Text Analytics

Student: Audrey Shannon, 2019-2020

Currently, there is no single solution that the global McDonald’s has leveraged to automatically source, parse, and analyze the increasing amount of data. The current solution is simple because it is all digital. However, not all McDonald’s restaurants use the digital app, hence there is no existing solution for these restaurants. For those that do have a current solution, it involves a lot of manual, repetitive work that is not efficient.
Looking at some numbers, every month, McDonald’s Customer Service Centers field more than 10,000,000 points of feedback globally in the form of emails and phone calls. In addition to this, millions more data points are generated via social media, app store reviews, and survey responses. It is estimated that McDonald’s currently only captures and quantifies about 10% - 15% of this data for Corporate consumption and only 3% is analyzed programmatically for customized machine classification, mostly from digital sources (mobile app, app store, survey SDKs). Hence, the goal of the project is to bridge the gap and bring custom classification up to 15% and consolidate data sources to increase the penetration into available data to 30% - 40%.