How could AI assistants transform energy retail?

Since the launch of Chat GPT-3 in late 2022, business executives can’t stop talking about AI. It’s set to fundamentally change the way in which many companies operate, how they engage with their customers and the skills they will require from their workforce.

For energy retailers, AI assistants or “digital agents” could transform the way in which retailers interact with their customers, which will could have major impacts in wholesale markets.

You can think of a digital agent as a virtual personal assistant acting on your behalf in the background. They do all the “admin” tasks that may be time-consuming or we can’t be bothered to do regularly, like constantly scanning for the best-priced deals on insurance, phone and electricity bills. They will always act in the consumer's best interest.

This concept may feel far off, but “Conversational AI”, the key to making this technology work, is a frontier has now been crossed.

What does this mean for energy retail?

At Australia’s first Generative AI Summit, Telstra Energy’s Ben Burge, who is also a former Powershop CEO, put it simply, “consumer agents kill the loyalty tax”.

How does this work?

According to the Australian Energy Regulator’s (AER), Annual Retail Markets report (2020-21) consumer switching rates have varied between 3% to 9% on an annual basis. This shows a reasonably low engagement between consumers and the energy market, perhaps because the savings of switching are too small, the task too difficult or simply not important enough.

For a digital agent, no task is too boring, no saving too small. With digital agents operating on behalf of consumers, constantly looking for the best deal, it’s conceivable that switching rates will skyrocket to a point where consumers are switching continually, looking for the best deal.

For energy retailers, this could also mean that marketing efforts would lose their impact almost completely as AI bots constantly scour the market for the lowest price, impervious to emotional pitches and unswayed by flashy advertising

Energy retail would quickly become a competition of who can best manage wholesale market price volatility and who can get the lowest wholesale price (and best margin).

To quote Ben, “How can you win them back if they can’t hear you?”.

A golden opportunity for regulators?

Recently energy market participants have seen an increased attention from market regulators, including a recent $17m fine for Origin Energy from the AER.

Ben pointed out thatenforcement is challenging and “the cost of enforcing these rules all falls back to the consumer eventually, through the way of cost.”

This cost could significantly fall with digital agents, with each agent keeping a record of all interactions between the consumer and retailer and able to report directly to regulators and ombudsmen to ensure the enforcement of rules. Ben described it as an “amazing opportunity for regulators and enforcement.”

Is AI the catalyst for demand response?

Consumer level demand response has long been touted as a way to manage load, but DR programs are incredibly difficult to implement. Price signals alone have proven ineffective in getting households to substantially change usage behaviour. However, digital agentsand could realistically have meaningful impact during short, temporary demand/supply shortfalls that become more common in a predominantly renewable grid.

Digital agents will build intimate relationships with their users over time, building up the buy-in to manage large parts of their lives. This could include things like switching an air-conditioner or washing machine on or off in response to price signals. Something that utilities have long hoped for.

Ultimately these agents would always work to benefit their user, but with the right price signals, they can also benefit the grid.