How We Scaled Airmorn to 3.5x ROAS in 3 Weeks

Airmorn, a crowdfunding steam iron brand, approached our media buying agency during their early-stage launch phase. They were burning budget on undefined audiences, unable to scale daily spend, and stuck at 1.5x ROAS. Within three weeks, we delivered 3.5x ROAS and increased monthly ad spend by 300%. Here's the methodology.

The Challenge

Early-stage crowdfunding campaigns face a unique media buying challenge: no historical conversion data, undefined target demographics, and limited budget tolerance for learning. Airmorn's account exhibited three critical failures:

• Unclear audience targeting: Campaigns were targeting broad, undifferentiated segments, resulting in 70%+ wasted spend on non-converting traffic
• Budget inefficiency: Daily spend caps were artificially low ($200-500/day) due to poor performance, preventing scale
• ROAS stagnation: Account-level ROAS hovered at 1.5x, below the 2.5x minimum viable threshold for sustainable growth
• Creative fatigue: Initial launch creatives showed declining CTR and conversion rates after 2 weeks

The brand was operating in a blind optimization loop: low ROAS prevented budget increases, but without higher budgets, the algorithm couldn't exit the learning phase. This media buying agency intervention required a systematic reset.

The Solution

We implemented a three-phase recovery framework: deep data diagnosis, audience reconstruction, and agile creative iteration. Each phase was executed sequentially to eliminate guesswork and maximize budget efficiency.

1. Data Insight & Diagnosis

We conducted a comprehensive audit of 14 days of historical campaign data across Meta Ads and Google Ads. The analysis revealed three root causes:

• Demographic leakage: 45% of spend was allocated to age groups (18-24) with <0.5% conversion rates, despite representing only 12% of actual purchasers
• Interest targeting mismatch: Campaigns targeted "home appliances" and "kitchen gadgets" (high competition, low intent), missing the actual buyer profile: "professional women 30-45, urban, high disposable income"
• Placement waste: 38% of budget was allocated to Facebook Feed placements with 0.2% CTR, while Instagram Stories (unused) showed 2.1% CTR in competitor benchmarks

We eliminated all assumptions. Instead of testing new audiences, we reverse-engineered the buyer profile from actual conversion data, then rebuilt targeting from first principles.

2. Audience Reconstruction

We deconstructed the existing audience strategy and rebuilt it using a data-driven segmentation model. The new targeting architecture eliminated invalid traffic and penetrated the core consumer cohort.

• Primary audience (60% budget): Custom audience built from website visitors (last 30 days) + lookalike 1% (seed: purchasers). Age: 30-50, Gender: 60% female, Interests: "professional lifestyle", "time-saving products", "premium home goods"
• Secondary audience (30% budget): Broad targeting with detailed exclusions. Removed: age 18-29, interests "budget shopping", "discount deals". Added: income brackets (top 25%), job titles (executive, manager, entrepreneur)
• Retargeting (10% budget): Website visitors (last 7 days) + video viewers (75% completion) + cart abandoners. Excluded: purchasers (last 90 days)
• Placement optimization: Removed Facebook Feed. Prioritized: Instagram Feed (40%), Instagram Stories (35%), Instagram Reels (25%)

This media buying agency approach shifted spend from low-intent segments to high-probability converters. Within 5 days, CPA dropped 42% and ROAS improved to 2.1x.

3. Agile Creative Iteration

We implemented a rapid creative refresh cycle: monitor performance daily, identify fatigue signals (CTR drop >20%, CPA increase >15%), and deploy new variants within 48 hours. This maximized every dollar of budget.

• Creative audit: Analyzed 8 existing video ads. Identified top performers: lifestyle shots (steam iron in use, professional setting) outperformed product close-ups by 2.3x ROAS
• New creative production: Commissioned 6 new 15-second videos: 3 lifestyle variants, 2 testimonial formats, 1 before/after comparison. All optimized for Instagram Stories and Reels (vertical 9:16)
• A/B testing framework: Launched new creative set with $1,500/day budget. Tested 3 variants per creative: hook variation (problem vs. solution vs. benefit), CTA placement (early vs. mid vs. end), product focus (feature vs. outcome)
• Iteration cadence: Weekly creative refresh. Paused underperformers (ROAS <2.0x) after 3 days. Scaled winners (ROAS >3.0x) by 50% budget increment

This agile media buying methodology prevented budget waste on fatigued assets. New creative variants maintained CTR above 1.8% and conversion rates above 2.5%, enabling sustained scaling.

The Results

Performance transformation occurred within 21 days. The account stabilized, scaled, and exceeded pre-intervention benchmarks.

Key Takeaways

• Data diagnosis eliminates guesswork: Deep analysis of historical performance identified three root causes (demographic leakage, interest mismatch, placement waste) that were invisible to surface-level metrics
• Audience reconstruction is non-negotiable for early-stage brands: Rebuilding targeting from first principles (using actual conversion data) reduced CPA by 57% and enabled 3x budget scaling
• Agile creative iteration maximizes budget efficiency: Weekly creative refresh cycles prevented fatigue-driven performance decay, maintaining CTR above 1.8% and enabling sustained scaling
• Rapid scaling is possible with proper foundation: Once targeting and creative were optimized, we scaled monthly spend from $8k to $32k in 3 weeks without ROAS degradation

This case study demonstrates that early-stage crowdfunding campaigns can achieve profitable scale within weeks, not months. The methodology: eliminate assumptions through data diagnosis, rebuild targeting from first principles, and maintain creative freshness through agile iteration. This media buying agency approach transforms underperforming accounts into scalable growth engines.