The leading global food service retailer wanted to more quickly and accurately interpret customer experience data for improved decision-making. While unstructured text and natural language data can provide actionable insight, data from customer satisfaction surveys and social media comments were analyzed by different text analytics solutions, resulting in disparate outputs. Written comments were processed as a whole, leading to less accurate results than analyzing sub-comments that contain unique sentiments and tones. The QSR\u2019s consumer Insights and social listening teams sought a more robust and flexible text analytics solution to effectively uncover sentiment and context at scale.<\/p>\n
Wavicle implemented ActiveInsights\u2122, a text analytics solution that integrates raw data from multiple sources, such as real-time social listening and customer satisfaction surveys, in an Amazon Redshift database. ActiveInsights\u2122 captured the full scope of written feedback by first breaking complex sentences into sub-comments, then assigning a positive, negative, or neutral sentiment to each referenced topic, ensuring consistency across categories and themes. To preserve information and minimize information loss, the data model captured and replaced all previous references and co-references, offering tremendous visibility and insight into millions of surveys and comments already collected.<\/p>\n
Approximately 95,000 user comments are processed daily by Talend cloud data integration software and Amazon Comprehend, a natural language processing service that uses machine learning to find insights and relationships in text specific to the QSR\u2019s topics and sentiments. The output is transformed into dashboard intelligence and ad-hoc reporting by Tableau data visualization.<\/p>\n
Wavicle\u2019s ActiveInsights\u2122 helped the global food service retailer to improve the speed, completeness, and accuracy of customer feedback analysis. The new text analytics solution processes unstructured text and natural language data\u00a010x faster<\/strong>\u00a0than before, with up to\u00a070% greater accuracy<\/strong>.<\/p>\n With the ability to gauge the sentiment and tone of complex comments across myriad specific items, themes, and categories, the QSR\u2019s consumer insights and social listening teams can quickly discover trends in customer experience with little to no lag time, driving meaningful action.<\/p>\nOriginal Text Analytics Solution<\/h2>\n