The client wanted to deliver tailored offers or a passenger personalized and was looking for an opportunity to Cross-Sell, Up-Sell, & gain customer loyalty. IGT helped with advance analytical solutions that helped them identify the opportunity area for upsell, cross-sell and deliver tailored offer or solutions to the customer.
Share this Post
The client was looking for a partner to help them:
- Identify potential opportunity areas using insights from the data analysis for revenue growth and develop a platform to upsell, cross-sell and target dormant passengers
- Ability to enable crew or the passenger’s backseat screen to deliver the tailored offer to the passenger.
- To focus on passengers past behaviour of inflight purchases and deliver personalized offers.
- Increase in revenues by 20% in inflight retail purchases from last year
- Setting up of a reusable data analysis platform for future model building
- Product recommendation which has a close similarity of the past purchases. Ex. perfumes, jewelry, watches and accessories
The client is the largest airline in the Middle East, operating over 3,600 flights per week to more than 150 cities in 80 countries across six continents.
The client was unable to make full use of the right data due to gaps in processing, aggregation and data analysis capabilities. As a result, they failed to address passenger delight and deliver tailored offers or a passenger personalized solution, which was resulting in losing opportunities of Cross-Sell, Up-Sell, & customer loyalty.
- Transformed weak quality data from source systems to a consumable data by writing data pipelines which were primarily a Data Engineering activity
- Setting up Azure Databricks environment
- Creating visualization for the business to represent the product affinity predictions in an intuitive manner
- Established predictive model to deliver an affinity score for a customer against every brand and category of a product in the inflight catalogue
Python, R language, Spark/Pyspark, Dataiku, Databricks
- Supervised Machine Learning Models
- Unsupervised Machine Learning Models
- Time-Series Models