[Strategy] Testing Video Creatives in Sponsored Brands Campaigns

Video creative testing is a core lever for improving Sponsored Brands performance, but many sellers aren’t sure how to structure these experiments for reliable results. This thread surfaces a practical question: how do you isolate which video actually drives better ACOS, conversion rate, or click-through rate when running multiple creatives simultaneously?

Why Video Creative Testing Matters

  • High variability in creative performance: The same product can see drastically different engagement depending on video framing, pacing, music, and messaging
  • Account-level impact: A winning video creative can lift overall campaign efficiency and scale faster than text/image ads alone
  • Data-driven iteration: Without structured testing, you’re guessing—and Amazon’s own reporting on video performance is limited

Common Testing Approaches

  • Split campaigns: Run identical targeting, keywords, and bid strategy across separate campaigns, each with one video creative. Requires patience for statistical significance
  • Campaign budget allocation: Weight spend toward your hypothesis winner while maintaining enough traffic to the challenger to detect a real difference
  • Timeframe consistency: Run each test for at least 1–2 weeks (or until you hit 100+ video views per variant) to filter out day-of-week noise
  • Conversion-weighted metrics: ACOS and conversion rate matter more than impressions or clicks alone—focus on the full funnel

How Squatio helps

Squatio Optimize (PPC intelligence) lets you pull detailed performance data from your Sponsored Brands campaigns in one view, making it easier to spot which video creative is actually moving the needle on conversion rate and efficiency. Rather than relying on fragmented Amazon reports, you can isolate video performance by campaign and compare ACOS across creatives side-by-side—so you know which winner to scale.

Discussion

How long do you typically run a video creative test before deciding to pause a loser or scale a winner? And do you isolate video testing in separate campaigns, or do you test multiple creatives within a single campaign and rely on Amazon’s reporting?

Source: Reddit