How Grüns Achieved 247% ROAS Increase with Vidsight Creative Intelligence

The Client

Grüns is a DTC supplement brand founded in 2023 that delivers daily nutrition through gummy vitamins containing 60+ organic ingredients. With 250K+ customers, $1.8M monthly Amazon revenue, and distribution in 400+ Sprouts locations, they've positioned themselves as the convenient alternative to AG1 for people who hate green powders. Their primary marketing channels are TikTok, Instagram Reels, and YouTube Shorts.

The Challenge

Despite viral TikTok success, Grüns faced three problems:

1. Creative fatigue — Same formats recycled without performance data. Average RPI had dropped to 0.67x (34% below baseline).

2. Paid media plateau — ROAS stuck at 1.8x for 4+ months. No framework connecting organic virality to conversion potential.

3. Virality ≠ conversions — High-view videos weren't converting. Average purchase intent was just 8.2%.

The Vidsight Approach

Vidsight analyzed 1,847 videos across competitors (AG1, Bloom, Olly, SmartyPants), wellness creators, and Grüns' existing content. The AI scanned every video frame-by-frame for hooks, CTAs, format types, tone, pacing, and audience signals—generating 23,500+ engagement datapoints scored for Relative Performance Index (RPI) and purchase intent.

Key Findings

Finding 1: Optimal Visual Formats

• Visceral product interaction (close-up gummy shots, ASMR audio): +127% RPI lift

• First-person POV morning routines (hands-only, no face): +89% RPI lift

• Before/after energy reveals (split-screen transformations): +64% RPI lift

Finding 2: Purchase Intent Analysis

• Product demonstration videos: 24.7% purchase intent

• Routine integration content: 21.3% purchase intent

• Trending/hype-only content: 4.1% purchase intent

The insight: formats that activate sensory simulation and habit visualization drive both engagement and conversions.

Finding 3: Viral-to-Ad Conversion Vidsight identified 12 organic videos with RPI 1.5x+ AND purchase intent 25%+. When converted to paid ads, these averaged 6.2x ROAS—compared to 1.8x for creatives selected by intuition.

Finding 4: Pre-Upload Scanning 47 videos were scanned before publication. Results: 89% prediction accuracy, 32 videos improved pre-post, +71% average RPI lift, and 8 underperformers caught before wasting ad spend.

Results (90 Days)

Key Takeaways

1. Virality and conversions require different creative signals—optimize for both

2. Format selection matters more than posting volume

3. Pre-upload scanning prevents wasted spend on underperformers

4. A systematic organic-to-paid pipeline outperforms intuition-based creative selection