Path Analysis of Data Empowerment Driving the Upgrading of China

Consumer Market

Qingyang Hong

Krieger School of Arts & Sciences, Johns Hopkins University, 555 Pennsylvania Avenue, U.S.A.

Keywords: Consumption Upgrading, Digital Economy, AI-Driven, Personalization, Regional Disparities.

Abstract: China’s consumer market has exhibited a dual-track phenomenon since the post-pandemic era, characterized

by consumption downgrade and upgrade simultaneously. Cost-effective products dominate among middle

and low-income populations due to economic uncertainty and rising price sensitivity. In contrast, high-income

demographics continue to drive premium consumption in sectors like health, technology, and education.

Regional disparities also further fragment the market: Eastern China adopts selective downgrade strategies by

leveraging mature digital infrastructure for data-driven precision marketing, whereas central and western

regions face comprehensive expenditure declines due to weaker technological capabilities. This study

investigates how Tmall and Douyin exemplify their algorithmic innovation to withstand these complexities,

to drive consumption upgrades, and to sustain growth. Tmall commands 62.2% of China’s e-commerce

market with the support of its real-time behavioral analytics, dynamic ranking, and gamified engagement. It

employs AI-driven recommendation systems to optimize personalized product matching during mega-events

like Singles’ Day. Douyin integrates user profiling and interest-tagging algorithms to revolutionize live-

stream e-commerce and transform interactions into purchase intent, achieving $200 billion GMV by 2022.

Findings underscore the pivotal role of AI in reconciling China’s dual-track consumption trends, offering

insights into the synergies between technological innovation and evolving consumer behavior in a stratified

market.

1 INTRODUCTION

China’s consumer market has undergone profound

transformations and expansion over the last decade,

driven by rapid digitalization, economic

stratification, and evolving consumer preferences. As

one of the leading forces, e-commerce penetration

surpassed 40% in 2023, and platforms like Tmall and

Douyin are emerging as the game changers in

reshaping consumption patterns (Bain and Company,

2024). This growth is underpinned by integrating

advanced data technologies, such as AI-driven

recommendation systems and real-time behavioral

analytics, into retail ecosystems, enabling

unprecedented precision in understanding and

influencing consumer behavior. However, the market

exhibits a dual-track phenomenon: while premium

consumption persists in sectors like health and

technology, economic uncertainty and regional

disparities have intensified cost-consciousness,

https://orcid.org/0009-0008-3889-1490

particularly among middle- and low-income groups

(Han, 2024; Chen and Li, 2020). Eastern regions

prioritize high-quality and personalized consumption

due to their mature digital infrastructure. Meanwhile,

central and western areas face broader expenditure

declines, exacerbating market fragmentation

(Deloitte, 2023). To turn around this backdrop,

algorithmic innovation has become a critical weapon

for bridging these divides, optimizing resource

allocation, and sustaining growth in this increasingly

complex landscape.

Existing scholarship has extensively explored

China’s consumption downgrade-upgrade duality and

regional imbalances (Han, 2024; Chen and Li, 2020),

yet few studies systematically examine how digital

platforms harness algorithmic tools to navigate these

dynamics. Prior research emphasizes macroeconomic

drivers of stratification, such as income inequality

and savings rates, but overlooks the micro-level role

of data-driven technologies in mediating consumer

576

Hong, Q.

Path Analysis of Data Empowerment Driving the Upgrading of China Consumer Market.

DOI: 10.5220/0013849900004719

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on E-commerce and Modern Logistics (ICEML 2025), pages 576-581

ISBN: 978-989-758-775-7

choices (Bain and Company, 2024). Similarly, while

regional disparities are well-documented, the

intersection of digital infrastructure gaps and

algorithmic efficacy remains underexplored

(Deloitte, 2023). This study addresses these gaps by

analyzing how leading platforms leverage AI-

powered systems to reconcile market fragmentation,

stimulate demand, and drive consumption upgrades.

The significance of this research lies in its timely

examination of technology’s role in sustaining

China’s digital economy amid shifting consumer

priorities. Platforms compete to balance efficiency

with hyper-personalization. At the same time,

building in volatile markets, their algorithmic

strategies offer insights into resilience. For instance,

Tmall’s AI recommendation engines, deployed

during the 2024 Singles’ Day Festival, boosted sales

for brands like Midea by 22% through real-time

personalization and dynamic resource optimization,

while Douyin’s interest-tagging algorithms achieved

$200 billion GMV by aligning livestream content

with user preferences. These cases illustrate how

algorithmic precision transforms transient

interactions into sustained purchasing intent, moving

beyond transactional relationships to foster brand

loyalty and recurring engagement.

Methodologically, this paper employs a mixed-

methods approach, combining case studies of Tmall

and Douyin with quantitative analysis of sales data,

user engagement metrics, and regional consumption

trends. By dissecting Tmall’s AI-driven mega-event

strategies and Douyin’s short-video e-commerce

model, the study highlights the operational mechanics

of algorithmic systems, their impact on consumer

behavior, and their implications for market cohesion.

The findings indicate the need for scalable, adaptive

technologies to address heterogeneity and advocate

for enhanced data infrastructure in underserved

regions to mitigate regional, economic, or behavioral

disparities. Ultimately, this research contributes to the

broader discourse on algorithmic innovation’s

capacity to harmonize efficiency, personalization,

and inclusivity in China’s evolving digital economy.

2 ANALYSIS ON THE CURRENT

SITUATION OF CHINA

CONSUMER MARKET WITH

DATA EMPOWERMENT

In recent years, Chinese consumers have increasingly

prioritized cost-effective products, reflecting a

pronounced trend of consumption downgrade. This

shift is linked to economic uncertainty, rising

household savings rates, and heightened price

sensitivity among consumers (Han, 2024). Despite

the prominence of consumption downgrade,

consumption upgrade persists in certain sectors. For

instance, expenditures on health, education, and

technology continue to grow, underscoring

consumers' pursuit of high-quality lifestyles (Chen

and Li, 2020). This "dual-track phenomenon"

highlights the complexity of China's consumer

market. Market stratification will become more

evident. On one hand, high-income groups will

continue driving premium consumption; on the other,

middle- and low-income demographics will prioritize

value-for-money, potentially prolonging the

consumption downgrade trend (Sing, 2024).

Against the backdrop of consumption downgrade,

regional consumer markets exhibit distinct coping

strategies. In eastern China, consumers lean toward

selective downgrade-reducing spending in certain

categories while maintaining or increasing

consumption in others. In contrast, central and

western regions demonstrate comprehensive

downgrade, marked by an overall decline in

consumer expenditure (Bain and Company, 2024).

This divergence further accentuates regional market

fragmentation. Eastern coastal areas (e.g., the

Yangtze River Delta and Pearl River Delta) boast

higher economic development and household income

levels compared to central and western regions. These

markets are more mature, with consumers prioritizing

high-quality, personalized goods and services

(Deloitto, 2023).

The efficacy of data empowerment also varies

regionally. Eastern regions, supported by robust

digital infrastructure, have achieved notable success

in data-driven precision marketing and supply chain

optimization. Conversely, central and western

regions lag due to insufficient data collection and

application capabilities, leaving the full potential of

data empowerment untapped. Strengthening data

infrastructure in these areas could help bridge

regional disparities (Zhang and Zhan, 2023)

Mintel’s 2024 China Consumer Report likely

highlights evolving consumer preferences and

adaptive marketing strategies, noting that companies

leveraging big data to analyze purchase histories and

browsing patterns can deliver targeted advertisements

for products/services matching consumer interests,

thereby enhancing marketing efficacy.

Despite the prevalence of consumption

downgrade, consumption upgrade trends persist in

specific sectors. Data analytics allow businesses to

identify these dynamic shifts and formulate agile

Path Analysis of Data Empowerment Driving the Upgrading of China Consumer Market

577

market strategies. For example, by tracking rising

consumer interest in health and sustainability,

companies can introduce high value-added products

aligned with these emerging demands (Zipser et al.,

2024). Data empowerment further fosters cross-

industry data sharing and collaborative innovation. In

service-oriented manufacturing, for instance, data

sharing enhances supply chain management and

reduces operational costs. Such synergies are

particularly vital amid consumption downgrade, as

collaborative resource integration strengthens overall

competitiveness.

3 CASE STUDY

Personalized Recommendation Systems are

intelligent algorithms grounded in user historical

behavior, preferences, and interests, designed to

provide tailored product or service recommendations.

These systems analyze behavioral data (e.g.,

browsing duration, click patterns, purchase history)

using machine learning and big data technologies to

predict user preferences, thereby enhancing user

experience and commercial conversion rates. Since

2020, personalized recommendation systems have

been implemented through four key steps: data

collection, feature extraction, model training, and

real-time prediction, enabling efficient and precise

recommendations. Both Tmall and Douyin use

personalized recommendation systems to accurately

depict users, thus improving consumers' purchasing

power.

3.1 Tmall

3.1.1 Tmall's Dominance in China's

E-Commerce Market

As the undisputed leader in China’s e-commerce

landscape, Tmall holds an unshakable position

characterized by its overwhelming market share,

strategic brand collaborations, and cutting-edge

technological prowess. Dominating the B2C sector

with a 50.8% transaction share and commanding 62.2%

of China’s total e-commerce market, Tmall’s

supremacy reflects not only consumer trust but also

its integrated strengths in supply chain efficiency,

logistics innovation, and technological infrastructure

(Verot, 2024). The platform has emerged as a critical

partner for brands, incubating over 4,000 brands in

2023 alone-many achieving annual sales exceeding

RMB 1 billion-while solidifying its reputation as the

gateway for both international and domestic brands to

penetrate China’s premium and luxury markets.

Leveraging Alibaba’s ecosystem, Tmall pioneers AI-

driven innovations, such as personalized

recommendation systems during mega-sales events

like Singles’ Day, which enhance user engagement

and conversion rates. Its category dominance,

particularly in fashion (30% market share) and beauty

(25% market share), further cements its appeal to

consumers and brands alike, making it the go-to

platform for product launches and trendsetting (Wang,

2024). By synergizing scale, technology, and

consumer insights, Tmall continues to redefine

industry standards and maintain its leadership in an

intensely competitive market.

3.1.2 AI-Driven Consumer Trends and Data

Analytics in Tmall's 2024 Singles' Day

Festival

During Tmall's 2024 Singles' Day Festival (October

14–November 11), AI recommendation systems were

pivotal in shaping consumer trends and enabling data-

driven decision-making. Concurrently, the festival

revealed evolving consumer priorities, with

heightened interest in green technologies and

sustainable consumption, as evidenced by Tmall’s

first full-scale adoption of cloud computing to

minimize environmental impact. Emerging brands

also thrived, with a 70% surge in new brand

registrations in Q3 2024 and selected newcomers

achieving a 239% sales boost during the festival,

signaling growing consumer openness to innovative

products (Zhang, 2024). Tmall’s AI recommendation

systems were instrumental in decoding these

dynamics. The platform delivered hyper-personalized

homepages to users by analyzing vast behavioral

datasets, streamlining product discovery, and

boosting conversion rates. Beyond sales, these

systems empowered brands to refine inventory

management, pricing, and promotions while

elevating user experience. As a barometer of China’s

consumption resilience, Singles’ Day data revealed

nuanced behavioral shifts, such as the rise of eco-

conscious purchasing and preference fragmentation.

Tmall’s AI recommendation system combines

retrieval, ranking, and mechanism modules. Enabling

precise, adaptive product matching during Singles'

Day, driving user satisfaction and revenue growth

(Chen, 2019). As AI technology evolves, its

integration into e-commerce will unlock new

frontiers in hyper-personalization and real-time

market responsiveness. Tmall’s AI recommendation

systems are critical in enhancing user experience and

driving sales during the Singles’ Day mega-sale.

ICEML 2025 - International Conference on E-commerce and Modern Logistics

578

However, the large scale of real-time data and

extreme system loads pose significant challenges to

these systems, particularly in balancing real-time

responsiveness with operational stability. For

example, avoiding errors from lost, duplicated, or

disordered data is vital because users' trust is built by

maintaining data consistency and integrity across

high-frequency streams. With millions of user actions

and purchases per second, the real-time processing

system of vast data streams demands instantaneous

analysis for personalized recommendations. Solid

architecture, load balancing, and fault tolerance are

required to ensure system stability under peak loads.

Resource allocation must dynamically match

fluctuating demand without overprovisioning and

prevent performance degradation from overwhelming

user requests. These challenges demand solutions to

ensure seamless performance and user satisfaction

during the event.

Tmall’s AI recommendation system-combining

retrieval, ranking, and mechanism modules-enabled

precise, adaptive product matching during Singles'

Day, driving both user satisfaction and revenue

growth (Chen, 2019). As AI technology evolves, its

integration into e-commerce will deepen, unlocking

new frontiers in hyper-personalization and real-time

market responsiveness. Tmall’s AI recommendation

systems play a critical role in enhancing user

experience and driving sales during the Singles’ Day

mega-sale. However, the sheer scale of real-time data

and extreme system loads pose significant challenges

to these systems, particularly in balancing real-time

responsiveness with operational stability. Below is an

analysis of the primary challenges.

3.2 Douyin

Douyin has rapidly ascended since its launch in 2016

and has become China's leading short-video platform.

It has profoundly reshaped the landscape of social

media and e-commerce. Douyin’s evolution and

market influence have not only transformed users'

content consumption habits but have also created

innovative marketing and sales channels for brands

and businesses. Since 2019, Douyin has emerged as a

formidable marketing platform through AI-driven ad

targeting and precise user profiling. Its short-video e-

commerce model showcased its significant potential

in the sector by achieving over $200 billion in Gross

Merchandise Volume (GMV) by 2022. The global

expansion strategy of Douyin has also yielded

remarkable success. Its international counterpart,

TikTok, amasses hundreds of millions of users

worldwide and replicates the same strategy globally.

By analyzing user behavior and interest tags, Douyin

delivers hyper-personalized content

recommendations, which enhance user engagement

while equipping brands with efficient marketing tools.

For instance, live-stream hosts leverage interest-

based tagging to recommend relevant products during

broadcasts, significantly boosting sales conversions.

Douyin's live-streaming feature has revolutionized

traditional e-commerce by breaking the fourth wall

between the host and viewers. As a consequence,

enabling real-time interaction allows the host to

showcase the product features and stimulate purchase

intent vividly. The platform's short-video and live-

stream formats offer brands novel marketing avenues,

enabling targeted outreach through creative content,

particularly resonating with younger demographics.

Its multi-channel network (MCN) model has garnered

significant industry attention, fostering collaborations

between content creators and brands. Douyin's rapid

growth has not only solidified its dominance in the

short-video market but has also spurred industry-wide

innovation, with competitors emulating its successful

model, thereby accelerating market maturity and

diversification (Xu et al., 2019). User profiling and

interest tag matching on Douyin play pivotal roles in

e-commerce marketing. By leveraging precise user

profiles and interest-based tagging, livestream hosts

can effectively attract target consumers and drive

sales growth (Koç, 2023). Below is an analysis of

how interest tags function in e-commerce marketing

and their practical applications.

Douyin constructs detailed user profiles by

analyzing behavioral data such as viewing history,

interactions (likes, comments, shares), and purchase

records. These profiles capture demographic

attributes (age, gender, geographic location),

spending habits, and nuanced preferences through

interest tags (e.g., “beauty & cosmetics,” “fitness,”

“culinary”). Interest tags are algorithmically

generated by analyzing user behavior-for instance,

frequent engagement with specific video categories

or participation in related topic discussions prompts

the system to automatically assign corresponding

tags.

The primary function of interest tags lies in

enabling precision marketing. By aligning content

and product recommendations with user tags,

livestream hosts can deliver hyper-targeted

promotions. For example, users tagged with “beauty

& cosmetics” are prioritized for livestreams or ads

featuring makeup tutorials or skincare products. This

targeted approach not only enhances user engagement

and purchase intent but also reduces ad waste by

minimizing irrelevant exposure (X. Wang & Cao,

Path Analysis of Data Empowerment Driving the Upgrading of China Consumer Market

579

2024). Furthermore, Livestream hosts on Douyin

tailor their content to align with users’ interest tags,

creating highly resonant experiences. For instance,

users tagged with “culinary interests” may encounter

livestreams showcasing gourmet snacks or kitchen

gadgets, accompanied by live cooking

demonstrations. By analyzing these tags, hosts craft

interactive segments-such as Q&A sessions or prize

giveaways-to amplify engagement and purchasing

intent. This precision-driven approach directly

influences consumer behavio, content aligned with

user interests accelerates purchase decisions by

reducing cognitive friction. Moreover, accurate tag

matching fosters platform trust and loyalty,

encouraging repeat participation in livestreams and

sustained purchasing activity. Ultimately, interest

tags serve as a dual catalyst-empowering hosts to

refine marketing tactics while deepening user-

platform relationships through relevance and

reliability (Liu and Liang, 2025).

4 CONCLUSION

Through research, this paper concludes that China’s

consumer market is marked by a dual-track dynamic,

where consumption downgrade and upgrade trends

coexist. Economic pressures drive cost-conscious

behavior among middle- and low-income groups, yet

demand for premium products in health, education,

and technology persists, reflecting deepening market

stratification and regional disparities. Eastern regions

emphasize selective, high-quality consumption

supported by robust digital infrastructure, while

central and western areas grapple with broader

expenditure declines. Amid this complexity,

algorithmic innovation has become a linchpin in

bridging gaps and reshaping consumption patterns.

Platforms like Tmall and Douyin exemplify this

transformation. Tmall’s AI-driven recommendation

systems, showcased during Singles’ Day, optimize

real-time personalization, dynamic exposure, and

gamified engagement to boost sales and user

satisfaction. Douyin leverages interest-tagging and

livestream marketing to convert user interactions into

purchase intent, achieving remarkable GMV growth.

Both platforms illustrate how algorithmic precision

addresses consumer heterogeneity, balancing

efficiency with hyper-personalization.

Looking ahead, advancements in AI, edge

computing, and federated learning will further

enhance real-time processing and adaptive

capabilities, enabling platforms to manage extreme

data volumes during mega-events while maintaining

stability. Strengthening data infrastructure in

underdeveloped regions could mitigate regional

disparities, fostering inclusive growth. Sustainability

will likely gain prominence, with AI optimizing green

supply chains and cloud computing minimizing

environmental impacts, aligning with rising eco-

conscious consumer preferences. Douyin’s global

expansion and integration of immersive technologies

(e.g., AR/VR) may redefine cross-border e-

commerce, while Tmall’s ecosystem could pioneer

AI-hardware co-design for seamless omnichannel

experiences.

However, challenges persist, including data

privacy concerns, algorithmic transparency, and the

need for agile resource allocation amid fluctuating

demands. As competitors emulate these models,

continuous innovation will be critical to sustaining

leadership. Ultimately, the synergy between

algorithmic agility, consumer insights, and

infrastructure resilience will shape China’s digital

economy, reinforcing its capacity to navigate

evolving market dynamics while driving global e-

commerce innovation.

REFERENCES

Bain & Company & Kantar Worldpanel, 2024. Changing

with the times: New consumption trends of FMCG

industry in China: 2024 China shoppers report, Series

2.https://www.bain.cn/pdfs/202412121036451898.pdf.

Chen, J., Li, H., 2020. Development prospect of China’s

new consumer economy in the new situation—

Concurrently discussing the impact of COVID-19.

Open Journal of Business and Management 8(3), 1201–

1205.

Chen, Q., 2019. How does the recommendation system

work on TMall? Alibaba Cloud Community.

https://www.alibabacloud.com/blog/595335.

Deloitte, 2023. 2023 White paper on consumer insight and

market outlook in China. https://www2.deloitte.com/c

ontent/dam/Deloitte/cn/Documents/consumer-business

/deloitte-cn-cb-consumer-insight-zh-230118.pdf.

Han, S., 2024. Analysis of MUJI’s development strategy in

China under the background of consumption

downgrade. Deleted Journal 12(1), 36–39.

Koç, B., 2023. The role of user interactions in social media

on recommendation algorithms: Evaluation of

TikTok’s personalization practices from user’s

perspective. Istanbul University.

Liu, H., Liang, J., 2025. A study on the factors influencing

Chinese costume consumers utilizing live streaming

platforms to purchase products: A case study of

Douyin. Journal of Theoretical and Applied Electronic

Commerce Research 20(1), 38.

Mintel, 2024. Consumers in China in 2024: Towards

diversification.

ICEML 2025 - International Conference on E-commerce and Modern Logistics

580

Sing, C.K., 2024. China consumer is epitome of delayed gr

atification. Reuters. https://www.reuters.com/breaking

views/china-consumer-is-epitome-delayed-gratificatio

n-2024-11-04/.

Verot, O., 2024. What is Tmall? China's #1 e-commerce pl

atform. Marketing China. https://marketingtochina.co

m/tmall-the-ultimate-guide/.

Wang, G., 2024. The global growth of e-commerce and Ch

ina’s leading position-Chinese digital marketing. Jade

mond. https://www.jademond.com/magazine/global-e-

commerce-made-in-chinas/.

Wang, X., Cao, S., 2024. Harnessing the stream:

Algorithmic imaginary and coping strategies for live

streaming e-commerce entrepreneurs on Douyin. The

Journal of Chinese Sociology 11, 9.

Xu, S. et al., 2019. Research on business model innovation

of short video enterprises from the perspective of

community economy—Taking “Douyin” as an

example. Atlantis Press.

Zhang, S., 2024. LIVE BLOG: Alibaba’s 2024 11.11 glob

al shopping festival. Alizila. https://www.alizila.com/al

ibaba-2024-11-11-live-blog/.

Zhang, Y., Zhan, R., 2023. Does online retailing narrow the

regional economic gap? -Empirical analysis based on

the data of prefecture-level cities in China.

https://fddi.fudan.edu.cn/cd/38/c19046a642360/page.h

tm.

Zipser, D., McKinsey & Company, 2024. China brief: The

truth about Chinese consumption. https://www.mckins

ey.com/cn/~/media/mckinsey/locations/asia/greater%2

0china/our%20insights/the%20truth%20about%20chi

nese%20consumption/mckinsey%20china%20brief%2

0the%20truth%20about%20chinese%20consumption

%20en.pdf.

Path Analysis of Data Empowerment Driving the Upgrading of China Consumer Market

581