Maximize Your Ad Performance with Subtext

Subtext is an advanced ad testing product, automating valid scientific measurements that benchmark and diagnose problems to improve performance. Subtext combines eye tracking, facial coding, and implicit emotional associations to forecast engagement & influence with greater accuracy.

Improve Ad Performance at Any Stage of Development

From preliminary concepts to the final cut, Subtext measures emotion and reason to diagnose problems, predict behavior and optimize any type of media, providing actionable recommendations for improving engagement and influence within the creative phase or in-market.

Reliable Insights for Ad Improvement

Performance benchmarks that provide greater context and interpretability. Weed out low performers, reveal strength of memory encoding, predict desirability and identify which demographic responds best. True implicit data sets empirically validated with the largest normative database in the industry provides greater reliability.

Real Correlations to Market Performance

The Proportion of Emotion model of sales forecasting is academically published with multiple global industry awards. Currently used by the largest brands in the world to forecast ad success and marketing effectiveness.

Stella Artois – Change Up The Usual

Subtext has analyzed thousands of ads to reveal key principles for improving performance. This snapshot reveals the ad's impact on memory as well as moment-by-moment valence.

  • Memory recall is maximized with the Stella brand integrated into the ad’s narrative.
  • This ad creates a desired scenario using what is own-able about the brand as a plot device.
  • In this ad the sophistication and the aspirational taste conferred by the characters impacts memory.

Subtext behavioral science benchmarks reveal the vital ad performance metrics you need to predict consumer behavior.


Advancing The Model of Persuasion

Subtext doesn't just deliver reliable scores, it compares & contrasts observations with synthesized results that provides real recommendations for improving engagement and influence.

Attention measures the interest level generated.

Subtext uses eye-tracking to measure the average percentage of consumers actively watchin the screen at any given second during the ad.

How it’s Measured:
Online eye-tracking with visual saliency analysis

Optimization Metric:
Precent of Valid Fixations + Hierarchy & Locus of Fixations

Success Metric:
Average percent of Valid Fixations during time-course of content exposure expressed as a percentile rank against normative benchmarks


The higher the attention score the better; particularly when the brand is being viewed.

Affect measures the degree of emotional response generated.

Subtext facial coding measures the net positive percentage of consumers expressing an emotional response at any given second during the ad.

How it’s Measured:
Online facial coding technology is used to indirectly measure the strength & valence of expressed emotion

Optimization Metric:
Degree of emotion expressed during content exposure

Success Metric:
Average degree of emotion over the time-course of exposure, expressed as a percentile rank vs. normative benchmarks


Both positive & negative emotion is good, when intended to engage the audience in the story.

Memory measures impact of content on brand recall & brand perceptions.

Timed response tests measures the % of consumers that can quickly and correctly recall the brand in connection with the content.

How it’s Measured:
Sentient Prime® implicit technology is used to measure both speed & accuracy of brand recall as well as implicit perceptions of the brand resulting from exposure to content

Optimization Metric:
Changes in individual brand perceptions measured before & after exposure, but real-time variations in emotional response during exposure are correlated with emotional outcomes, revealing moments where memory encoding is likely occurring

Success Metric:
Depth of memory encoding (Brand Recall), expressed as a percentile against normative benchmarks


Higher memory scores mean that more people will remember your brand in connection with the ad.

Desirability measures the impact of content on near-term demand.

Choice tests measure the percent difference in brand preference, due to ad exposure; while Association Tests measures the percent difference in personal attachment (loyalty) to the brand.

How it’s Measured:
Near-Term Preference is calculated with a highly predictive mathematical model that combines emotional & reason-based inputs, & comparing those who are exposed vs unexposed to content.

Optimization Metric:
Lift in reason-based preference & implicit emotional appeal are reported as independent metrics

Success Metric:
Desirability is calculated by equally weighting lift in near-term & long-term measures & expressing as a percentile rank against normative benchmarks


Lift in these metrics must be significant for an ad to succeed. For near-term demand the score must exceed +10%. Long-term loyalty must exceed +4.5%.

Contact Sentient for a demo of Subtext and make an impact on your research.