Know what your customers say!
Client: A Fine Dining Restaurant in New Delhi, India
Time: April 2015
————————–
Problem Statement
A fine dining restaurant in New Delhi, India failed to achieve a more than 10% growth rate over the past two years.
Approach
No one so far had analysed the thousands of reviews they had received online/offline over the years. So we decided to focus on the reviews. In doing so, we asked the following questions:
- What did people say about the restaurant?
- What were the keywords and topics of discussion peppering the comments?
- What aspects of the restaurant would they like to see improved?
- Did the customer demographic affect the way customers perceived the various operational aspects like taste of food, ambience, music, lighting, quality of service etc.
This project involved a heady blend of text and structured (customer demographics, POS sales information) data, which enriched the data in a multitude of ways. We used some advanced techniques of text analytics like machine learning for topics classification, sentiment analysis, keyword parsing and relations extraction.
Outcome
Most (80%+) customers were satisfied customers. Even so, a significant minority also raised pointed questions and asked for specific improvements in the day-to-day operations. Furthermore, we also discovered that it’s not recommendable to rely solely on ratings – which tend to paint a misleading picture. Customers might well give you an entirely commendable rating of 8 or 9 out of 10, but tear you to pieces in their reviews.
Based upon these findings, we suggested operational improvements in eight areas of fine dining; menu design, non-vegetarian fare, room temperature, music, staff interaction, drinks, portion sizes and speed of service.
Impact
The client has taken up our recommendations and made appropriate changes. We are currently awaiting the next round of reviews to help quantify the impact of the changes made.
