For years, energy retailers have faced an uncomfortable technology trade-off. Move quickly with standardised systems and processes or invest the time required to configure sophisticated technology around the pricing logic, products and commercial models that make the business different.

There are good reasons for that compromise - B2B energy is complicated after all. Pricing logic can reflect everything from customer consumption and market exposure to risk, product structure and individual commercial strategy. Configurability matters precisely because no two retailers operate in quite the same way. But it comes at a cost: time.

Our Energy Margin Intelligence Index found that 57 per cent of energy retailers take 13–18 months to move a new product from inception to billing at scale. In a market where customer expectations, technology and competitive pressures are moving considerably faster, that becomes more than an implementation problem, it presents a constraint for innovation. AI creates an opportunity to rethink that trade-off.

The interesting application of AI in energy retail isn't simply generating answers faster. It is using AI to remove some of the work surrounding complex commercial logic without removing the logic itself. The opportunity lies in getting more value from the technology and commercial foundations already available, rather than recreating them.

That distinction has shaped the next evolution of Gorilla. We're introducing AI capabilities across our Energy Margin Intelligence platform, using AI alongside the structured commercial data and calculation engine already underpinning retailers' pricing and margin decisions. The intention isn’t to make every interaction AI-led but to apply it where it can meaningfully make complex work faster, easier or more useful.  In practice, that means teams can interrogate why margin has moved on a contract, understand the logic behind a price, configure pricing workflows faster and identify potential calculation or data errors, all without replacing the underlying rules that govern the answer.

The principle matters as much as the functionality. In commercially sensitive decisions, an answer that arrives quickly but cannot be explained has limited value. The calculation, therefore, remains governed and auditable, while retailers retain control over their own commercial logic.

Speed without confidence offers little advantage. If every AI-generated answer has to be manually reconstructed before someone will act on it, much of the promised productivity gain disappears.

Equally, the value of AI depends on the foundations beneath it. A sophisticated AI interface cannot substitute for the calculation engines, connected data and industry-specific logic required to support complex energy decisions. Those capabilities need to be dependable before AI can make them more accessible and efficient.

The bigger opportunity isn't to use AI to standardise energy retail. It's almost the opposite: to make sophisticated, configurable technology easier and faster to use.

For energy retailers, that could change the equation considerably. Instead of choosing between moving quickly and preserving the commercial logic that differentiates them, AI increasingly offers the possibility of doing both. It allows retailers to concentrate their expertise on the pricing strategies, products and commercial decisions that distinguish their business, supported by tried-and-tested technology designed to handle the underlying complexity.

And that may prove far more valuable than simply doing what we already do, a little faster.