Case Study: XpresSpa
Leveraging data and developing algorithms to predict demand in XpresSpa locations in airports nationwide
XpresSpa is the world's largest airport spa company, offering services that are tailored specifically to the busy consumer.
XpresSpa is committed to providing exceptional customer experiences with its innovative premium spa services, as well as exclusive luxury travel products and accessories through partnership with some of the leading cosmetics brands in the world. XpresSpa serves almost one million air travelers each year at its locations in the United States, Holland and the United Arab Emirates.
XpresSpa has over 750 employees, including talented teams of professionally licensed massage therapists, cosmetologists and nail technicians who are committed to providing exceptional customer experiences.
By nature, airport foot traffic can be unpredictable and volatile. Factors such as weather patterns, flight delays, cancellations, and the number of flights in any given week may vary drastically. This presents a challenge to airport vendors to forecast sales on any given day.
The inability to accurately forecast daily sales presented XpresSpa challenges that revolved around the inability for spa managers to react to unexpected demand spikes, adopting a universal, streamlined process to manage operations and getting a better, more wholistic view on sales data.
Icreon was able to partner with XpresSpa to analyze the relevant data points that directly and indirectly affect the throughput of spa customers on any given day. Factors such as weather patterns, number of flights, the number of seats on each flight, delays, and cancellations. After an initial analysis of sales data to determine how the data should be structured, the data was cleansed to remove missing values.
We identified external data sources for flights, weather and traffic then integrated with these sources via APIs. Once the data was identified, it was separated into different tables to form the appropriate relationships between all data sets where we built models for individual stores in each location using backwards elimination and stepwise regression. The tools used were Power BI, SQL, R, and Python to achieve the desired outcome.
Our team built out an algorithm that predicted the sales of a location based on specific coefficients and variables. The end result was an end-to-end automation, communication and analytics platform that effectively standardizes how XpresSpa actualizes potential from each of its locations.