accessed 22 November 2023

37 Hong Kong Institute for Monetary and Financial Research (HKIMR), “Artificial Intelligence in Banking: The Changing

Landscape in Compliance and Supervision,” HKIMR Applied Research Report No. 2/2020, 21August 2020,
https://www.aof.org.hk/docs/default-source/hkimr/applied-research-report/airep.pdf, accessed 1 November 2023

ST138 10
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
Big Data Analytics and Artificial Intelligence (BDAI).38 Beyond data privacy and security, in the same
HKMA survey, the banking industry also expressed several other concerns with AI, in general, most of

os
which revolved around liabilities that may result from AI applications. These included operational risks,
reputation risks, strategic risks and legal risks.39

HSBC’s policy regarding generative AI blended optimism and caution, as communicated below by
Luanne Lim, CEO of HSBC Hong Kong during Hong Kong’s Fintech Week:

rP
“AI has been around for a long time now. What has changed in the last couple of years is the emergence
of generative AI… it’s going to be like you’re talking to a human; it’s going to be very personalised. It
will be a game-changer one day in terms of how we interact with our customers…”

However, Lim added that such technology will not fully replace in-person services, and it will also need
to be overseen by proper governance and risk management, with the necessary tooling. “There’s a
fiduciary duty here, and that lies with the bank.”40

yo
Conclusion
HSBC was always dedicated to delivering the best to their customers and Customer Lifecycle
Management was a powerful methodology to raise the bar on customer engagement. By harnessing the
power of AI and data analytics, HSBC supported customers, especially millennials, to rethink their future
and achieve their personal financial goals. Hyper-personalised customer experience was HSBC’s ambition
op
and Jiahao’s team was a valued internal resource to make it happen. Amy fully understood that HSBC
planned to target the most digitally savvy customers the bank had ever served, and that the CLCM approach
would be key to this endeavor. While she noted HSBC’s intent to continue leading AI adoption within the
retail banking industry, she also understood that privacy, security and safety were fiduciary requirements
that HSBC took seriously. She also understood that AI’s emergence was accelerating at a pace that was
alarming to many. When Amy shared her thoughts with Jiahao, he agreed, stating that ‘to support HSBC’s
tC

position at the forefront of deploying AI to capture the benefits for customers and the business, my team
has implemented enhanced AI governance mechanisms to ensure structured risk and privacy assessments
which will support our business in accelerating the realization of AI benefits in a safe manner.’

AI was already here to stay and was expected to be a transformative technology. Moreover, technology
was known to be a major disruptive force in many industries. So, while HSBC had considerable confidence
in their primary market in Hong Kong, Amy and Jiahao understood that to maintain their market position
No

with the millennials they were targeting, they had to fully understand this market segment now, and would
have to continue to leverage every technology resource they could to effectively serve millennials over their
lifetime.
Do

38 Ibid.
39 Ibid.
40 Reggie Ho and Morning Studio Editors, “New generation of fintech and banks ‘a great marriage’, HSBC Hong Kong chief

says,” South China Morning Post, 2 November 2023, https://www.scmp.com/presented/business/banking-finance/topics/hsbc-

hong-kong-ceo-talks-fintech/article/3239935/new-generation-fintech-and-banks-great-marriage-hsbc-hong-kong-chief-says,
accessed 22 November 2023.
ST138 11
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 1 – LARGEST BANKS IN HONG KONG

os
rP
yo
op
tC

Source: Benjamin Quinlan and Eashan Trehan, “Branching Off: The Outlook for Hong Kong’s Virtual
Banks,” Quinlan & Associates, March 2021,
https://www.gov.hk/en/about/abouthk/factsheets/docs/financial_services.pdf, accessed April 2021.
No
Do

ST138 12
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 2 – HSBC’S PROFITS BREAKDOWN BY REGION

os
rP
yo
op
Source: HSBC, Annual Results-2022, https://www.hsbc.com/investors/results-and-announcements/all-
tC

reporting/annual-results-2022-quick-read, accessed 3 November 2023

No
Do

ST138 13
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 3 - DATA INFRASTRUCTURE

os
rP
yo
op
tC

Source: Keystone Strategy as referenced by Iansiti, Marco, and Karim R. Lakhani. Competing in the Age
of AI: Strategy and Leadership When Algorithms and Networks Run the World. Boston: Harvard
Business Review Press, 2020.
No
Do

ST138 14
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 4 – OPPORTUNITES & CHALLENGES WITH EXTERNAL DATA

os
rP
yo
op
tC
No

Source:
Do

Forrester Consulting, “Digital is Driving the next Generation of Data Marketplaces,” commissioned study
on behalf of Pitney Bowes, 2017, https://www.pitneybowes.com/content/dam/pitneybowes/us/en/white-
papers/pitney-bowes-forrester-data-us.pdf, accessed 8 November 2023

ST138 15
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
os
EXHIBIT 5 – HSBC’S DATA MACHINERY

rP
yo
op
tC

Source: HSBC, Communication with Case Writers, November 2023

No
Do

ST138 16
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 6 – HSBC’S HK APP

os
rP
yo
op
Source: HSBC, Communication with Case Writers, November 2023
tC
No
Do

ST138 17
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 7A – HSBC: TECHNOLOGY SPEND

os
rP
yo
op
EXHIBIT 7B – HSBC: GROWTH IN DIGITAL ENGAGEMENT
tC
No
Do

SOURCE: HSBC Holdings plc 4Q22 Results – Presentation to Investors and Analysts,
https://www.hsbc.com/-/files/hsbc/investors/hsbc-results/2022/annual/pdfs/hsbc-holdings-plc/230228-
annual-results-2022-presentation-to-investors-and-analysts.pdf, accessed 3 November 2023
ST138 18
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 8 – HONG KONG’S AGING POPULATION

os
rP
yo
op
tC

Source: Snapshot of the Hong Kong Population: Older Persons, 2021 Population Census, Census and
Statistics Department, HKSAR,
https://www.census2021.gov.hk/doc/pub/21C_Articles_Older_Persons.pdf, accessed 21 November 2023
No
Do

ST138 19
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 9 – HSBC: CUSTOMER LIFE CYCLE MANAGEMENT

os
rP
yo
Source: HSBC, Communication with Case Writers, November 2023
op
EXHIBIT 10 – CLCM: RELATIONSHIP DEEPENING
tC
No
Do

Source: HSBC, Communication with Case Writers, November 2023

ST138 20
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 11A – SEGMENTATION-BASED STRATEGY

os
rP
yo
Source: Adapted by the authors from Gupta, “Segmentation and Targeting”. Boston, MA: Harvard
op
Business School Press. 2019.
tC
No
Do

ST138 21
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 11B – TOP-DOWN MARKET SEGMENTATION BY AGE AND INCOME

os
rP
yo
op
tC
No

Source: Miklós Dietz, Jay Datesh, Attila Kincses and Carlos Gonzalez Valverde, Charting Retail
Banking Revenues by Revenues, Panorama Financial Institutions and Insights Consulting (PFIC),
McKinsey & Co., 21 December 21 2020, https://www.mckinsey.com/industries/financial-services/our-
insights/charting-retail-banking-revenues-by-generation, accessed 16 November 2023.
Do

ST138 22
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 11C – BOTTOM-UP MARKET SEGMENTATION

os
Two approaches to market segmentation
Traditional top-down approach vs. algorithmic bottom-up approach

rP
Top • Traditional approach in marketing strategy
• Focus on geographic & demographic variables
Down • Easy to conceptualize and implement

Bottom
yo Segments that fit criteria described above!
op
• Algorithm-driven approach in market research
• Focus on empirics, follow whatever algo suggests
up • Difficult to make sense of the results at times

18
tC

Source: Adapted by the authors from Gupta, “Segmentation and Targeting”. Boston, MA:Harvard
Business School Press. 2019.
No
Do

ST138 23
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 12A – AI USE CASES ACROSS THE BANKING VALUE CHAIN

os
rP
yo
op
Source: Deloitte, “Artificial Intelligence in Banks”.
https://www2.deloitte.com/content/dam/Deloitte/ca/Documents/audit/ca-audit-abm-scotia-ai-in-
banking.pdf. Accessed 22 November 2023.
tC

EXHIBIT 12B – AI DEPLOYMENTS WITHIN THE BANKING SECTOR

No
Do

Source: Deloitte, “Artificial Intelligence in Banks”.

https://www2.deloitte.com/content/dam/Deloitte/ca/Documents/audit/ca-audit-abm-scotia-ai-in-
banking.pdf. Accessed 22 November 2023.

ST138 24
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 13 – CHATBOT TRAINING

os
rP
yo
Source: Gaurav Singh, “AI Training: 5 Easy Ways to Train a Chatbot,” Verloop.io, 15 June 2022,
https://verloop.io/blog/train-a-chatbot/, accessed 22 November 2023.
op
tC
No
Do

ST138 25
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 14 – AI AND CUSTOMER INTERACTIONS

os
rP
yo
op
tC
No

Source: Violet Chung, Malcolm Gomes, Sailee Ranee, Shwaitang Singh and Renny Thomas,
“Reimagining Customer Engagement for the AI Bank of the Future,” McKinsey & Company, Global
Banking and Securities, 13 October 2020, https://www.mckinsey.com/industries/financial-services/our-
insights/reimagining-customer-engagement-for-the-ai-bank-of-the-future, accessed 31 October 2023
Do

ST138 26
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 HKUST Business School Thompson Center for Business Case Studies

t
EXHIBIT 15 – AI GOVERNANCE RISKS PERCEIVED BY RETAIL BANKS

os
rP
yo
op
tC

Source: Hong Kong Institute for Monetary and Financial Research (HKIMR), “Artificial Intelligence in
Banking: The Changing Landscape in Compliance and Supervision,” HKIMR Applied Research Report
No. 2/2020, 21August 2020, https://www.aof.org.hk/docs/default-source/hkimr/applied-research-
report/airep.pdf, accessed 1 November 2023
No
Do

ST138 27
UST138 HSBC: Leveraging Data Analytics and AI to Enhance Customer Life Cycle Management

This document is authorized for educator review use only by Diseno Instruccional ESI School of management, ESI - Guatemala until Jun 2025. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860