What is text analytics?

Text analytics uses specialised software to identify topics, sentiment and text patterns in text data. It is usually used in circumstances where the volumes of data are very high, or reporting is required in real-time.

What can it deliver?

  • Identification of key topics in the data
  • Identification of links or associations between these topics (text patterns)
  • Full categorisation of the topics – usually structured to include ‘macro’ categories and a series of subcategories
  • Analysis of prevailing sentiment in the data and/or sentiment associated with specific categories
  • Analysis of the potential impact of different categories on sentiment and/or a KPI measure

Who do I contact?

Stephen Yap
Head of Customer, Ipsos Loyalty

More about text analytics

Text analytics is an analytical approach that efficiently and reliably structures and identifies the main themes in text data; thus making sense of large volumes of unstructured data. This could be open ended verbatim responses, social media data, or text fields in administrative or customer data.

There are a number of computer processing techniques that can be used to extract meaning and identify sentiment, all of which involve computer programmes which seek to replicate or apply human decisions across the data.

Text analytics can also explore relationships between topics, and quantify the strength of relationships between unstructured text and structured data such as survey responses, market data or other performance indicators – this helps understand what is driving good or bad performance.

Case studies

Merging consumer voices for an airline to improve customer service

Understanding mobile network switching

Automotive deep dive to improve customer satisfaction

Understanding key drivers of satisfaction in higher education