E commerce service provider

Overview:

An e-commerce website provider was concerned about the customer behaviour in the websites. They had the review system in place. However, going through 1000’s of reviews wasn’t practical. The client was focussed on improving and making the lives of their customers easier. The client wanted to extract the data and understand what their customers thought of their services and product catalogue.

Challenge:

The reviews were plain texts. Understanding from the data at the moment was not set up. No analytics department was set up due to the budgeting. However, a small budget was allocated to get a PoC done for it.

Implementation:

First things first, connect an API to extract the data from the data source. A lot of data had to be segregated based on the metrics, measures and dimensions. The sentimental analysis algorithm was used to measure the magnitude (on what scale is the negativity or positivity of the comment is) and the indicator (whether at all the comment is positive or negative). The report was created to display the percentage of positive and negative comments by percentage along with the positive and negative words by its usage.

Benefits Achieved:

• 3-day delivery
• Budget approved
• Reports in fingertips (tableau on mobile)