About AI Creative Insights by Decode
AI Creative Insights by Decode is a tool from Entropik that predicts how audiences will react to ad creatives before media spend goes live. Users upload video, banner, social, or OOH assets and receive attention heatmaps, second-by-second emotion curves, and prescriptive suggestions for changes. The tool also includes category benchmarking and side-by-side variant comparison, with a Synthetic Audience feature on the roadmap for persona-level predictions.
Review
AI Creative Insights takes a specific bet: that pre-launch creative testing can move from weeks-long research studies to a few minutes of automated analysis. The tool draws on Entropik's underlying emotion and attention AI, originally built for their Decode research platform, and applies it to a faster, self-serve workflow. For teams that currently ship most ads on gut feel, it introduces a consistent, repeatable check before budget gets committed.
Key Features
- Attention heatmaps that forecast where viewers' eyes land first and what gets missed.
- Second-by-second emotion curves tracking likely viewer sentiment across the duration of a video creative.
- Prescriptive suggestions that name what to change, why, and the expected impact, shaped with help from Claude by Anthropic.
- Category benchmarking to see how a creative stacks up against norms for its vertical.
- Side-by-side variant comparison for evaluating multiple versions of an ad in one view.
Pricing and Value
New users can test 5 creatives free using the code PHCREATIVE. Beyond that, the page does not detail standard subscription or per-test pricing tiers. The reference content mentions that traditional pre-testing carries a large research budget and timeline, positioning this tool as a faster alternative, but specific post-trial costs are not yet defined publicly.
Pros
- Returns results in minutes, which fits into tight campaign timelines where traditional pre-testing would lag.
- Heatmaps and emotion curves give visual, second-by-second feedback rather than a single aggregate score.
- Integrates with Figma and Canva, so design teams can test frames directly from their working files.
- Prescriptive suggestions read as actionable creative advice, not raw data output.
- Built on attention and emotion models trained on real viewer data, as described by the engineering team.
Cons
- The engineering lead notes that predictions are a forecast, not a guarantee, and won't replace live testing for a team's largest campaign of the year.
- Synthetic Audience, which would show predictions broken down by persona, is listed as early access only and not yet part of the standard product.
- The tool is not well suited for teams that require statistically validated, human-subject research for every creative decision, since it models rather than measures actual viewer response.
AI Creative Insights fits teams that produce a high volume of ad variants and currently test few of them. It works as a triage layer - flagging which creatives deserve more investment and which have clear attention or emotion drop-offs. For a flagship campaign with a seven-figure media buy, the makers themselves position this as a complement to, not a replacement for, research with real people.
Open 'AI Creative Insights by Decode' Website
Your membership also unlocks:
