Overall Score: Understanding Baseline Metrics in junbi

When you test an ad in junbi, the platform generates an Overall Score and three specific effectiveness drivers. These metrics are not raw data points. They are normalized percentile scores. This means your results are automatically compared against a massive benchmark database of YouTube video ads to show exactly how your creative stacks up against the competition.

The Strategic Value

If you only look at raw visual attention data, it is incredibly difficult to know if your ad is actually performing well. By translating raw data into normalized percentiles based on your specific Ad Type and Viewer Mode, junbi does the heavy lifting for you. You get immediate contextual insights without needing to build your own database of thousands of ads.

How your scores are calculated

1. The Overall Score: This is your ad-level prediction summary. It is the normalized average of your three effectiveness drivers: Cognitive Ease, Brand Attention, and Ad Breakthrough. It features a color-coded gauge showing both a numerical percentile score from 1 to 100 and a qualitative quintile ranking from "bad" to "excellent." For example, a score of 96 means your ad performs better than 96 percent of ads in the benchmark database. This is a summary metric and does not have a timeline.
2. Brand Attention: This metric measures the total amount of visual attention your brand assets secure throughout the entire video. A higher percentile indicates your logo or product is successfully seen and remembered.
3. Ad Breakthrough: This score measures how effectively your creative grabs raw attention and competes against surrounding platform interface distractions. A higher percentile means your ad cuts through visual clutter rapidly.
4. Cognitive Ease: This evaluates how easy your visual message is for the human brain to process and understand. A higher percentile indicates that the frame layouts are clean, simple, and easy for viewers to absorb.

The Raw vs. Normalized concept

To make your data actionable, junbi automatically converts raw physical data into a percentile. For example, the raw score for Ad Breakthrough calculates the exact percentage of physical visual pixels landing on your YouTube ad relative to interface elements like the comment section or sidebar thumbnails. The platform then processes this raw number through the database to output the finalized, normalized percentile circle you see on your dashboard.

💡 Pro Tip

Never optimize blindly. Because the Overall Score is an ad-level summary metric without its own timeline, you cannot fix it by tweaking the video at random. Always look at the second-by-second data inside the effectiveness drivers over time dashboard for the three underlying performance drivers. Locate the precise frames where a score dips, adjust those specific elements, and re-test.