Predictive CRM Analytics Implementation - 15 days

Predictive CRM Analytics

Digia Predictive CRM Analytics is a service concept for predicting customer behaviour using machine learning algorithms.

Two prediction use cases are in primary focus:

1. Prediction of winning of a sales opportunity Model calculates probabilites for an opportunity to be won in pre-defined future time period. In implementation, all open sales opportunities are scored periodically (e.g. once a week) as batch. Individual open opportunities can be scored whenever triggered.
2. Prediction of customer churn Model calculates probabilites for customer churn. Meaning of ”customer churn” may vary depending on the nature of business, and it must be defined exactly before modeling. In implementation, all customers are scored periodically (e.g. once a week) as batch. Individual customers can be scored whenever triggered

NEED: Selected use cases can benefit decision-making in several ways :

• Gain understanding which factors are behind successful sales
• Gain understanding which factors increase risk of customer to leave
• Time savings : machine-learning based customer/opportunity scoring is automatic, fast and consistent
• Optimize resource and time usage with ranked customers: o put effort on the most likely win sales cases o target actions to the customers who are most likely to leave

SOLUTION AND DELIVERABLES : Predictive analytics implementation in customer’s MS Azure environment • Refined scoping of the predictive CRM analytics in customer business context • Identification of data sources • Data modification work (joining, aggregations, derived variables) in Microsoft Azure Data Platform for predictive modeling purposes. • Machine learning algorithm implementation for the selected use case. • Pipeline for scoring customers / opportunities in production with created predictive model

PRICING AND SCHEDULE Pricing is estimated, final pricing according to customer scenario complexity and requirements.