Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Luxury Retail Customer Insights: Social Media & Sentiment Analytics

Business Problem

Luxury retailers invest heavily in social media, but high follower counts do not always translate into meaningful engagement or positive brand perception.

This project analyses Harrods’ social media performance against Selfridges to:

  • Diagnose engagement decline despite strong brand equity
  • Identify gaps between passive and active engagement
  • Extract customer sentiment and themes from verified reviews
  • Translate insights into actionable content and CX recommendations

Data & Scope

The analysis combines platform performance data and customer-generated text data:

  • Social media KPIs (30-day window: 15 Feb – 16 Mar 2025)
    • Instagram, Facebook, TikTok
    • Engagement, posting frequency, follower growth, applause vs conversation
  • Customer reviews
    • Verified Trustpilot reviews (Harrods)
    • Sentiment scores, polarity categories, topic prevalence

Competitor benchmark: Selfridges


Analytical Approach

1. Social Media Performance Analysis

  • Cross-channel KPI comparison
  • Engagement per post vs posting frequency
  • Applause (likes) vs conversation (comments) diagnostics
  • Competitor benchmarking (Harrods vs Selfridges) Visual reference:
    Harrods vs Selfridges – Engagement & Conversation Comparison
    engagement per post

2. Sentiment Analysis

  • Context-aware sentiment scoring using R (sentimentr)
  • Sentiment categorisation (Positive / Neutral / Negative)

Visual references:*

  • Sentiment Score Distribution
  • Sentiment Category Breakdown

3. Topic Modelling

  • Structural Topic Modelling (STM, K=8 topics)
  • Topic–sentiment and topic–time relationships

Visual references:

  • Topic Summary & Labels
  • Sentiment vs Topic Effects
  • Topic Trends Over Time

Key Insights

Engagement Performance

  • Harrods generated ~65,700 total engagements, but this represented a 20.9% decline versus the prior 30-day period.
  • Despite lower posting volume, Harrods achieved ~1,530 engagements per post, significantly higher than Selfridges (~96 engagements per post).
  • Harrods dominated passive engagement:
    • ~63,500 applause interactions (likes/favourites)
  • Selfridges outperformed in active engagement:
    • +52.9% increase in comments, while Harrods comments declined 21.7%

Interpretation:
Harrods maintains strong brand affinity but underperforms in conversation-driven engagement.


Content Strategy Gap

  • Harrods posting frequency declined 10.4% during the analysis period.
  • Engagement decline coincided with reduced posting and limited interactive formats.
  • Selfridges posted more frequently and adopted trend-led formats, particularly on TikTok, where it held ~166K followers vs Harrods’ ~19.9K.

Sentiment Analysis

  • Trustpilot sentiment distribution showed:
    • Majority positive sentiment, aligning with Harrods’ 4-star average rating
    • Smaller but persistent negative cluster
  • Topic modelling revealed:
    • Negative sentiment strongly associated with pricing, service delays, and payment issues
    • Positive sentiment driven by staff quality, ease of shopping, and overall experience
  • Complaint-related topics showed declining prevalence over time, indicating partial resolution but not elimination.

Business Implications

  • Engagement strategy:
    Harrods should prioritise conversation-led formats (polls, Q&A, UGC prompts) to convert passive admiration into interaction.

  • Channel strategy:
    TikTok represents a clear competitive gap; increasing posting cadence and UGC activation is critical to reach younger luxury consumers.

  • Customer experience:
    Sentiment and topic analysis highlight pricing transparency and service efficiency as the highest-risk perception drivers.

  • Measurement maturity:
    Combining KPI diagnostics with sentiment analytics provides a fuller view than engagement metrics alone.


Tools & Techniques

  • R: sentimentr, ggplot2, stm
  • Analytics methods: KPI benchmarking, sentiment analysis, topic modelling
  • Text analytics: tokenisation, polarity scoring, topic prevalence
  • Business frameworks: SWOT-informed recommendations, KPI diagnostics

How to Navigate

  • README.md → business context, insights, decisions
  • code/ → sentiment analysis & topic modelling scripts (R)
  • visuals/ → engagement diagnostics, sentiment distributions, topic outputs
  • data/ → cleaned Trustpilot review dataset

About

Customer and social media analytics project analysing engagement, sentiment, and brand perception for a luxury retailer using KPI analysis, sentiment scoring, and topic modelling

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors