Cognitive Sales Analytics - overview
Sales organizations manage a pipeline of deals that they try to close with different clients. The deals for IT services, for example, become contracts to deliver IT services such as software, cloud computing, and mobile computing. These deals go through a sales process of several stages/steps: first, an opportunity is identified with a client, then qualified, validated, and later negotiated, until it is either won by the service provider or lost to another competitor.
Our research project focuses on bringing innovations to enterprise sales domains by utilizing massive business data and novel analytical methods in the area of optimization, machine learning, text analytics and other disciplines of AI and cognitive computing. Example research problems include:
- Deals Win Prediction: Predicting which deals will be won and which will be lost as well as efficiently managing the deal pipeline. Deal win prediction is based on both structured data of deal attributes and unstructured textual data of seller comments after their meetings with the clients.
- Revenue Prediction: Predicting overall quarter-end revenues for all deals that would be signed by the end of that quarter.
- Revenue Change Prediction: Predicting changes in revenues from clients receiving periodic services.
- Up-Sell Analytics: Analysis and optimization of sales and solution offering bundles.
We closely work with business stakeholders to identify key business challenges, formulate them into adequate models, design and implement solutions, and realize them as tools, services and reports for actual use. Our work is published in academic conferences and journals, as well as patents.
- We are honored to have received the second place for the 2016 Innovative Applications in Analytics Award from INFORMS for "Analytics for the Engagement Life Cycle of IBM’s Highly Valued IT Service Contracts".
- We were a finalist for the 2016 best cluster paper award of INFORMS Service Science for our paper entitled: "Optimizing Precision in Machine Learning Models for Actionable Predictions of Revenue Change".