Text analytics to improve contact center performance


for a Pharma Major

Type
Pharma Manufacturer
Duration
10 Weeks
Size
10,000+ Employees

Background

  • The contact centre of a large pharmaceutical company wanted to analyse their contact centre text data to understand trends, patterns and improve customer centricity
  • The text data received was primarily transcripts of conversations between call centre support team and clients
  • DataZymes used text analytics algorithmic approaches and natural language processing (NLP) based on statistical, linguistic and machine learning (ML) methods
  • The project involved two phases – one that involved data preparation of the data and the other that involved data analysis
  • The customer wanted to analyse their contact centre text data that mostly included call transcripts to understand trends and improve customer centricity

    Data Preparation

    A Eight-Step approach was followed for data preparation:

  • Sentence segmentation
  • Build corpus or dictionary
  • Tokenization
  • Stemming/lemmatization
  • Part-of-speech tagging
  • Parsing
  • Named entity recognition
  • Co-reference resolution

  • Data Analysis

    The DZ Team identified the top reasons for delays in improvements for operational efficiency. Using a combination of machine learning and NLP, hot topics and customer sentiments were analysed.

    Impact

  • The top reasons for delays leading to improvements in operational efficiency were identified
  • Hot topics were identified by understanding customer behaviour
  • Sentiment analysis identified the sentiments of the customers
  • Customer experience was improved by 15%