Short answer

Proactively address negative consumer sentiment related to product functionality and usage by refining marketing communications and informing product improvements.

Field: Innovation & Markets
Source: Journal of theoretical and applied electronic commerce research (2025)
Method: Text Mining, Topic Modeling (LDA), Sentiment Analysis (LSTM), Social Network Analysis
Sample: 14,078 reviews
Evidence: Strong effect

Analyzing consumer reviews using text mining techniques reveals key product aspects and sentiment trends, enabling targeted marketing adjustments for cross-border e-commerce. This innovation & markets research insight is drawn from a 2025 study published in Journal of theoretical and applied electronic commerce research. Using Text mining, topic modeling (lda), sentiment analysis (lstm), social network analysis with 14,078 reviews, researchers explored how this design variable affects real-world outcomes. The key design takeaway: Proactively address negative consumer sentiment related to product functionality and usage by refining marketing communications and informing product improvements.

Study

Innovation & MarketsNew This WeekStrong effect

Consumer sentiment analysis of online reviews can significantly improve cross-border e-commerce marketing strategies.

Analyzing consumer reviews using text mining techniques reveals key product aspects and sentiment trends, enabling targeted marketing adjustments for cross-border e-commerce.

Journal of theoretical and applied electronic commerce research · 2025

Key Findings

01Consumers primarily focus on functional features, quality/cost-effectiveness, usage effectiveness, post-purchase support, and design/assembly.
02Negative sentiment is disproportionately high for reviews concerning functional features and usage effectiveness.
03Topic and sentiment evolution trends can be identified over different time periods.

Application

Design takeaway

Proactively address negative consumer sentiment related to product functionality and usage by refining marketing communications and informing product improvements.

How to apply

Implement automated text analysis tools to regularly process customer reviews, identify recurring themes and sentiment patterns, and use these insights to guide marketing campaigns and product feedback loops.

Project actions

01Clearly define the scope of your review data collection (e.g., specific product category, platform).
02Document the specific text mining and sentiment analysis tools and techniques used.
03Visualize the sentiment trends over time to show evolution.

Method & Evidence

AimHow can text mining of consumer online reviews be used to understand sentiment tendencies and inform marketing strategies for cross-border e-commerce?

MethodText Mining, Topic Modeling (LDA), Sentiment Analysis (LSTM), Social Network Analysis

ProcedureCollected and analyzed 14,078 online reviews for top-selling products on a cross-border e-commerce platform. Utilized Python with Jupyter Notebook to apply LDA for topic extraction, LSTM for sentiment classification, and social network analysis to identify key themes and their associated sentiments over time.

Sample14,078 reviews

ContextCross-border E-commerce Marketing

Variables

IV["Consumer review text","Time period"]

DV["Identified topics","Sentiment distribution (positive, negative, neutral)","Sentiment trends over time"]

CV["E-commerce platform","Product category","Analysis tools (Python, LDA, LSTM)"]

Strengths & Limitations

Strengths

+Utilizes a large dataset of actual consumer reviews.
+Employs advanced text mining and machine learning techniques.
+Integrates theoretical frameworks (4P/4C) for practical application.

Limitations

The accuracy of sentiment analysis can be affected by sarcasm, irony, and context-specific language.

Reliability & validity

Reliability can be enhanced by using multiple annotators for sentiment classification if done manually, or by validating the automated model's accuracy against a manually labeled subset. Validity is strengthened by using established topic modeling and sentiment analysis algorithms and by ensuring the review sample is representative of the target market.

Think critically

To what extent can sentiment analysis accurately capture the complexity of consumer opinions, and what are the ethical considerations when using this data for marketing?

Design Principles

"Leverage consumer-generated data to iteratively improve product marketing and development."

Understanding the nuances of consumer feedback, particularly sentiment towards specific product features and post-purchase experiences, is crucial for refining marketing messages and product development in the competitive global e-commerce landscape. This data-driven approach allows businesses to proactively address concerns and capitalize on positive perceptions.

What This Means for Your Design

By reading what customers write in online reviews, businesses can figure out what people like and dislike about products, especially for international online shopping. This helps them make better ads and improve the products themselves.

How to use in your project

1.Use the methodology as a framework for analyzing qualitative data in your own design project.
2.Cite the findings to support the importance of user feedback in market analysis.

Add to My Project

Quick Cite

Paragraph starter

This study demonstrates that analyzing consumer sentiment from online reviews using text mining techniques, such as Latent Dirichlet Allocation (LDA) for topic modeling and Long Short-Term Memory (LSTM) for sentiment classification, can reveal critical insights into product perception. Specifically, it highlights that consumers often express negative sentiment regarding functional features and usage effectiveness, which can inform targeted marketing strategies and product development for cross-border e-commerce enterprises.

Source

Journal of theoretical and applied electronic commerce research

Text Mining for Consumers’ Sentiment Tendency and Strategies for Promoting Cross-Border E-Commerce Marketing Using Consumers’ Online Review Data

journal · 2025

View source

Questions About This Research

What does the research say about consumer sentiment analysis of online reviews can significantly improve cross-border e-commerce marketing strategies?: Proactively address negative consumer sentiment related to product functionality and usage by refining marketing communications and informing product improvements. Evidence: Journal of theoretical and applied electronic commerce research (2025).
Why does "Consumer sentiment analysis of online reviews can significantly improve cross-border e-commerce marketing strategies." matter for design?: Understanding the nuances of consumer feedback, particularly sentiment towards specific product features and post-purchase experiences, is crucial for refining marketing messages and product development in the competitive global e-commerce landscape. This data-driven approach allows businesses to proactively address concerns and capitalize on positive perceptions.
How can designers apply this research?: Proactively address negative consumer sentiment related to product functionality and usage by refining marketing communications and informing product improvements.
What were the main findings?: Consumers primarily focus on functional features, quality/cost-effectiveness, usage effectiveness, post-purchase support, and design/assembly.. Negative sentiment is disproportionately high for reviews concerning functional features and usage effectiveness.. Topic and sentiment evolution trends can be identified over different time periods.
What research method was used?: Text Mining, Topic Modeling (LDA), Sentiment Analysis (LSTM), Social Network Analysis with 14,078 reviews.
How strong is the evidence?: Evidence strength is rated Strong effect, based on a 2025 journal from Journal of theoretical and applied electronic commerce research.
What should I do differently in my next project?: Implement automated text analysis tools to regularly process customer reviews, identify recurring themes and sentiment patterns, and use these insights to guide marketing campaigns and product feedback loops.
What are the limitations?: The analysis is specific to one platform and product category, and sentiment analysis models may not capture all nuances of human language.