Société Générale Group bank uses big data analytics to manage network
We share with you the experience of our projects and respect the NDA
Objective:
To increase the efficiency of a subsidiary's network in the Eastern European market.
Context:
The group's subsidiary bank in Eastern Europe has suffered an efficiency slowdown in managing its office network.
Key indicators:
Over 300 branches. Over 4,000,000 customers. Increasing efficiency more than 18%
Solution:
In order to increase efficiency, the company decided to digitize its processes. All data that might be useful were brought to a uniform digital format. The data included depersonalised customer data, data on the company's offices, their footfall, sales for each product, and a customer portrait in each office. GML experts assessed the level of office visibility for the target audience, potential advertising efficiency, and marketing and product sales potential.
The company began implementing new location intelligence technology to manage the branch network. This location intelligence technology aggregates data on all bank branches and departments, assesses the potential and load, calculates the efficiency of prospective offices based on customer activity data, competitor banks, population size, traffic on city streets and other statistical information. As a result, the bank has a "heat map" for each city where it operates along with an assessment of branch location potential at a walking distance (100 m) proximity level.
How does this happen? AI models—neural networks—are trained using the collected data. The AI then uses the same criteria to assess all of the locations in the city and region. During the second stage, the best locations individually and their combinations are compared, because an ideal network is an optimal network that works better than the competition, attracts targeted traffic, and in which points don't interfere with one another.
The project has been underway for several years. Using high-tech products, GML handles the opening and relocation of bank branches. Applying advanced solutions makes it possible to maintain a high development speed and use a scientific approach to improve efficiency in network management. Geodata and predictive models allow for more accurate and faster decision-making, minimizing errors, and planning actions while taking into account the location potential of each sales point.
Result:
The project has already increased the network's efficiency by more than 18% and continues to harness the potential for development. The unique feature of AI models is that system training and new data help increase prediction accuracy and improve results. This allows the companies that started using AI solutions earlier to maintain a competitive advantage over many years.
© 2019 - 2024 Global ML
