Skip to main content

Vers un système Smart CRM d’analyse prédictive pour la création d’une vision 360 des clients de Safar Flyer : conception et mise en œuvre d’un système de scoring utilisant les algorithmes du Machine Learning

Engineer: Doha Raissouni
Organisation: Royal Air Maroc
Language: French
Promotion: 2019
Year: 3

Abstract #

Customer loyalty is one of the main concerns handled by relationship marketing services, so

for ROYAL AIR MOROCCO as a service company, its ultimate goal is to have how to retain

its customers and especially customers who travel often and which contribute to the

continuous increase of turnover. In this fact, ROYAL AIR MAROC needs to have a complete

vision of its Safar Flyer customers in order to carry out more targeted marketing actions for

the loyalties.

It is in this perspective that the end-of-study project that we developed within the ROYAL

AIR MAROC was set up and which had as its objective: the 360 vision of the client Safar

Flyer, in fact 6 sub- objectives served as a framework in this perspective which are :

➢ Have a culture in marketing and aerial field.

➢ Identify data sources and combinations.

➢ Define the KPIs and computed ones to have meaningful information.

➢ Used the machine learning to predict the churn score.

➢ Calculate the value of the customer’s life.

➢ Create a scoring system based on KPIs, churn, customer value, and RFM segmentation

for 360 vision.

In this fact, we have adopted three methods of collating data and information : the

documentary method that allowed us to search, identify and find separate documents related

to our research topic, the Inquiry to fully understand the need through the interaction with the

professionals of the field and the benchmarking to compare the used tools thus to know which

is the most relevant solution to answer the needs in an optimal way.

In this regard, we adopted an agile method that allowed my tutor to focus on the result.

Indeed, the adoption of this method allowed us to meet the challenges of design and

implementation.