Transformation of the bank's branch network

Aim

Global transformation of the bank's network of 400 offices

Context:

The largest private bank in Russia is implementing a global network transformation project within the framework of the phygital concept, which takes into account the trends of digitalization and offline customer service. The project involves the reformatting of 400 bank offices across the country. The network required transformation due to the changing format of banking services, digitalization, optimization opportunities, as well as the potential accumulation of" obsolescence " of the existing format over the next years.

Key indicators

Number of branches of the network – 400 offices
The average service life of offices at the time of transformation is 6-7 years
The number of data layers for the project is more than 300
Project implementation period – 3 years
Potential optimization reserves – more than 200 million annually.
case-key

Solution

First of all, the Marketing Logic and bank teams combined all the data available and necessary for the project implementation: both internal and external. Given the innovation and high technical level of the financial organization, it took a little time to bring all the data layers to a single digital format, even taking into account the large scale of the project. Based on the described business processes and the collected data layers, the ML developers built a complete digital copy of the bank's network – they created a "digital twin". It takes into account the quantitative, qualitative, and financial characteristics of the network from the addresses of branches and staff to data on sales of banking products and functional modules that make up each office of the network, marketing expenses, rent, repair and relocation plans, customer segments, as well as a large number of other intra-bank indicators. The digital twin is necessary in order to see in its entirety the various variants of forecasts offered by neural networks trained during the project on the retrodata on the work of the branch network in previous years. To improve the accuracy of forecasts, ML uses the voting technology of an ensemble of predictive models. Models "vote" for certain scenarios and decisions when planning, which allows you to strengthen the overall expertise.

The forecast is built "from the bottom up", from the smallest components to the overall economy of the network and the profitability of the bank. First, all the calculations are carried out, then they are digitized into money, which allows you to evaluate each component and stage from a financial point of view: both investments and returns. This approach eliminates the need to draw up a detailed financial plan at each stage, since all calculations are initially available. At the same time, all top-level indicators have an action plan with step-by-step elaboration.

For opening, moving and reformatting offices, the Atlas geoinformation system generates solutions based on the commercial premises rental market, which eliminates recommendations for areas where opening is technically impossible. All network changes take place taking into account long-term strategies and principles of maintaining the existing customer base. At the same time, the strategy takes into account the digital ecosystem, ATMs and new office formats as ways to support and serve customers.

Result

As a result of using the geomarketing approach, Geonet and Atlas GIS platforms, big data analysis and machine learning, the bank's team and Marketing Logic created more than 500 scenarios for the development of the network, taking into account macroeconomic indicators, the bank's development strategy, digitalization processes, and trends in the transition of banking products and services to online. The "smart" approach to transformation allows you to optimize the area of offices without losing customers, move and update offices, taking into account the savings on rent, repair costs, as well as marketing support in advance. For 1 year of the project implementation, the bank has optimized more than 200 million rubles.