AdLens metric definition

    Attention

    How effectively the video creates and sustains reasons to keep watching across its runtime, including the strength of its opening and pacing.

    What AdLens evaluates

    • Opening strength, information density, visual change, pacing, and unresolved questions
    • Moments where repetition, complexity, or slow progression may weaken momentum
    • The shape of the attention curve, not only its average score

    What it does not prove

    • The attention curve is a model-based creative assessment, not eye tracking or platform watch-time data.
    • Attention alone does not show whether viewers understood or believed the message.

    How to read the score

    The same shared interpretation bands apply to every AdLens metric. The model rationale and evidence explain why a particular creative falls within a band.

    0–65

    Needs work

    Important signals are missing, inconsistent, delayed, or difficult to follow.

    66–85

    Effective

    The creative communicates well, with identifiable opportunities to make it stronger.

    86–100

    Strong

    The relevant signals are clear, consistent, and well integrated across the video.

    Report example

    See Attention in context

    87
    Kellogg’s Raisin Bran

    Observed in the report

    The concept is entertaining, but multiple mission vignettes diffuse momentum through the middle.

    Proposed creative edit

    Compress one middle vignette and reach the product-benefit payoff earlier.

    This proposed edit has not been re-analyzed and does not claim a new score or uplift.

    Open the evidence in the demo report

    Ways to improve

    1. 1.Create an immediate reason to keep watching in the opening seconds.
    2. 2.Compress repeated beats and move the product payoff earlier.
    3. 3.Use changes in scene, sound, or information to renew attention with purpose.

    Benchmark context

    The metric score evaluates the creative itself. Its percentile and difference from average are calculated afterward from the eligible comparison set.

    Read benchmark methodology