Data Optimisation Network is an end to end application which connects multiple databases together to create a 360° view for the business on a same geospatial index through using machine intelligence and data science techniques. The demonstration of the app is based on SSE Airtricity. However, the data is anonymised for the solution but depicts the real-life problem.
This research is based on real-life experiences in which a business would have raised questions around New Sales Strategy: How to target the “right” New Customers; Customer Retention Strategy: How to retain Existing Customers; Cross / Up-Sell Strategy: What additional products/services to be offered to which all Existing Customers; Generation Green Strategy: How to undertake and support initiatives to promote greater environmental responsibility?
The Data Optimisation Network (DON) seeks to improve business decision making through use of machine intelligence and data science techniques. It predicts customer behaviour through data-driven insights and enables business to acquire new customers intelligently. The key requirement with most of the companies is not just to acquire more customers but it is to acquire the right customers and retain them for long term. DON is a combination of three core modules: 1) Data Extraction and Manipulation, 2) Advanced Analytics and 3) Visualisation Dashboards.
DON will empower teams such as Sales, Marketing, Customer Value Management and more via interactive Real-Time Dashboards. Furthermore, having all critical and most relevant information consolidated in one place; enhances user experience with ease of information understanding, identification of focus areas and supports decision making process.
The Data Optimisation Network is built in R Shiny utilising data science and machine learning techniques. DON helps to build; an end to end application that empowers users with better understanding of key data points to make strategic business decisions.
DON is an automated interactive dashboard where the first step is Data Extraction and Manipulation in R via the SQL queries and loading other data extracts from the Central Statistics Office Portal.
The Second Step is generation of models such as RFM Model, Propensity to Churn, Propensity to Additional Products, Customer Lifetime Value and Anomaly Detection Model on Consumption.
The address matching to create single address view is built in Python using the FuzzyWuzzy package which obtains the closest matches from the Geo-Directory using a match rate threshold. It uses the Levenshtein distance between the two strings as the comparative measure.
Finally; an interactive dashboard is built as a shiny application to show trends, customer behaviour and segmentation.
DON will empower businesses with valuable data insights; vital for building business strategies and associated action plans to drive business sustainability and growth. Through the innovative use of business data sets, CSO data, and the application of data science techniques and machine intelligence, the DON will work as an automated proactive and reactive solution platform which will enhance business strategic decisions.
This solution demonstrates the power to CSO data in business decision.