At EMC 26 in Houston, a panel of retail energy CIOs and technology leaders sat down for what was billed as a conversation about the role of IT in shaping competitive advantage. What unfolded was something more interesting: a fairly clear consensus on what comes after data integration as the next frontier for energy retailers, and a useful disagreement about where the disruption to today's playbook is going to land first.

Gorilla's own Joris Van Genechten, co-founder and VP of Product & Engineering, joined a cast of prominent tech leaders from across the energy industry. On stage with him were Sajjad Mussani (CIO of 174 Power Global and Chariot Energy), Tyler Falkenhagen (CIO of Japan Electric Renewable Americas), Brian Hines (formerly Senior Director of Retail Technology at Direct Energy), and Matt Goldman (VP of Retail Energy at CG Infinity).

Below are the threads that ran through the discussion, and what they suggest about where retail energy IT is heading next.

The CIO as a commercial leader

How hasthe role of IT has changed over the last ten to twenty years? Brian Hines offered the cleanest data point. At Direct Energy in 2015, the call centre was the number one sales channel and the mobile app accounted for around 17% of customer interactions. A decade later, partners are the leading channel, organic is second, and mobile app usage sits at roughly 85%, with most customers using the app at least once a month. The benchmark for customer experience, he noted, is no longer the retailer down the road. It is Amazon, and the last great experience the customer had anywhere.

The CIO is not in a back room writing code. The CIO is in the boardroom, responding to unending questions about AI. When Sajjad Mussani was asked this, he asked the board member back what they meant by AI, and the board member could not answer. That gap, he argued, is the modern CIO's job. The skills you need in a senior team have shifted accordingly: commercial thinking first (looking at churn, conversion, and business outcomes rather than IT delivery metrics), with a strong instinct for guardrails and controls in the AI era.

Competitive advantages

Two things emerged when discussion turned to competitive advantages in the next five years.

Tyler Falkenhagen made the case for speed on data. The companies that can pull their data together quickly are the ones that can act quickly, and that effect is amplified by every layer of AI you put on top. Joris took the same logic one level deeper. Beyond data, the bet is on companies that can embed their institutional knowledge into the systems themselves. The data needs to encode how the business actually runs: how margins are made, what sets the company apart, what the niche is. If that knowledge sits inside the systems, the analysis rolls out automatically rather than being repeated by people every time a question comes up.

Most retail energy IT roadmaps over the last five years have been about getting the data trustworthy and integrated. Plenty of retailers are still working on that and Gorilla is there alongside to support them. Now the emergence of AI is demanding that access to all of this data is unfettered, without bottlenecks caused by manual interventions or the need to get input from your most experienced people every time. At the same time, every company is aware of the horror stories of AI deleting databases or exposing secrets, so retailers will need to tread a fine line to ensure systems deliver on both speed and security.

AI without domain expertise is confidently wrong

AI ran through the conversation as both a promise and a warning. The warning came up most pointedly when the panel discussed where AI actually creates measurable business value in retail energy.

Joris's framed it thusly: Use AI in retail energy without deep energy expertise, and you will get a confidently wrong decision. The reason is that retail energy decisions sit on top of decades of accumulated context: regulatory nuance, market structure, contract logic, customer behaviour. Generic AI does not have that context, and pretending it does produces output that looks credible but is structurally wrong.

The version that does work is augmentation. Senior analysts and portfolio managers traditionally look at the top ten or twenty contracts in detail because that is all human bandwidth allows. With AI tooling that holds the same context as the analyst, the same analysis can run across the entire book. The senior analyst's role shifts from execution to validation: confirming the analysis is right, identifying where to tweak it, scaling impact further than they could on their own.

Tyler Falkenhagen added the accountability point with a generation-side example. A pump goes out on turbine number one; an operator can ask AI how to replace it and receive the procedure, parts, and steps. That works. But for critical infrastructure activities, the question of who is accountable for the AI's recommendation is unresolved.

Outcomes, not scope

A consistent theme on the vendor relationship: the value comes from sharing what you are actually trying to achieve, not just what you want built.

Joris described how Gorilla's project conversations have moved away from scope-led planning. The starting question now is what the outcome needs to look like. Scope is something the Delivery team works out from there. Sajjad reinforced the point from the customer side: he shares his numbers and his metrics with both internal teams and external vendors, and asks them all to come back with an ROI that matters to the business.

Most retail energy implementations are still scoped on functional requirements, with outcomes inferred. Reversing that order, in the panel's experience, both shortens the project and produces a system that is genuinely tied to the metric the business is being measured on.

The next disruption

Where the panel actually disagreed, productively, was on where the next disruption is going to come from.

Sajjad's bet was on speed of reaction. Volatility, regulation, customer expectations, and new sales channels are all adding complexity. Whichever retailer can absorb that complexity and react fastest is ahead. Matt Goldman's bet was on adaptability. Five years ago nobody was talking about generative AI; today everyone is; in five years the cost curve of AI may force another shift. Companies that build their architecture on a single technology assumption will get caught out. Brian Hines bet that the digital channel itself is going to change shape. Google has publicly said it expects search volume to drop materially as users move from browsing to asking AI. If your sales engine is built on Google Ads and online partner traffic, that is going to be affected.

None of these bets is obviously wrong. Together they suggest the same underlying point. The retailers that win the next decade will be the ones whose architecture, operating model, and decision processes can absorb a market that keeps surprising them.

What this means for retail energy

The panel did not agree on where the disruption arrives first, but they did agree on what makes a retailer ready for it. Trustworthy, integrated data as the floor. Institutional knowledge embedded in the systems doing the work, so the analysis scales beyond the bandwidth of the people who hold the expertise. AI used to multiply senior experts rather than replace them. And a willingness to operate on outcomes, not scope, both internally and with vendors.

For retail energy specifically, this maps directly onto the margin question. Margin in this industry is made and lost in the gap between pricing, forecasting, hedging, and billing. Closing that gap takes more than data. It takes the institutional knowledge of how margin actually behaves in your business, embedded in the systems doing the work, with the senior people who hold that knowledge freed up to validate and steer rather than execute.