AustraliaMedia and entertainment2023 – ongoingTabcorp
Machine-learning models for customer offers and safer-gambling interventions
Tabcorp's offer-allocation models cut its A$385M annual generosity budget by 12%
Problem
Tabcorp's customer data sat in separate legacy systems. Piecing together a customer journey took days, which delayed decisions on offers and customer care.
Approach
Tabcorp built a single customer view on the Databricks platform and used MLflow to manage its machine-learning models. The models included a Next Best Experience recommendation engine, a daily allocation model that fills in-app offers, and a player risk model for safer-gambling interventions. The company set up an AI Next Lab in 2023 and retired seven legacy systems. The number of models in production grew from five to more than 40.
Results
Takeaway
Tabcorp tied its models to a budget line it already tracked, the offers it gives customers. That made the saving visible in the annual report.
Categories
- Industry
- Media and entertainment
- Business function
- Marketing and sales
- Innovation type
- Process innovation
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks describes how Tabcorp replaced siloed legacy data with a single customer view and added machine-learning models for offers, recommendations and player risk. It reports a 70-90% gain in generosity efficiency and a 12% cut to the A$385M annual generosity budget, citing Tabcorp's FY24 Annual Report. It also reports that models in production rose from five to more than 40.
Full details are in the original sources.
