What is a Brand Attention Score?

The Brand Attention Score in junbi answers a key question: How well does your brand capture attention?

This score is based on the percentage of attention your brand receives at every moment within your ad. It helps determine whether viewers notice your brand consistently—not just in the first few seconds but throughout the entire video.

How Brand Detection Works

junbi currently offers two types of brand detection:

1. Text-Based Detection – Identifies specific brand-related text within the video.
2. Iconic Detection – Uses a custom-trained model to recognize logos, symbols, and other brand elements, ensuring every variation is detected. (Learn more about Iconic brands here.)

In the example below, junbi successfully detects the brand name in a Lancôme ad using text-based detection. The detected brand text is highlighted with a blue bounding box.

Understanding Your Score

Your Brand Attention Score is displayed as a percentile, comparing your ad’s performance to a dataset of over 20,000 real YouTube ads.

• An 76th percentile score means your ad outperforms 76% of other ads in brand attention.

• A higher percentile indicates strong brand visibility and engagement.
• A lower percentile suggests your brand may not be getting enough attention, signaling areas for improvement.

By analyzing this score, junbi helps ensure your brand is not just present in the ad—but truly memorable.