Skip to contentGet startedWhat Foresight predicts, your first analysis and how to read the results page.How it worksFrom upload to results, the attention and emotion models, images and videos.KPIsWhat every score measures, how it is built and how to read it.EmotionsThe eight predicted emotions, intensity and tone, and the video emotion timeline.AttentionHeatmaps, fogmap, gaze plot, attention build-up and areas of interest.BenchmarksWhat your scores are positioned against and what the comparison value means.PlatformsWhat changes when you analyse a creative inside a TikTok, Instagram, Facebook or X screen.AI insightsHow Copilot writes summaries and recommendations from your results.Using ForesightUploading, projects, Compare, exports, sharing, plans and finding past analyses.The scienceHow the models learned from human data, how they are tested and their limits.
Foresight Knowledge Center
How Foresight works
The models, KPIs and benchmarks behind every Foresight result, explained.
Popular questions
- How is the E-Score built, and is it the main KPI?
- Is a strong negative emotion less engaging than a positive one?
- Why does Clarity change so much when I pick TikTok?
- Why does a story-driven creative get a low Mental Demand?
- Which benchmark are my scores compared with?
- Can I compare face, logo and text results with each other?
- Are image and video scores directly comparable?
- Where are the campaigns I tested months ago?