Case study · Supplements · Meta
$263,235 in Meta-attributed revenue in 82 days: a historical AI creative case.
Tubentum creative appeared in 149 ad IDs for a DTC supplement brand from June 28 to September 17, 2026. These include repeated use of videos, rather than 149 unique videos produced. Meta attributed $263,235 in revenue and 3,858 purchases to those ads on a Meta attribution window. The client managed media buying during this engagement. Before it began, static images had been the account’s established production format. The work developed from a small paid production project into recurring creative batches.
By Soyeon Yoon · Published October 12, 2026 · Client name withheld
The numbers
REVENUE AND PURCHASES ARE WHAT META ATTRIBUTED TO ADS USING TUBENTUM CREATIVE ONLY. SPEND, ROAS AND COMPARISONS AGAINST THE CLIENT’S OWN ADS ARE NOT PUBLISHED. RESULTS ARE NOT GUARANTEED.
The situation
A DTC supplement brand wanted more video directions for its Meta account. Static images were an established format. The assignment was to develop different creative concepts and presentations of the product.
What I made
I developed concepts, scripts and hook variations in recurring production batches. Product references, customer problems and account feedback informed the work. The client controlled media buying in this historical engagement.
The production explored AI presenters, interview-style cuts and B-roll with voiceover. These were different creative approaches to test, rather than a guarantee that one format would win for another brand.
What the results showed
Review the full performance history. The initial test looked stronger than its later cumulative result. A useful review considers the full reporting window and test conditions before deciding what comes next.
Results concentrated in one historical batch. In that batch, two videos produced almost all of the revenue and the other eight barely spent. That is an observation from this batch. It guides the next creative decision, rather than a promised hit rate for another account.
Not every batch beat the last. Some delivered material received too little spend to support a useful verdict. Low delivery does not by itself establish that an idea failed.
“Separate a weak result from an under-tested idea.”
Soyeon Yoon, founder, Tubentum
Plan the next test for your account.
Bring your product and recent account results. A planning call establishes the fit; the $3,000 creative test scope, launch plan and deliverables are confirmed before starting.