AI-supported energy advice: next-generation customer experience

For years, the energy industry has been relying on artificial intelligence (AI) for smart energy distribution, grid load management, and protection against cyberattacks. But the new generation of AIs, exemplified by ChatGPT, Midjourney and others, is ushering in a completely new era. It takes the potential of the technology to a whole new level – especially for customer interaction – and thus also becomes the decisive lever for energy suppliers to remain competitive in the future.

According to a digitization study by the consulting firm PwC, an outstanding customer experience is a decisive factor in the purchase decision for 73% of those surveyed. Those who offer a positive customer experience can even charge up to 16% higher prices. Above all, speed counts: expectations for better, more personalized advice, faster response times to service requests, and better services around the core product are constantly increasing. The possibility of using intelligent AI assistants in precisely this area will quickly raise the bar for companies. The potential for improving the customer experience with the help of generative AI in a variety of ways is enormous and goes beyond mere chatbot applications. The use of consumption data as a basis for individual customer advice is just one further area.

Many companies are also recognizing this trend: According to a Gartner survey, almost five months after the launch of ChatGPT, almost half of the executives stated that they had increased their investments in AI because of it. The customer experience and loyalty were clearly the focus of the AI investments. Companies that are not yet exploring the possibilities of the technology will face real competitive disadvantages in the medium term. It is also high time for the energy industry!

The starting point: internal knowledge management

Currently, generative AI models are used by companies primarily as efficiency drivers. According to a study by the Boston Consulting Group, employees who use AI tools achieve up to 40% better results. They can complete up to 12% more tasks – for example, by automating and optimizing online ads or coding for websites and platforms in marketing. The E.ON Group has been working with its own AI since December of last year. “E.ON GPT” supports employees of the energy provider, for example, in research activities, word processing and ‘as a virtual sparring partner for energy industry issues’. Although internal knowledge sharing and automation are definitely useful for a first approach to the technology, they fall well short of tapping the full potential. We will also see the first applications in direct interaction with customers in the near future.

AI as a translator for complex topics

The German Association of Energy and Water Industries has identified customer orientation as one of three key areas for action on the road to the digital and ecological transformation of the energy industry.

Generative AI is particularly important in this context, because it is creating a completely new generation of chatbots. Currently, chatbots provide strictly predetermined conversations in which the user selects from predefined answer options and at the end receives a link to a (possibly) suitable landing page. In the future, intelligent AI assistants will be able to answer customer queries the way we would: in a dialog.

Behind every application is a central language model. These Large Language Models (LLMs) form the basis for conducting complex conversations, delivering customized solutions, and referring to more in-depth sources. Energy companies that want to set up an intelligent assistant need a proprietary enterprise AI as a basis. The key here is not the model itself, but how it is trained with content-verified data. The new thing about this is that this data can also be available in “unstructured form”, for example as website text, customer brochure, manual or the course of a conversation from customer service. Based on this, the AI provides in-depth knowledge and comprehensible explanations to specific questions. This is the first time we have had a technology at our disposal that allows us to rethink digital customer interaction and make it human-centered, natural and highly personalized. Imagine having 1:1 support for every customer, millions of personal conversations, without long Google searches, endless click navigation or waiting times. An AI assistant that patiently answers every question, always in the right tone, that quickly and easily helps customers and empowers them to make informed decisions based on sound information.

Because of these properties, generative AI is currently particularly suitable for use cases in which the aim is to play synthesized information. The technology can also already provide good support in the transactional area – for example, when filling out forms or grant applications. However, the aspects of autonomy and security are essential for a positive user experience. What does that mean in practice? The AI is welcome to fill out the form. But for checking and sending, users still prefer to do it themselves. AI assistants reach their limits when it comes to critical and emotional topics that address a high need for security. In situations where compassion and life experience are required, a human counterpart will always be needed.

Next-generation energy advice

So what will the intelligent energy consultant of the future look like? Enriched with website information and anonymized data from customer service – such as the most frequently asked questions and their corresponding answers – energy companies can use AI-based chatbots and intelligent platforms directly. This results in an actionable AI solution for targeted and intelligent advice. By providing all relevant information in an immediate and comprehensible way, as well as through dialogic exchange, customers are guided towards the right product and given the necessary transparency for sustainable purchasing decisions.

This applies not only to the right electricity contract. Politicians want to see a rapid turnaround, but many consumers are unclear about how solar panels or heat pumps work, whether or which technology is suitable for their own living situation, and what long-term advantages they offer. This is where AI can help by explaining complex relationships in a simple and user-centered way. Particularly when it comes to purchasing photovoltaic systems or heat pumps, as well as the associated subsidies, energy suppliers are currently often losing customers to specialized providers. Intelligent energy advice can provide the decisive leverage here to catch up.