Decision Augmentation Platform

About

AI for telcos Sourse accelerates AI adoption using prebuilt models engineered specifically for telcos using telco data. AI is used for decision support with strategic planning, and for ‘always on’ optimisation of KPIs such as churn. Acquisition: Expected lifetime value and financial data is modelled alongside marketing, customer, and behavioural data. This data is used to define target customer segments using predictive lifetime value so that richer lookalike audience segments can be targeted by digital marketers. Yield: Causal factors along with historical data, customer data, and behavioural data are usedto describe yield opportunities and recommend treatments to optimise cross-sell, upsell, and loyalty offers for all customers simultaneously. Retention: The predicted lifetime value of every customer is calculated and used to predict the future value across the entire customer base. The reasons that positively and negatively impact customer loyalty and the best loyalty offers to impact retention are available to customer care, marketing, and other stakeholders. Churn: Sourse predicts the reasons that individual customers will churn using client data and competitive insights and is able to calculate the expected timeframe. This allows clients to intervene at the earliest point to avert churn. This data is used to dynamically cluster customers and prioritise treatments of the highest value or most influential customers. Forecasting and anomaly detection: AI is applied to learn complex over time trends and seasonal changes to predict future business performance across all KPIs and all product and audience types. The application of AI allows real time forecasting and up to the minute prediction on whether KPIs will be achieved. Predictive Insights: Each AI model is explainable at a population and individual level to support planning and use across the customer lifecycle. This unlocks a detailed understanding of historical trends and changes, reasons that impact customer behaviour, previous activities that have impacted each KPI, and an understanding of the clusters of customers that are described within the data. Scenario planning: What-if analysis to simulate decision making. Decision makers provide simple information to allow neural forecasting and causal machine learning to recommend the best strategic decision or competitive response. Recommendations provide insights into product changes, pricing changes, optimal budget to be allocated, and the optimal delivery to different customer cohorts. Experimentation: Small scale testing to validate AI insights and treatments to improve the probability of success. Experiments can be easily run through client’s digital channels to elicit a statistically reliable prediction of success before committing budget and resources to deploy large scale treatments or optimisations. Optimisation: Automated management of a KPI so that many strategies can be simultaneously ‘on.’ The application of AI allows self-learning so that continuous improvement is automated. Uplift modelling is applied to cause maximum impact.