AQ is an expert system for creative optimization, built on a theoretical framework of how creative communication works. It reads each piece of creative against five dimensions, with the strategy, the audience and the competitive position in hand. Each dimension is scored at the level of individual variables, and the scores compound into a single index that can be compared across concepts, across campaigns and over time.
It is built to optimize creative work, not to pass or fail it. The composite tells you where you stand. The reasoning behind every variable tells you why — and that is the part you can act on.
What the model is measuring
Strategic Fit reads the brief the creative was built against. A high-scoring ad on a weak brief is a brilliantly executed wrong answer, so this is settled before a single creative variable is scored.
It then qualifies everything downstream. An ad that earns attention against the wrong competitive context, or argues persuasively to the wrong audience, is not commercial value — and the gate makes that visible before the creative gets blamed.Attention is the gateway and the multiplier. It is scored as a normalised index rather than a raw total, and applied multiplicatively to engagement plus persuasion. An ad that fails to register cannot deliver any downstream effect, however well it is made.
It is not one thing either. Attention runs through four processing stages — pre-conscious arrest, relevance and curiosity, sustained grip, and distinctive processing — assessed independently before they combine. No single variable carries the index alone.Engagement measures the depth and quality of processing after attention is secured. Getting attention and holding it are separate achievements, and the model scores them separately.
Emotional engagement without distinctive brand character produces engagement that does not accrue to the brand. Deep narrative immersion can even suppress brand processing — the viewer feels it and cannot say who made it.An engaging ad is not automatically a persuasive one. Persuasion measures whether the work produces the attitudinal and belief movement the brief requires — and whether that movement attaches to the brand specifically or drifts to the category.
It runs through two independent routes: the rational one (argument, credibility, proof) and the meaning one (cultural reframe, association, brand integration). Either route alone can persuade. They are not interchangeable, and their commercial implications differ by campaign stage.Brand Lift is not a sixth measure but a translation. Five commercial outcomes are predicted from the pattern of upstream variables, each driven by a specific combination of them rather than by the composite score.
The predictions are qualified by Strategic Fit. The same AQ score means different things depending on the brief it answered — strong engagement on a weak gate generates brand warmth aimed at the wrong audience, which does not convert.AQ, across three car ads
AQ = ATTENTION × (ENGAGEMENT + PERSUASION)Car advertising is one of the most contested creative categories — big budgets, similar briefs, and broadcast breaks where your ad runs next to your competitor's. Three executions, scored across the framework.
The multiplier is doing the work here. Concept B's engagement (64) is not dramatically below Concept C's (85), but its attention failure halves the final score. An ad that generates real feeling and fails to register funds brand warmth for the category — and for the competitor in the next slot.
Each dimension resolves into its own variables, and each concept draws a different shape. The same three ads, read three ways.
Variable-level breakdown
The AQ report delivers findings at the variable level — identifying precisely where performance is strong and where intervention is warranted.
See the full report ↗