Go back to Blog
Aurelie Touboul
Published on 4 June 2026
For the launch of Flam on TF1+, Lunii applied the same Creative Intelligence logic to Premium CTV as it does on YouTube — and proved that broadcaster inventory can also be performance-driven.
CTV buying has gone fully programmatic.
Media optimization layers are now extremely sophisticated: targeting, bidding, reach management, frequency capping, cross-platform orchestration and measurement.
But creative optimization remains largely intuitive and static.
The media infrastructure is optimized.
The creative layer is not.
Paradoxically, studies show that creative accounts for 50% of a video campaign’s intrinsic performance.
Yet most Premium CTV campaigns still run with a single Master video, a primarily branding logic, very few creative variants, and virtually no cross-platform creative learning loop.
Lunii is a French audio entertainment brand for children, known for the Fabrique à Histoires and now expanding into the teen market with Flam.
The brand wanted to invest in a Premium CTV environment without entirely sacrificing its short-term performance KPIs: qualified site visits and demand generation.
On premium inventory, all creatives reach a 99% view-through rate. That metric validates the format – not the creative.
It says nothing about what drives a viewer to visit the site, what the intro triggers in terms of purchase intent, or which variant actually deserves the budget.
Already running creative performance on its video campaigns with VideoRunRun on YouTube CTV, Lunii naturally wanted to apply the same logic to TF1+ premium CTV inventory.
VideoRunRun deployed its Campaign-Level Creative Intelligence Layer on the TF1+ Flam campaign, as an increment to its YouTube CTV campaign.
The CI Layer was activated across three phases Pre-Flight, In-Flight and Post-Flight to
1/ Pre-Flight: From a Single Master to a Creative System
The strategy was to move from a single branding video to a system of iterations and analysis.
VideoRunRun applied its Predictive Scoring approach to generate creative optimization recommendations before the campaign launched.
Three variants were deployed alongside the Master: