9:30 - 10:30
October 25, 2018
Grand Cypress A
101: A Portfolio for the Utility of the Future: Using Predictive Data Science To Optimize Entergy’s Products and Services Portfolio and Manage the Modernized Grid
Session Category : Grid Operations
This project was designed to leverage cutting-edge analytics to address three objectives:
- Develop a fresh, profitable portfolio of customer product and service offerings
- Prioritize offerings that increase customer engagement, generate new top line revenues, and improve customer satisfaction
- Understand the impacts on load and the distribution grid over time
Because limited data was available on a broad portfolio of products and services, Entergy conducted a survey to explore the participation potential of new offerings. To create an analytics-ready data set, Entergy’s survey and other customer data was fused with a third-party data set set rich with demographic, financial, lifestyle, behavioral and structural attributes on every household and business in the United States. Using machine learning analytics, Entergy developed product adoption rates based on propensity models for each household and business relative to that individual entity. Customers were segmented for each of 30 products and services in Entergy’s portfolio.
In parallel to this new products and services study, Entergy was conducting a grid modernization study which included the objective of more responsively planning for anticipated DERs at a feeder level. By developing propensities and load forecasts at the customer level, the results of this study could be aggregated up to each feeder. As such, the grid modernization group was able to include data from this study’s findings in its treatment of PV penetration, DR and EE impacts, and other elements of the planned product and services portfolio in its long-term planning activities for grid asset upgrades and replacement.
Using predictive data science methods, this study allowed the grid modernization team to provide input into Entergy’s products and services strategy on which offerings would be most beneficial at the feeder level and helped shape not only the portfolio but also the regional roll-out strategy.
Predictive data analyses, when fused with detailed household and business data, can also be designed to improve the productivity of customer acquisition, for which propensity to participate is developed for each household and aggregated to the feeders of most interest to grid modernization planning.