UK Power Networks is working to deliver support at scale to vulnerable consumers at risk of being left behind by the energy system transition. This project will develop a new digital tool that aims to utilise cutting edge technologies such as machine learning and artificial intelligence to enable tailored support to be effectively delivered to vulnerable consumers at scale. By working closely with current service delivery partners, the aim is to build a tool that supports existing services while also providing new options for consumers that would prefer to self-serve using digital channels.
Benefits
During RIIO-ED2, UK Power Networks is working to deliver support at scale to vulnerable consumers at risk of being left behind in the energy transition (aka Leaving No one Behind, LNB). This requires a step change in volumes of vulnerable and disadvantaged customers that are identified and supported. This is especially important in a context where other factors, such as the increasing cost of living, digital exclusion, increased inflation and increasing energy costs will contribute to increasing numbers of people at risk of being left behind.
A key challenge currently faced is an inability to scale services to meet the required significant increase in volumes compared to RIIO-ED1. Many of the partners funded by Distribution Network Operators (DNOs) to provide support to vulnerable consumers do so manually and are operating at capacity so unable to address the volumes of customers that require support. Many partners have growth plans to expand their reach, but this requires additional time and resources.
It is relevant to note that across the existing portfolio of services and initiatives that UK Power Networks provides, there has currently been a low uptake of digital products despite novel solutions being available. This comes despite an increased expression of interest from consumers in digital self-serve channels. This project looks to address this challenge in two ways:
Utilising innovative technologies to provide a tailored experience for customers. This will allow customers to interact with the tool to provide bespoke support, with the aim of providing a high level of customer service.
Working with frontline service delivery partners from the outset in design of the tool. Delivery partners could utilise the tool to enhance their existing service, support the growth of their service, or promote the tool to customers who can self-serve. This reduces the risk of under-utilisation of the tool relative to investment and will help increase the reach.
Learnings
Outcomes
Identified challenges with delivering support to vulnerable customers: Through a series of interviews and workshops, the service delivery partners shared their insights on the behaviours, needs and attitudes of vulnerable consumers. Challenges that the delivery partners face and how they provide effective support was also discussed and collated.
These activities aimed to identify where technology such as artificial intelligence could perform tasks for front line staff, while identifying the parts of the process where a human touch provides the best outcome. Identifying which step of the process is best performed by staff member or technology is important to optimising the outcomes of customer satisfaction with effective use of available resource.
Validated problem statement: The service delivery partners agreed that enabling some consumers to self-serve will free up resources for other consumers that need a hand-holding approach. For example, empowering consumers to find and apply trusted information themselves rather than being reliant on an interaction with a human can alleviate call wait times.
The developed proposals for the digital tool were then presented back to the service delivery partners and received positive feedback, validating that it would be valuable for consumers They confirmed that the proposed designs enable to tool to be used by customers to self-serve, or by service delivery partners to support clients during home visits, therefore maximising its value.
The work to date has confirmed that there are a number of technology solutions that could utilise artificial intelligence, machine learning and data analysis to deliver support to vulnerable consumers. We are learning from other tools available in the market as well as the direct learnings identified through this project to ensure that the proposed tool is fit for purpose and meets customer needs effectively.
The customer validation exercise combined with ongoing engagement with internal and external stakeholders produced insightful learnings that are impacting the development of the tool, and improving the final outcomes. Therefore, the tool concept proposed is being further developed and tested prior to being built.
Lessons Learnt
Customer validation is essential in digital projects: We tested the aims of the digital tool, as well as the design proposal, with customer working groups. Customers were initially distrusting of a free digital tool that offered advice, until the role of DNOs was explained to them, and then trust increased. Customers recognised the value in the tool, and the ability for customers to self serve at a time and pace that worked for them. However, customers were also concerned that not everyone may be able to use the tool effectively if they were lacking in digital skills, and shared examples where a digital service that had seemed useful actually was not able to deliver what they needed. This highlighted the importance of managing customer expectations, and ensuring the tool has sufficient information and useability.
Digital projects need to maintain flexibility in delivery: Due to the rapidly evolving nature of the current AI field, we have discovered that the development process should be agile to enable maximum benefits of the technology. Further exploration of the complexity and costs relating to the design, build and operation of the tool, especially the artificial intelligence elements, highlighted the challenges and opportunities linked with this technology.
Delivery of technical tenders: Assess the market size and estimate the volume of responses prior to issuing a invitation to tender, and also allow additional time and resource contingency for higher volumes than anticipated. Due to the current growth in the AI market and high interest from organisations looking to demonstrate their expertise in this field, we received a significantly higher volume of responses to the tender than anticipated. This required additional resource, including time from internal stakeholders and experts to assess the tender responses and provided feedback to all applicants. It also required us to manage expectations of the tenderers, as the higher volume of responses extended the planned timeline of the tender process.
