Project - Social Media Sentiment

Using Natural Language Processing and text analysis to evaluate customer sentiment about specific campaigns and customer journeys.

How can we get an unvarnished view of how we are doing, direct from our customers?


Social media provides a direct, real-time and unvarnished view of your customer experience rather than insights from complaints or small sample NPS scores.

Our value goal was to identify social media content relating to specific customer journeys and also understand point-in-time customer experience and trends from sentiment expressed. Natural Language Processing and text analysis was used to evaluate customer sentiment about specific campaigns and customer journeys.  


Changed views on how to create exceptional customer experience and influenced marketing strategy. For the first time, our client didn’t only see complaints, but also feedback on parts of the process that truly delighted their customers.

One of the most surprising insights was the strength of positive feeling about the welcome gifts given to new mortgage customers… our client had intended to stop this campaign but quickly reversed their decision!

Our approach

Classify each post

We applied Natural Language Processing to the text in social media posts to classify them based on the customer journey or marketing campaign they related to.

Measure the sentiment

We performed text analysis using customised sentiment libraries to score each post based on the tone of voice and types of words used.

Visualise trends

We created visual representations of changes in sentiment over time, overlaying campaigns and events relating to different journeys in order to understand their impact

This site uses cookies and by using this site you are consenting to this. Find out why we use cookies and manage your settings here.