In 2017, one of the biggest banks in Poland reached out to Neontri. They had an innovative project in mind that would help them accommodate clients with their help and services.
Objective
Simply put, the client needed a Big Data platform for analyzing customer behavior.
With the platform’s help, the bank aimed to better adapt to customers’ needs and demands. Implementing such a solution would protect the bank from sharing their clients’ data with a third-party company.
Expectations
At first, our client needed only to collect information concerning the usability of specific app functionalities and analyze how customers move around the app. The aim was to understand clients’ needs and provide them with improved experiences.
Outcomes
We proposed an architecture for the solution and implemented the basic set of reports. We focused on visualization and the option to extract more detailed data. Our proposal allowed the analytics team to connect different data sources in various formats comfortably.
We’ve delivered a set of clickstream reports related to the bank’s client database, which were visualized through Sankey diagrams, pie charts, line charts, and tabular reports.
The Big Data platform we developed was BEHEX. It allowed the bank to collect customer activity data from a wide range of internal systems. To achieve this, BEHEX linked individual customer numbers with their actions. It could then track their general UX flow across different bank applications.
Additionally, BEHEX was able to generate detailed cross-sectional reports. These reports provided the bank with valuable insights, such as which mobile app options their clients were using and their habits. The bank also gained access to data regarding their mobile applications, including which buttons and screens generated the most conversions.
BEHEX itself consists of three key components. First, a Presto database that allows for the creation, analysis, and storage of valuable data. Second, various interfaces enable BEHEX to track clients across the bank’s internal systems. Third, tools for generating detailed reports were implemented.
Most importantly, our client gained the ability to analyze their customers’ behavior. With BEHEX, they could track which elements of their applications people interacted with and which buttons, screens, and forms increased engagement and conversion rates. Such data is incredibly useful for optimizing marketing campaigns, improving user experience, and maximizing profitability.
Process
We provided a dedicated solution that addressed the client’s specific needs. We carefully analyzed the bank’s needs and looked for opportunities to provide a product that could help the business grow. We ensured it was easily scalable and could be developed further by adding new features. We’ve also taken care of end-to-end implementation.
Challenges
The entire project was an incredibly complex endeavor. The tool we were tasked with developing had to record the actions of hundreds of millions of records simultaneously. Then, this data had to be analyzed and presented in an organized format, such as a pie chart or a table. We had to combine various data sources, such as logs, databases, and text files, to prepare valuable reports in a short time.
The challenges we faced contributed to creating a much better product than initially expected. We enabled the tool to develop further and provide even better, more insightful data for the bank to use and learn from. This gave them huge business opportunities.
Cooperation
We’ve used the agile methodology to iterate the product and communicate with the client regularly about the development process. Our team consisted of four people from Neontri and four people from the bank.
Technology
The key technologies used in the project were Open-Source Big Data solutions, custom database connectors, SAP Lumira, and AI to show results.
Results
We delivered a solution that allowed the client to analyze how often users utilize specific options and how they use advanced app functionalities, what is problematic for them, how they find what they need, and when they stop using the app.
The tool we created also allowed the bank to group customers based on selected characteristics and analyze their behavior. The data could be aggregated, sorted, and grouped according to customer category, app versions, day of the month, and other parameters. This way, the client could use the data to support the sales processes, increase app safety, and identify when a sale was concluded.
BEHEX was a base for further development as it’s been used for many years. Of course, systems, functionalities, and databases are being updated with time, but they still use the solution we provided. This proves our software is made with potential for growth and scalability.
Dorota Jasińska
