A global household appliance manufacturer wanted to create a new steam iron that met its customers’ needs and addressed their pain points.
Objective
To mine social media and product review posts to identify the key product features that customers liked and disliked and the key drivers impacting customer experience and perception.
Approach
- Collected reviews and ratings data from Amazon.com, Amazon.co.uk, Walmart.com, and Currys.co.uk, as well as tweets regarding the top five steam iron brands.
- Cleaned and pre-processed data for text analytics (i.e., cleaning, tokenization, removing stop-words, stemming, lemmatization, creating n-grams, and vectorization).
- Used bag-of-words and VADER sentiment analysis to identify key topics, categorized into user experience, product features, etc., and respective customer sentiment.
- Leveraged ML models to identify product features that impacted customer rating.
Outcome
- A shortlist was developed of the five most significant factors impacting customer experience and perception, along with their relative weighting.
- The company developed product concepts that focused on these key factors, resulting in a more targeted and successful concept test.
Algorithms Used: NLP, VADER, and bag-of-words analysis
