Statistical Analysis of the Impact of Short Video Marketing on
Consumers' Purchase Intention
Lihan Chen
a
College of Mathematics and Statistics, Northeastern University at Qinhuangdao, Qinghuangdao, Hebei, 066003, China
Keywords: Short Video Marketing, Consumer Purchasing Behavior, Brand Recognition.
Abstract: With the rapid development of digital technology, short video platforms have gradually come to occupy the
core position of brand marketing. However, short video marketing faces challenges in terms of traffic
conversion, content quality, and homogenization, and based on this. The purpose of this research is to make
up for the difference in theoretical and practical aspects. This study employs a combined methodology of
questionnaire surveys and empirical testing to systematically deconstruct the mechanistic pathways through
which the technological features of Video-on-Demand (VOD) services influence consumer purchasing
behavioral tendencies, Pearson Correlation Analysis, Regression Model and Factor Analysis. Based on the
reliability analysis, Cronbach's alpha is above 0.8, which means that the data is reliable. Among them, the
consistency of practical information, cultural connotation, and celebrity image is particularly prominent in
driving consumer decision-making. In addition, the in-depth dissemination of short videos and corporate
brand culture can effectively enhance consumer trust and brand recognition.
1 INTRODUCTION
In today's technological progress, the traditional
novels, pictures and text as a representative of digital
media entertainment have been unable to meet the
needs of people's lives. Short video platforms are also
emerging, and the traffic liquidity modes such as
advertisements and live broadcasts are continuously
innovated. Their high-quality content is bringing
more business opportunities and value dividends to
brands (Wang, 2020). Since 2014, short video
platforms have risen rapidly and gradually become
popular, bringing new experiences and feelings to
users. User engagement on short video platforms is
significantly higher than that of traditional social
media (Voorveld et al., 2018), a phenomenon closely
related to Gen Z consumers' preference for visual,
interactive content (Djafarova & Bowes, 2021).
Under this trend, technological progress has brought
about changes in content production models,
communication methods, and audience acceptance
habits, while intensified market competition has led
to the gradual fading of traffic dividends (Liu 2022).
Many brands are promoting myopic frequency, but
a
https://orcid.org/0009-0004-3080-8956
they also face a lot of problems, such as short video
traffic conversion rate being very low, the content
obtained by Yao (2018) through data and literature
analysis is inferior and vulgar, Li &Wang (2021) has
too much homogeneous content through data and
literature analysis, and Nilobar & Zheng (2021)
highlight the inherent dilemmas in short video content
creation under capital and algorithmic pressures, and
the trust crisis caused by lack of supervision, Elti &
Zheng (2021) has obtained the monopoly crisis under
capital operation through the analysis of existing
literature data. So, how to affect the customer's
buying intent in the short video marketing is a subject
worth further research. Based on these, the author
makes an analysis of the video style, discusses the
spread level of Music Video (MV), and researches the
influence of MV on customers's perception and
buying intention.
2 RESEARCH DESIGN AND
EMPIRICAL TESTING
This article did a study by means of a conventional
questionnaire. Among the major possible variables
Chen, L.
Statistical Analysis of the Impact of Short Video Marketing on Consumers’ Purchase Intention.
DOI: 10.5220/0013815300004708
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 2nd International Conference on Innovations in Applied Mathematics, Physics, and Astronomy (IAMPA 2025), pages 129-133
ISBN: 978-989-758-774-0
Proceedings Copyright © 2025 by SCITEPRESS Science and Technology Publications, Lda.
129
are the Video-on-Demand (VOD) system's message
quality, the short video content, the video style, and
the short video operation method. Based on the
empirical framework of Kim and Ko (2012) on social
media operation strategies, this study further
optimizes the measurement dimension of short video
operation methods, covering innovative strategies
such as gamified advertising and influencer
cooperation. Formal questionnaires are distributed
through professional online questionnaire platforms
and online social platforms for sample data collection.
The report consists of three parts. One is about
personal data, the second part is consumers'
perception of the information quality, short video
content, the third chapter discusses how the star effect
affects customer's shopping intention through brand
effect. A total of 300 questionnaires were sent, with a
1:1 ratio of men to women.
On this basis, through Pearson correlation
analysis, regression analysis, factor analysis and other
means, to study how short video advertising
communication through the customer's effective
impact, thus promoting buying desire. In order to
further verify the visual appeal of short video
keyframes, the eye tracking technology of Wang and
Kim (2021) can be used to analyze the user's attention
distribution of color and dynamic elements. In
addition, AR technology (augmented reality) and
gamification elements can be combined in interactive
advertising design, and Xu and Sundar (2023) have
experimentally demonstrated that such designs can
significantly improve users' immersion and purchase
intention (β=0.29, p<0.01)
3 RESULT
3.1 Validity and Reliability Verification
When the squared values of the simple correlation
coefficients generally exceed those of the partial
correlation coefficients, and the KMO (Kaiser-
Meyer-Olkin) measure approaches 1, this indicates
significant correlations among the variables, making
it appropriate to perform factor analysis. Conversely,
if the values progressively approach 0, it reveals that
the associations between them are not strong.it
suggests minimal mutual influence between
variables, rendering the data unsuitable for factor
model analysis. Table 1 shows the KMO values for
this study and the approximate chi-square.
The validity analysis results demonstrate that the
structural validity of the research data meets optimal
criteria: The Kaiser-Meyer-Olkin (KMO) measure of
sampling adequacy reached 0.931 (>0.9 threshold),
and Bartlett's test of sphericity showed statistically
significant results (p < 0.001), confirming the
suitability of the data for factor analysis. Through
principal component analysis, six key factors with
eigenvalues exceeding 1 were extracted, collectively
accounting for 73.17%.of the total variance,
indicating strong explanatory power of the
constructed model.Among them, the factor loadings
of "Short Video Content" (Factor 3, loadings 0.763 -
0.867) and "Consumer Attitude" (Factor 1, loadings
0.756 - 0.864) are relatively high, verifying the
rationality of the scale design.
Table 1: KMO Test Results
KMO test and Bartlett test
KMO value 0.931
Bartlett
sphericity
test
Approximate chi-square 6158.578
df 351
P 0.000***
Note: ***, **, and * represent the significance levels o
f
1%, 5%, and 10%, respectivel
y
The research was done with SPSS 26.0 software
after the data processing stage. The Cronbach alpha
ratio was 0,941 as shown in Table 2 and all Adjusted
Item-Sum Correlations (CITCs) were greater than
0.4. Notably, no significant increase in the α
coefficient was observed even when deleting any
single measurement item. These results demonstrate
high internal consistency of the questionnaire and
excellent data reliability.
Table 2: The Influence of different short video advertisement content on consumption intention
The average
after the item
is delete
d
The variance
after the item is
remove
d
The relevance of the
deleted item to the overall
after the item was delete
d
Cronbach's α
coefficient after the
term is remove
d
8. How much does the different content o
f
short video ads affect your willingness to
spend? - Make teasers (e.g. movie trailers,
roduct sales trailers)
86.891 478.735 0.641 0.939
9. Introduce the product itself (e.g. Xiaomi
b
rand features
)
86.894 487.939 0.481 0.94
IAMPA 2025 - The International Conference on Innovations in Applied Mathematics, Physics, and Astronomy
130
Table 2: The Influence of different short video advertisement content on consumption intention (cont.).
The average after
the item is
delete
d
The variance after
the item is
remove
d
The relevance of the
deleted item to the overall
after the item was delete
d
Cronbach's α
coefficient after the
term is remove
d
10. Introduce how to use it (such as the
inline control headphones of Xiaomi
mobile
p
hones
)
86.844 485.051 0.533 0.94
11. Storytelling (e.g. Coca-Cola's magic
story)
86.81 487.192 0.517 0.94
12. Do activities (e.g. Disney-themed
events)
86.857 484.911 0.548 0.94
The correlation between the deleted items in 9-12
and the overall after deletion was greater than 0.4, and
there was no significant change in Cronbach's α
coefficient after deletion
3.2 Correlation Characteristics
Between Variables
Pearson correlation analysis showed that the five
core variables of short video marketing (the degree of
influence of content consumption intention, the
degree of influence of style purchase intention, the
degree of influence of techniques, the degree of
influence of celebrity effect,it has good correlation
with customer satisfaction (P < 0.001), and the
correlation range is 0.33 to 0.528.
Table 3: Pearson correlation was used to analyze the results
The degree of influence
of content consumption
intention
The degree to which
the willingness to buy
is influenced by style
The degree o
f
influence of the
manipulation
The degree o
f
influence of the
celebrity effect
Attitude
The degree of influence
of content consumption
intention
1(0.000***) 0.33(0.000***) 0.507(0.000***) 0.43(0.000***) 0.427(0.000***)
The degree to which the
willingness to buy is
influenced b
y
st
y
le
0.33(0.000***) 1(0.000***) 0.528(0.000***) 0.437(0.000***) 0.448(0.000***)
The degree of influence
of the manipulation
0.507(0.000***) 0.528(0.000***) 1(0.000***) 0.497(0.000***) 0.499(0.000***)
The degree of influence
of the celebrity effect
0.43(0.000***) 0.437(0.000***) 0.497(0.000***) 1(0.000***) 0.436(0.000***)
Attitude 0.427(0.000***) 0.448(0.000***) 0.499(0.000***) 0.436(0.000***) 1(0.000***)
Note: ***, **, and * represent the significance levels of 1%, 5%, and 10%, respectivel
y
As Table 3, the correlation between the degree of
influence of tactics and consumer attitudes was the
strongest (r=0.499), indicating that the innovation of
short video operation methods (such as hot topic
interaction and brand integrated packaging) had a
significant driving effect on consumer attitudes.
The correlation between celebrity effect and
content consumption intention was high (r=0.43),
indicating that the consistency of celebrity image and
advertising content can effectively improve users'
content attention.
3.3 Linear Regression Models of
Consumer Attitudes
Based on the results of linear regression, it can see
that the four dimensions of VRT (content consuming
intent, fashion preference, operating mode, and
celebrity influence) are significantly positively
influenced by Tables 4 and 5 (p < 0.01){as shown in
Table 4}, the model goodness of fit R²=0.347,
adjusted R²=0.339, and F test was significant
(F=42.066, p<0.001), which was as follows:
The standardization coefficient of the influence of
operational methods was the highest (β=0.222),
indicating that the innovation of operational strategies
(such as gamified ad placement and influencer
cooperation) was the most critical factor in improving
consumer attitudes.
Style preference (β=0.202) and content
consumption intention (β=0.178) were the second,
indicating that the creative style (e.g., humor, cultural
atmosphere) and practical content (e.g., product
Statistical Analysis of the Impact of Short Video Marketing on Consumers’ Purchase Intention
131
preview, usage method) of short videos have a
continuous impact on consumer attitudes.
The formula of the model is:
𝐴𝑡𝑡𝑖𝑡𝑢𝑑𝑒 =
0.542 0.197𝑋1 0.214𝑋2 0.218𝑋3 0.173𝑋4 (1)
Among them, X1X1 to X4X4 represent the
influence of content, style, technique, and celebrity
effect, respectively.
Table 4: Linear regression analysis results n=321
Non-normalized
coefficients
Normalizatio
n facto
r
t P VIF
Adjust
F
B standard erro
r
Beta
constant 0.542 0.22 - 2.467 0.014** -
0.347 0.339
F=42.066
P=0.000***
The degree of influence of content
consumption intention
0.197 0.06 0.178 3.284 0.001*** 1.427
The degree to which the willingness
to buy is influenced by style
0.214 0.058 0.202 3.668 0.000*** 1.469
The degree of influence of the
mani
p
ulation
0.218 0.06 0.222 3.665 0.000*** 1.780
The degree of influence of the
celebrity effect
0.173 0.06 0.161 2.901 0.004*** 1.482
De
p
endent variable: Attitude
Note: ***, **, and * re
p
resent the si
g
nificance levels of 1%, 5%, and 10%, res
p
ectivel
y
In view of the limitations of the current linear
regression model goodness-of-fit of only 0.339,
the three dimensions of theoretical construction,
method innovation and practical strategy were
systematically optimized. At the theoretical level, the
existing model can only explain 33.9% of the
variation of purchase intention, exposing the neglect
of key variables such as consumer psychological
distance and algorithm recommendation mechanism,
so it is recommended to introduce the SOR theoretical
framework to construct a chain mediation model of
"content characteristics→ perceived value
purchase intention", and include product involvement
as a moderating variable, and verify the interaction
effect between variables through the Bootstrap
method. At the methodological level, a hybrid
research design can be adopted, combined with
questionnaire survey and eye tracking technology,
Hotspot Analysis 3.2 software can be used to analyze
the visual attractiveness of short video keyframes, and
the panel data state space model can be established,
the SURE estimator can be used to deal with the
cross-section-dependent problem, and a 6-month
observation period can be set to capture the
attenuation trajectory of purchase intention. In terms
of practical strategies, it is recommended to construct
a dynamic content matrix, determine the optimal ratio
of utility (40%), entertainment (30%), cultural (20%),
and public welfare (10%) through A/B testing,
develop a UGC incentive model to calculate the ROI
threshold of KOL cooperation, and apply AIGC
technology to improve creative production efficiency,
but it is necessary to control the AI perception within
15% to avoid a trust crisis. Follow-up research can
further reveal the mechanism of short video
marketing effect by integrating machine learning
algorithms and structural equation models to
construct multimodal prediction equations including
neuroscience indicators.
4 CONCLUSION
Based on the results of questionnaire surveys and the
use of integrated model theory, this paper extracts
three main characteristics of short video marketing
influencing consumers through interviews and
relevant data and conducts an empirical study on how
these three characteristics affect consumer behavior.
The results show that the information quality, short
video content, video style. Both the architecture of
video platforms and the interaction mechanisms of
short-form video content exert significant positive
effects on product sales performance. Consumer
perceived value effectively enhances purchase
decision-making intentions. Based on these empirical
research findings, it is suggested that brands should
strengthen the innovation of operational methods in
short video marketing, such as designing interactive
ads based on hot topics, or improving user
engagement through gamification; Pay attention to
the practicality of content and the diversity of styles,
balance creative entertainment and rigorous
practicality to meet different consumer preferences;
Make reasonable use of the celebrity effect, choose
IAMPA 2025 - The International Conference on Innovations in Applied Mathematics, Physics, and Astronomy
132
spokespersons who are in line with the brand tonality,
and enhance the user's sense of trust; Optimize the
quality of short video information, enhance brand
image through public welfare content or cultural
narratives, and indirectly promote purchase intention.
REFERENCES
Djafarova, E., & Bowes, T., 2021. ‘Instagram made me buy
it’: Generation Z impulse purchases in the fashion
industry. Journal of Retailing and Consumer Services,
59, 102345.
Kim, A. J., & Ko, E., 2012. Do social media marketing
activities enhance customer equity? An empirical study
of luxury fashion brands. Journal of Business Research,
65(10), 1480–1486.
Liu, D., 2022. Research on how short video marketing
affects consumers' purchase intention [M.S. dissertation,
Shanghai University of Finance and Economics].
Master's.
Li, Q., & Wang, H., 2021. Research on the marketing value
of Douyin short video content. Science and Education
Papers, (07), 180–181.
Nilobar, E., & Zheng, L., 2021. Characteristics, trends, and
dilemmas of short video content production in the new
media era. China Editor, (03), 81–85.
Voorveld, H. A., van Noort, G., Muntinga, D. G., &
Bronner, F., 2018. Engagement with social media and
social media advertising: The differentiating role of
platform type. Journal of Advertising, 47(1), 38–54.
Wang, X., 2020. Research on the Impact of Short Video
Content Marketing on Consumers' Purchase Intention.
Master's Dissertation, Jilin University. Master's.
Wang, Q., & Kim, J., 2021. Eye-tracking analysis of visual
attention in short video ads: The role of color, motion,
and framing. Journal of Advertising Research, 61(4),
432–448.
Xu, J., & Sundar, S. S., 2023. Interactive short video ads:
How gamification and augmented reality enhance
consumer immersion and purchase intent. Journal of
Computer-Mediated Communication, 28(2), zmad005.
Yao, M., 2018. Discussion on mobile short video
communication strategy based on content marketing.
China Press, (16), 74–75.
Statistical Analysis of the Impact of Short Video Marketing on Consumers’ Purchase Intention
133