Spotlight addresses the challenge of identifying Priority Services Register (PSR), Fuel Poor (FP), and Leaving No One Behind (LNB) customers at a household level, surpassing current methods limited to high-level demographic trends. To do so, the project seeks to access data from sectors such as telecommunications and finance to enhance customer identification. Additionally, the project aims to optimise engagement strategies with PSR, FP, and LNB customers. By utilising data insights, it plans to enhance operational team interactions through effective engagement channel selection, tailored to various customer needs and vulnerability categories.
Benefits
Benefits:
This project will deliver substantial societal advantages for vulnerable customers, including reduced stress during outages, relief from financial debt, and a contribution to reduced carbon emissions through participation in the energy transition. Moreover, Spotlight will also depict the intersections amongst essential datasets that impact both established and emerging segments of consumer vulnerability. These insights will provide valuable guidance for addressing intricate tasks or challenges. Furthermore, current customers will also gain from this initiative, as adding them to the PSR program grants access to extra support during power interruptions, including a 24/7 helpline and proactive outage updates.
Learnings
Outcomes
The project has developed models targeting PSR, FP, and LNB customer identification and engagement. The first round of customer engagement has been completed to validate the PSR model and test a combination of various channels and message content across different segments. Using the email channel just over 88k customers were contacted, we saw that over 28k (32%) emails were opened, with a sign-up rate of 26% of those that opened the email. The SMS channel delivered over 27k text messages, 4.5k customers (16%) actively engaged with the SMS and clicked on the links, the sign-up rates were just over 7% for the delivered SMS and around 44% of all those that actively engaged with the SMS. This data will be used to refine the message content and the channel used for each customer segment that will then be used to increase customer engagement in the next round of PSR engagement.
The outputs from these project activities will facilitate improved support for vulnerable customers, offering actionable insights into successful FP initiatives and effective assistance during the energy transition.
Lessons Learnt
The project has identified several key insights:
Early identification of data access challenges
It is crucial to identify any potential challenges or delays in accessing required data sources early in the project. This allows for prompt action to resolve issues and prevents setbacks that could impede progress.
Importance of data quality
Data quality management should be prioritised from the beginning of the project. By ensuring data accuracy and integrity from the outset, delays caused by erroneous or incomplete data can be minimised. High-quality data also leads to more reliable and trustworthy project outcomes.
Comprehensive data collection
Thoroughly identify and collect all relevant data sources needed for the project at the outset. This comprehensive approach ensures that the models developed are based on a wide range of data inputs, resulting in more robust and accurate analyses and insights.
Data quality assurance
Implementing rigorous data cleaning, pre-processing, and integration procedures is essential for maintaining data quality and consistency throughout the project lifecycle. By adhering to these practices, potential errors or inconsistencies in the data can be identified and addressed early, ensuring the reliability of project outputs.
Limitations of standard business tools
During the development of the dashboard a limitation of the number of records that could be viewed was identified using standard business tool PowerBI, PowerBI limits the number of lines that can be viewed at once to 150,000. It was identified that this limitation wouldn’t impact accessing data for customer engagement and support, as the data is available in the backend system should data sets with more than 150,000 lines be required.