Examples of completed projects
Data Science
Increasing the effectiveness of call center calls
Before: Calling to customers completely at random – on random days and hours.
After: Optimal days and part of the days of contact for each customer segment.
Result: Efficiency of the call increased even 2 times for some segments.
Before: Calling to customers completely at random – on random days and hours.
After: Optimal days and part of the days of contact for each customer segment.
Result: Efficiency of the call increased even 2 times for some segments.
Identification of agents with high risk of loss
Before: A simple decision expert model;
identifying transactions with potential high operational risk,
using basic data on a single transaction.
After: Automatic process of assessing the agency’s quality on the basis
of its current behavior and sales portfolio
Based on a complex algorithm identifying the risk of high operating costs
Result: Reduction of operating costs by 30%.
Before: A simple decision expert model;
identifying transactions with potential high operational risk,
using basic data on a single transaction.
After: Automatic process of assessing the agency’s quality on the basis
of its current behavior and sales portfolio
Based on a complex algorithm identifying the risk of high operating costs
Result: Reduction of operating costs by 30%.
Analysis of the impact of competition on sellings
Before: Noticeable decrease in selling after the opening of competitors’ facilities
No information for which product groups there was an impact on selling.
No information on how the distance of a competitor’s facility affects sellings
After: Ability to reliably compare the level of influence of competition per store category
Determination of expected decrease in sellings on specific product groups
Result: Possibility of protection against a selling drop at the level of the facility.
Before: Noticeable decrease in selling after the opening of competitors’ facilities
No information for which product groups there was an impact on selling.
No information on how the distance of a competitor’s facility affects sellings
After: Ability to reliably compare the level of influence of competition per store category
Determination of expected decrease in sellings on specific product groups
Result: Possibility of protection against a selling drop at the level of the facility.
Reduction of collection costs in a telecommunications company
Before: Manual selection of one out of 4 collection strategies for a given client
Based on 2 simple criteria (amount and time of arrears) and the intuition of the decision-maker
After: Automatic vindication strategy selection process based on many criteria and maximizing profit
Result: For 40% of customers, internal debt recovery proved to be unnecessary and ineffective; Profits from debt recovery increased by 20%.
Before: Manual selection of one out of 4 collection strategies for a given client
Based on 2 simple criteria (amount and time of arrears) and the intuition of the decision-maker
After: Automatic vindication strategy selection process based on many criteria and maximizing profit
Result: For 40% of customers, internal debt recovery proved to be unnecessary and ineffective; Profits from debt recovery increased by 20%.
Increasing sales on the smartphone platform
Before: Random selection of several products for the screen viewed by the customer
No hint at global level
After: Automatic selection of products adapted to customer profile
Result: 30% conversion of recommended products.
Before: Random selection of several products for the screen viewed by the customer
No hint at global level
After: Automatic selection of products adapted to customer profile
Result: 30% conversion of recommended products.
Credit risk model for a financial institution
Before: A simple decision model
After: Automaticmonitoring ofquality of scoring model
Automatic model quality analysis
Result: Using advanced modelling methods, a more efficient model was prepared than the basic model.
Before: A simple decision model
After: Automaticmonitoring ofquality of scoring model
Automatic model quality analysis
Result: Using advanced modelling methods, a more efficient model was prepared than the basic model.
Price sensitivity model
Before: A simple decision model based on expert rules, using
basic data about the customer
After: Automatic price sensitivity process indicating optimal
price for the product, based on a complex algorithm identifying price that maximizes profit
Result: Increase in profit by 8-12% depending on the product group.
Before: A simple decision model based on expert rules, using
basic data about the customer
After: Automatic price sensitivity process indicating optimal
price for the product, based on a complex algorithm identifying price that maximizes profit
Result: Increase in profit by 8-12% depending on the product group.
Increasing the effectiveness of additional sales in a financial institution
Before: A simple decision model based on statistical linear model
Using monthly data
Based on data available in only one department
After: Automatic process identifying customers most likely to buy at any given time
Maximizing conversion
Result: 10-20% increase in conversion depending on the customer group.
Before: A simple decision model based on statistical linear model
Using monthly data
Based on data available in only one department
After: Automatic process identifying customers most likely to buy at any given time
Maximizing conversion
Result: 10-20% increase in conversion depending on the customer group.
Business Consulting
Revitalization of the bank's sales call center
Przed: Fragmentary, with no control over the rest of it, manual, with bottlenecks, without regular reports, without KPI standards, without a complete organizational structure, low employee motivation.
After: End-to-End – with full control. The automatic call-in process is fully configured. Managed through a clear corporate structure, accountabilities, reporting, and standards. The strong motivation of employees through the introduction of tools, training, and continuous coaching processes.
Efekt: 10-fold increase in sales performance – record-breaking call center sales results in the Polish banking.
Przed: Fragmentary, with no control over the rest of it, manual, with bottlenecks, without regular reports, without KPI standards, without a complete organizational structure, low employee motivation.
After: End-to-End – with full control. The automatic call-in process is fully configured. Managed through a clear corporate structure, accountabilities, reporting, and standards. The strong motivation of employees through the introduction of tools, training, and continuous coaching processes.
Efekt: 10-fold increase in sales performance – record-breaking call center sales results in the Polish banking.
Implementation of context-specific sales at a financial institution
Before: Contact channel selection based primarily on individual channel handling capability. Consistent and standard communication of products to all customers. Same sales process for all customers.
After: Adequate selection of communication channels. Presentation of the product in channels based on the description of features identical to the customer profile. Fully automated lead distribution process with communication profile.
Efekt: Increase conversions on individual shares from 30% to even 300%. Increase in credit card sales by 20%.
Before: Contact channel selection based primarily on individual channel handling capability. Consistent and standard communication of products to all customers. Same sales process for all customers.
After: Adequate selection of communication channels. Presentation of the product in channels based on the description of features identical to the customer profile. Fully automated lead distribution process with communication profile.
Efekt: Increase conversions on individual shares from 30% to even 300%. Increase in credit card sales by 20%.
Reorganization of a midsize service company
Before: Flat organizational structure, overlapping tasks, lack of incentive systems.
After: Efficient organizational structure, clear tasks, and incentive schemes.
Efekt: Increased operational efficiencies, faster project implementation.
Before: Flat organizational structure, overlapping tasks, lack of incentive systems.
After: Efficient organizational structure, clear tasks, and incentive schemes.
Efekt: Increased operational efficiencies, faster project implementation.
Automated evaluation of subcontractors' financial credibility
Before: The manual and time-consuming process of verification of contractors, based on individual and subjective assessment. Difficult access to financial data about contractors. Use of limited information.
After: Automatic analysis of financial data of the contractor and calculation of the most important measures and indicators and trading limits. Immediate knowledge of the current condition of the company. Periodic monitoring of the contractor’s situation.
Efekt: Shortening the process of analyzing a potential contractor from a few hours to a few minutes. Reduce financial risk and improve liquidity. Higher work efficiency and lower operating costs.
Before: The manual and time-consuming process of verification of contractors, based on individual and subjective assessment. Difficult access to financial data about contractors. Use of limited information.
After: Automatic analysis of financial data of the contractor and calculation of the most important measures and indicators and trading limits. Immediate knowledge of the current condition of the company. Periodic monitoring of the contractor’s situation.
Efekt: Shortening the process of analyzing a potential contractor from a few hours to a few minutes. Reduce financial risk and improve liquidity. Higher work efficiency and lower operating costs.
