Erste Group deploys FICO optimisation for retail lending
Erste Group has deployed FICO optimisation technology for lending and collections decisions across its markets, supporting tailored financing terms for retail customers.
The rollout covers mortgages, cash loans and other retail products, and is part of a wider programme Erste Group has been developing for nearly 14 years. The bank operates in Austria, Croatia, Czechia, Hungary, Poland, Romania, Serbia and Slovakia.
Erste Group uses mathematical optimisation to improve pricing and credit-limit decisions that were previously based more heavily on expert judgement. According to the bank, the models rely on the financial parameters of each financing case and do not use personal customer data.
The approach is intended to support risk management, enable more precise financing decisions and strengthen measures to prevent over-indebtedness.
Loan pricing
One of the clearest examples is pricing for small-business unsecured instalment loans. Under the previous set-up, nearly 90% of loan pricing was driven by branch-level price exceptions, with a heavy reliance on manual decision-making.
By combining machine-learning models that predict client behaviour, including take-up, prepayment and risk, with mathematical optimisation, Erste Group says it has been able to introduce more individualised pricing at scale. The change significantly reduced manual exceptions and delivered a 22% profit increase.
The broader use of optimisation has also produced measurable improvements for clients while supporting internal goals around profitability, customer satisfaction and operational efficiency. To widen adoption, Erste Group created an Optimization Expert role to guide analysts through model development, data preparation, deployment and evaluation.
Industry recognition
The cross-border nature of the project appears to have contributed to external recognition. Erste Group received a 2026 FICO Decision Award for AI, Machine Learning & Optimization.
FICO supplies the optimisation technology used in the project and describes Erste Group as among the more advanced users of the approach in retail banking. The US software company sells analytics and decisioning tools to financial institutions and other industries.
"Erste Group Bank has consistently innovated in its use of mathematical optimization to power smarter decisions," said Nikhil Behl, President, Software, FICO.
"They have proven results in multiple areas, demonstrating their excellence and the gains that this technology provides across the enterprise," said Behl.
"FICO considers them one of the most advanced users of optimization in retail banking today," added Behl.
A judge for the awards programme highlighted the bank's ability to apply work developed in one market across several regions, suggesting the project has moved beyond a single-country pilot into a broader operating model.
"The judging panel was impressed by how Erste Group Bank extended results from one country to multiple regions," said Shrimanth Adla, Senior Director, Credit Risk Strategy and Analytics, Comcast.
"They have made AI-powered strategies into an analytic advantage," added Adla.
Erste Group employs about 55,000 people and serves around 23 million customers through more than 2,100 branches in its core markets.