"Why my Reel looked good but didn't sell"
Léa R., a beauty creator, was preparing a highly polished sponsored Reel. Before sending it to the brand, she wanted to validate one thing: would the product and its benefit be understood early enough?
A realistic scenario built to show the level of detail of the final report.
Visible glow. No filter.
Product identified at 7.1s
Context
Léa has been publishing skincare Reels for two years. Her audience follows her for honest routines, not disguised ads. When a serum brand approached her for a sponsored collaboration, she wanted a video that stayed true to her editorial tone.
She filmed a 15-second GRWM: soft light, mirror, serum application at the end of the routine. The team liked the edit, but no one knew whether the promise landed early enough for a cold audience. Media spend had not started yet.
The original creative
"The serum everyone's talking about"
A fixed shot facing a mirror. The creator talks for 15 seconds. The serum visually appears at the 7-second mark. The product promise is spoken at second 9. The CTA is at the end of the video.
Protocol
Each reaction is generated in the target audience context, then aggregated without hiding minority objections.
What the panel detected
The visual universe is premium and the hook grabs attention. But the product and its promise arrive after the interest drop. The panel doesn't understand what's being sold until second 9 — too late.
500 synthetic agents matching the target audience profile watched the video. Here are their aggregated reactions.
Primary score
54/100
Critical signal
Attention drops 22% between seconds 4 and 7 — before the product even appears.
Panel reactions
“It's pretty, but I almost scrolled before seeing the product.”
“I didn't know it was a serum ad until the end.”
“The skincare routine is pleasant to watch but the product should have been shown immediately.”
The decision
Show the serum and its promise from the very first second.
The panel identified that attention is captured immediately by the visual hook, but drops as soon as the video becomes a routine without an identified object. The product must be visible in the frame from the opening, and the promise must land within the first 3 seconds.
The re-test
The revised version keeps the tone and lighting of the GRWM. The serum is visible from the opening frame and the benefit lands in the second second.
Before
54
/100
After fix
82
/100
Simulated improvement
+28 pts
Changes tested
- Serum visible in frame from second 1
- Product promise moved to second 2
- Routine kept but placed after product identification
- CTA kept at the end with a visual callback to the serum
“This time I immediately understood what was being sold. The routine becomes a demonstration, not a distraction.”
Next case
"The client had an opinion. The panel had an answer."
Find the correction worth testing.
Upload a creative, describe the audience, and get one priority decision before committing budget.
Test a creative for free