Developing
The agency's CEO describes his own firm's framework and system; results are anecdotal.
In 60 seconds
- Power Digital announced a proprietary AI ecosystem in August 2026: an intelligence layer called Iris, an operating system called Omega and a creative layer called Creative Affinity.
- Liquid Death now optimises campaigns daily, especially on retail media, tasks once reviewed weekly or twice weekly; it has produced 17 iterations of Omega in three weeks.
- Power Digital's CEO describes three phases: AI-curious, AI-assisted (a bespoke intelligence layer) and AI-native (adds an orchestration layer).
- Vatere's stated goal is to run a sizable budget with two people internally.
Impact Layer
- What changed
- A brand and its agency describe a working model for AI-run media operations.
- Why it matters
- It is a concrete picture of what organisations mean by AI-native, including who owns data and decisions.
- Who is affected
- Marketing operations, media and agency leaders.
- Martech stack area
- Analytics, AI
- Stack impact
- Ask where your team sits on the three phases and which data and orchestration layer would move you.
- Business impact
- Low Reported by Marketing Dive from a panel. The phases are Power Digital's framework, and the agency has an interest in selling the approach.
- Related technologies
- What to watch
- Measured outcomes from the approach.
Marketing Dive reported on October 7, 2026 on a panel at Advertising Week New York (October 5) with Benoit Vatere, chief media and digital commerce officer at Liquid Death, and Power Digital Marketing, the agency CEO Jeff Mason leads.
The system
Power Digital announced a proprietary AI ecosystem in August 2026: an intelligence layer called Iris, an operating system called Omega and a creative intelligence layer called Creative Affinity. Vatere says it lets him see campaign activity without Slack or writing queries. Liquid Death now optimises campaigns daily, especially on retail media, where it once reviewed weekly or twice weekly, and has produced 17 iterations of Omega in the last three weeks. His stated goal is to run a sizable budget with only two people internally.
Three phases (per Mason)
- AI-curious: uses frontier models such as Claude and ChatGPT with some structured and unstructured data added. Mason says most brands are here.
- AI-assisted: builds a bespoke intelligence layer that all employees can access and that learns how teams work.
- AI-native: adds an orchestration layer on top, enabling work inside custom applications.
Advice
Centralise and organise data (Mason calls it the long-term moat); check AI agents for accuracy; use and improve AI systems daily; stop relying on Excel; and automate any task that takes more than five minutes. Mason says frontier models no longer offer a competitive edge because every firm uses them.
Our read
The article offers no outcome metrics, and the agency sells the model. The transferable idea is putting your own data and rules between the models and the people.
Sources
- Marketing Dive: How Liquid Death is building an AI-native marketing organization (opens in a new tab)
Source of the panel details, the phases and the advice.