Customer-Oriented Strategic Planning for Hotel Competitiveness Improvement Based on Online Reviews

Yuan Yuan, Tian-Hui You, Tian’ai Xu, Yu XunSustainability 2022 · DOI

TL;DR

This paper introduces a novel, integrated framework for hotels to improve their competitiveness by leveraging online customer reviews. It combines text mining, sentiment analysis, and SWOT analysis to extract actionable insights from reviews, identify key competitors, and formulate customer-centric strategic plans.

What Problem Does This Paper Address?

The hotel industry is intensely competitive, making effective strategic planning crucial for survival and growth. Traditional strategic planning methods often rely on internal data or questionnaires, which can be limited by insufficient data and a lack of direct customer perspectives. This paper addresses this gap by proposing a method that uses readily available online reviews to understand customer needs and preferences, analyze competitive landscapes, and develop data-driven strategies.

Methods

The proposed method is a multi-stage process:

  1. Online Review Data Collection and Preprocessing: Web scraping is used to gather reviews from platforms like Ctrip.com, followed by standard text preprocessing techniques such as word segmentation, POS tagging, and stop-word removal using the Jieba library.
  2. Text Mining for Customer-Concerned Service Attributes:
    • Attribute Extraction: Latent Dirichlet Allocation (LDA) is employed to identify key service attributes that customers care about from the reviews.
  3. Customer Concern Level and Performance Evaluation: Customer concern level is gauged by the frequency of attribute mentions in reviews. Performance is evaluated using a multi-granularity sentiment analysis approach, which assigns sentiment scores (-2 to +2) to review segments related to specific attributes.
  4. Competitive Structure Analysis and Main Competitor Identification:
    • Correspondence Analysis (CA): CA is used to visualize the competitive landscape of hotels based on their perception frequencies of service attributes.
  5. Cosine Similarity: This metric is applied to the CA results to quantify the competitive proximity between the target hotel and its competitors, allowing for the identification of main competitors.
  6. SWOT Analysis toward Main Competitors:
    • Factor Identification: Strengths, Weaknesses, Opportunities, and Threats (SWOT) factors are identified by comparing the target hotel's performance against its main competitors. This includes a novel approach to identifying opportunity and threat factors based on performance differences.
  7. Factor Prioritization: The priorities of SWOT factors are determined by considering customer concern level, attribute performance, and performance differences, using a weighted approach.
  8. Strategic Planning: The identified SWOT factors and their priorities are used to generate strategic planning recommendations, presented in a TOWS matrix.

Datasets

The study uses online reviews collected from Ctrip.com for a five-star hotel in Shanghai (Mandarin Oriental Pudong Shanghai) and 13 other five-star hotels in the same area. A total of 35,558 valid online text reviews were used after preprocessing.

Key Insights and Ideas

  • Leveraging Unstructured Data: Online reviews, often overlooked or underutilized, are a rich source of customer sentiment and specific service attribute feedback that can directly inform strategic decisions.
  • Customer-Centricity is Key: The framework explicitly prioritizes customer concerns and perceptions, moving beyond internal assessments to truly understand what drives hotel competitiveness.
  • Integrated Approach: Combining text mining, sentiment analysis, competitive analysis (CA), and SWOT provides a holistic view, bridging the gap between raw data and actionable strategy.
  • Quantitative Prioritization: The method offers a more objective way to prioritize SWOT factors compared to traditional subjective approaches, leading to more targeted resource allocation.
  • Dynamic Competitor Identification: The framework allows for the identification of main competitors based on actual competitive positioning, rather than pre-defined lists.

Key Results

  • 12 Customer-Concerned Service Attributes Identified: Through LDA, attributes like "Room service," "Room layout," and "Check-in/out service" were found to be significant for customers.
  • Performance Differences Highlighted: The study revealed performance variations among hotels for specific attributes, such as the target hotel performing best on "Internal environment and cleanliness" but worst on "Check-in/out service."
  • Main Competitors Identified: Correspondence analysis and cosine similarity identified the four closest competitors to the target hotel as H8, H6, H9, and H7.
  • Actionable SWOT Strategies: The study generated specific SWOT factors and prioritized them, leading to strategic recommendations like prioritizing "Room service" for strengths, improving "Check-in/out service" for weaknesses, leveraging "Location/traffic" for opportunities, and addressing "Room layout" for threats.

Limitations and Future Work

  • Single Data Source: The study relied solely on Ctrip.com reviews. Future work could incorporate reviews from multiple platforms for a more comprehensive view.
  • Dictionary-Based Sentiment Analysis: While effective, dictionary-based sentiment analysis could be enhanced by machine learning methods for improved efficiency and accuracy.
  • Limited Data Types: The study focused on online reviews. Integrating internal hotel data with online review data could provide even richer strategic insights.
  • Managerial Weights: The study used equal weights (0.5) for customer concern and performance in factor prioritization. Future research could explore how varying these weights based on managerial judgment impacts strategic outcomes.

Summary updated

Abstract

The hotel industry has been facing fierce competition in recent years. It is important for hotels to conduct effective strategic planning for competitiveness improvement to achieve sustainable development. Prior studies on hotel strategic planning mainly use questionnaire data or hotel internal data, which have the problems of insufficient data or neglecting customer perspectives. The purpose of this study was to develop an integrated method for customer-oriented strategic planning for hotel competitiveness improvement based on text mining of online reviews. First, text mining of online reviews was conducted to extract customer-concerned service attributes and evaluate customer concern level and the performance of the service attributes through Latent Dirichlet Allocation (LDA) and sentiment analysis. Second, the competitive structures of the hotels were analyzed and the main competitors were identified from the competitive hotels through correspondence analysis. Third, SWOT analysis of the target hotel toward the main competitors was conducted, and the priorities of factors in each SWOT category were determined. An empirical study on a five-star hotel is given to illustrate the feasibility and effectiveness of the proposed method. The results indicate that the proposed method can help managers in strategic planning to obtain more specific strategies for hotel competitiveness improvement.
BibTeX
@inproceedings{yuan2022customer,
  author = {Yuan Yuan and Tian-Hui You and Tian’ai Xu and Yu Xun},
  title = {{Customer-Oriented Strategic Planning for Hotel Competitiveness Improvement Based on Online Reviews}},
  booktitle = {Sustainability},
  year = {2022},
  doi = {10.3390/su142215299},
}
Figure