Challenges

1 / Effective segmentation. Accurately segmenting clients to pinpoint the most profitable groups in the tech industry.

2 / Customer retention. Developing strategies to minimize client turnover and strengthen retention.

3 / Data analysis. Analyzing extensive customer data to discern preferences and trends in tech solutions.

4 / Predictive modeling. Constructing models to forecast future client value and tech needs based on current data.

Goals

1/ Identify the most profitable customer segments to tailor technology solutions and marketing strategies more effectively

2/ Develop models reflecting the common attributes of significant clients for future targeting

Results

By segmenting customer transaction data based on brand occurrences, the bank achieved a 15% increase in customer engagement.

The creation of heat maps to track payment card usage geographically led to a 20% improvement in targeted marketing efforts.

The launch of co-branded bank cards, informed by popular brands in transaction data, resulted in a 25% increase in new card sign-ups and a 30% rise in transaction volumes with these cards.

Selling depersonalized transaction data to retailers opened a new revenue stream, enhancing the bank’s profitability by 5%.

Challenges

1 / Data integration. Seamlessly integrating various data sources into a unified data lake.

2 / Transaction monitoring. Managing and analyzing the vast number of daily transactions.

3 / User behavior analysis. Understanding client actions and payment details across multiple channels.

4 / Real-time analysis. Ensuring the timely and effective analysis of incoming data for instant decision-making.

5 / Fraud detection. Identifying and preventing suspicious transactions and patterns of money theft.

Goals

1/ Analyze transaction history, client actions, and payment details for insights

2/ Set up predictive analytics

3/ Enhance the monitoring department’s efficiency with an intuitive interface for rule generation and editing

Results

1/ Our service advanced algorithms that enabled instantaneous insights, allowing for real-time fraud detection.

2/ The implementation of predictive analytics within our service allowed for a 90% confidence level in forecasting customer trends, utilizing sophisticated data modeling.

3/ Rounding loss fraud identification. Detected a unique fraud scheme involving multiple small transactions in different currencies, exploiting rounding differences.

4/ Predicted customer financial trends with a 90% confidence level, greatly enhancing foresight in fraud prevention.

Challenges

1 / Customer segmentation. Effectively using data to segment customers for targeted cross-selling and up-selling.

2 / Optimizing channels for clients. Understanding and optimizing customer interactions across diverse channels.

3 / Customer loyalty. Ensuring high levels of customer satisfaction and loyalty in a competitive market.

Goals

1/ Successfully conduct customer behavior analysis

2/ Use data-driven strategies to improve customer satisfaction and trust

3/ Predict customer departures and create targeted marketing campaigns.

Results

1/ A significant 20% increase in customer loyalty scores, as well as an improvement of 10 points in the Net Promoter Score (NPS). These results stemmed from the analysis and segmentation of the client base.

2/ A 20% increase in customer usage of web and mobile banking channels, achieved through the analysis of the elimination of user barriers.

3/ Development of a model reflecting the common traits of the most profitable customers from past years, so the bank could focus special retention programs on them, enhancing customer value.

4/ Segmentation and development of “customer profiles”. 

These profiles identified groups of customers with preferences for certain banking services, allowing for targeted and efficient advertising and marketing activities.

Business Challenge

Build a secure and scalable data aggregation and analytics cognitive platform that transforms key business processes.

Solution

IBA Group developers worked on the Cognitive Enterprise Data Platform (CEDP). This platform makes it easy to get, share, and integrate enterprise data. Various business units of the company interact with the platform as producers and consumers of data.

Result

CEDP currently contains over 100 enterprise data assets. Datasets include customer master data, product offering master data, unified taxonomy, and more.

CEDP has big data analytics capabilities such as BigSQL, HDFS, DashDB for performance-based analytics, unstructured data annotation capabilities, machine learning, deep learning, and access to IBM Cloud Services.

BUSINESS CHALLENGE

Develop an application to analyze photos made by a drone while flying over rice fields. Count the number of plants that have sprouted on the plots in the photo.

SOLUTION

The IBA Group team implemented image processing algorithms to accurately calculate the number of objects located on them. For end users, we created an application that processes photos, determines the number of plants on them and generates reports in *.csv and *.pdf formats.

Application algorithm:

• Create a polygonal image of the region of interest.
• Download satellite image for it (requires Internet connection).
• Take a calibration photo (taken at the desired measurement height containing two highly visible objects placed on the ground at a distance of 1 m from each other).
• Take a photo above the field for measurements.
• Perform calibration procedure using the calibration image.
• Generate report.

OUTCOME

Thanks to the application, the customer can receive reliable information about the state of rice fields using drones. The project is continuously improving.

Business challenge

Improve the quality of customer analytics and services. Collect additional information from public sources: social media and forums.

Solution

The solution collects data from various external sources. Using semantic and cluster models, it identifies objects of interest. It complements existing dictionaries with business terms, calculates a satisfaction rating based on sentiments in comments and reviews, and visualizes sentiment data. Visualization is based on Qlik Sense, however any BI system can be used.

Result

The solution is applicable to any enterprises and organizations that provide services or sell products to a wide segment of the population, which can be discussed on forums, edicated websites and social media.

Business challenge

The company set an objective to assess the efficiency of budget spending on the advertising of manufactured drugs, and to build an optimal advertising strategy.

Solution

IBA Group experts together with methodologists from the consulting company analyzed the costs of advertising in the media, on TV and radio, outdoor advertising and online promotional campaigns. Five-year data were used to take into account the impact of economic factors and seasonal fluctuation. Specialists created an interpretive forecast model with a seasonal component and identified key parameters that influence brand loyalty.

The results of the study showed that outdoor advertising is greatly overestimated and its efficiency needs to be evaluated more carefully. In addition, advertising a single product in a product line may increase sales of all products in a line – the so-called “halo effect,” when brand visibility due to the main product is extended to cover the entire line.

Result

Marketers obtained new tools: a dashboard for comparative analysis of the sales level and advertising efficiency in various channels, and a calculator for the optimal allocation of budgets for the forecast period. They were used to identify budget limits, above which there’s no point to invest in promotion, because the efficiency of advertising campaigns decreases sharply and does not lead to expected sales.

Financiers and budget owners use accurate data to plan investments in each advertising channel for various estimated levels of advertising budgets.