E.ON Next puts customer trust at centre of AI strategy

E.ON Next is using AI to give customers instant energy quotes from uploaded bills, while its digital AI enablement lead warns that confidence in the technology will determine how far businesses can automate customer transactions.

Tim Lawless set out the approach ahead of an appearance at CustomerX in London in November 2026. His comments, reported by InternetRetailing on 9 October, distinguish between improving services customers already use and preparing for AI platforms to become sales channels themselves.

What happened

Customers seeking an E.ON Next quote can upload a bill from another supplier rather than manually entering all the information required. Lawless said the earlier process created opportunities for prospective customers to abandon their application before receiving a price.

The bill-upload service addresses a specific obstacle in an existing sales journey. Lawless’s test for such investments is whether AI enables a useful improvement, rather than whether a business can add a new technology feature to its website.

Alongside those changes, he sees platforms such as OpenAI’s ChatGPT becoming a route through which businesses reach prospective buyers. Consumers are already using generative AI to gather information without working through multiple pages of search results, he said.

Moving from research to execution would be a more substantial change. An AI service could identify a cheaper energy supplier and then offer to arrange the switch, requiring the customer to trust it with more than a recommendation. Lawless presented that as a possibility businesses should prepare for, not an established behaviour across the market.

The commercial challenge is therefore broader than adopting a tool. Businesses need to consider how their products are found, how customers assess them and which parts of a purchase customers are willing to delegate. Retailers face a related question over visibility, with H&M’s SEO lead urging preparation for AI shoppers.

Lawless argued that developing these new routes should not come at the expense of existing ones. Companies will need to accommodate people who delegate tasks to agents, those who use AI only for research, and customers who continue to prefer established digital services or contact with staff.

The background

Lawless illustrated the gap between technical availability and public confidence through his mother’s experience of online shopping. He said she took 15 years to become comfortable enough with the internet to enter her credit card details online.

His argument is not that customers will reject AI outright. Instead, willingness to use it will vary by demographic group, sector and task, making a single adoption assumption unreliable for businesses planning customer services.

A person might accept AI-generated restaurant suggestions or holiday plans while refusing to let the same technology access a bank account or change an energy supplier. For Lawless, the important distinction is the responsibility being handed over, rather than whether someone is broadly enthusiastic about AI.

That leaves room for services which assist customers without taking full control. A business could support AI-led research while allowing the customer to complete the transaction through a familiar channel, rather than requiring an immediate move to autonomous purchasing.

The handling of personal information also sits alongside this debate about delegation. The Information Commissioner’s Office has questioned OpenAI, Meta and Anthropic over AI agents, bringing regulatory scrutiny to a technology businesses are assessing as a potential customer interface.

Inside organisations, Lawless said AI also differs from more familiar technology programmes. A customer relationship management or enterprise resource planning implementation usually has a broadly understood destination; he believes businesses remain much less certain about the eventual organisational benefits of AI and where to begin.

What people are saying

Lawless separates the work into two areas: improving products and customer interactions, and changing how employees carry out their jobs. Although connected, he said these require different approaches rather than a single transformation programme applied across the organisation.

Customer-facing improvements can offer employees a readily understandable purpose. Changes to working practices are more sensitive, particularly when the discussion begins with job replacement, automation or reducing the volume of tasks people perform.

He favours giving teams practical opportunities to test AI against frustrating or repetitive work. Presenting the technology as support for those tasks can help employees see where it is useful, while leaving them more time for work that contributes to business performance.

Participation matters because staff who have not experimented with the tools may remain anxious or resistant, he said. His approach is to let people build, test and assess applications themselves, rather than relying on instructions to adopt them. That emphasis on a defined operational problem is also reflected in Emma Jones’s call for small firms to trial AI for late payments.

Lawless also wants investment decisions to start with a business objective rather than a predetermined piece of software. At E.ON Next, he is trying to shift the emphasis from selecting individual AI applications to funding improvements in sales, customer effort or retention.

Under that approach, a product team would receive responsibility for an outcome and test whether AI helps achieve it. It could then adjust its work as technology develops or customers respond, instead of treating delivery of a chatbot as the measure of success.

He identified IKEA as an example of a business connecting customer-facing AI with workforce changes. In his assessment, its approach includes moving employees into work that creates additional value, rather than focusing only on reducing tasks. He did not suggest it offered a complete model for others to copy.

What happens next

Lawless is due to discuss AI in commercial teams at CustomerX next month. He sees the event as an opportunity to compare E.ON Next’s progress with businesses outside energy and identify both strengths and gaps in its approach.

He argues that energy suppliers, retailers and financial services businesses can learn from one another’s experience of combining AI, data and customer service. That exchange is intended to help organisations assess approaches beyond their own sector, rather than assume one company has already resolved every challenge.

For implementation, his proposed starting point is to establish separate customer and workforce priorities, deliver useful applications and learn from their performance. Experiments should remain tied to wider company objectives, so that testing leads to operational decisions rather than an indefinite collection of pilots.