What is Creative Guard?
Creative Guard is Bunker's native AI-powered creative pre-testing solution. It combines your brand's knowledge with artificial intelligence models to validate your creative assets before publishing, helping you detect messaging issues, guideline compliance problems, or legal risks before going live.
What does the tool evaluate?
Creative Guard analyzes each creative asset (image or video) based on two groups of modules:
Core validations:
Message Clarity: assesses whether the creative communicates its message quickly and clearly.
Platform Best Practices: verifies compliance with each platform's specific technical and framework standards (Meta, TikTok, YouTube, etc.).
Brand Guideline Compliance: reviews the correct use of logos, colors, tone, and other brand elements.
Legal & Brand Safety: identifies sensitive content to ensure safety and regulatory compliance.
Advanced intelligence (customizable):
Attention Capture: predicts the asset's potential to capture attention within the first few seconds.
Emotional Impact: identifies the emotions the asset generates and their intensity on the audience.
Performance Prediction: predicts the asset's potential performance (CTR, VTR, Engagement).
How does AI-based detection work?
Bunker processes each creative frame by frame with AI to automatically detect key elements, including:
Dominant colors
Logo presence (overall and within the first 3 seconds)
Presence and number of people
Product focus
Presence of music or voice-over
Objects present in the creative
Predominant emotion type
How do I upload and analyze a creative?
Go to the Creative Guard module from the side menu ("Power Cycle" section) and click New Analysis.
Collection Name (Required): enter an identifying name to group the assets for this analysis.
Platform (Required): select the network where the asset will be published (Google Ads, YouTube, Meta, TikTok, LinkedIn, X/Twitter, Pinterest, or Snapchat).
Placement (Optional): choose the specific placement(s) within the selected platform (for example, Instagram Feed, Story, or Reels; Facebook Feed, Story, Reels, In-Stream Video, Right Column, or Marketplace).
Report Language: choose the language in which you want the analysis results generated (English, Spanish, or Portuguese).
Files (Required): upload the creative asset by dragging it in or using the upload button.
Model (Optional, right panel): select the validation framework to apply. If the brand has a model marked as "Always active," that framework is applied by default to all analyses.
Additional criteria (Optional): if you need to add specific rules beyond the selected model, add them using the + Add rule button.
Once all required fields are complete, click Analyze (the button shows the credit cost of the analysis) to start processing.
Wait for the analysis to complete: you'll see the "analyzing..." status while it's processing.
Review the results: overall score, criteria-by-criteria checks, and recommendations.
How do I interpret the results?
Once the analysis is complete, you'll find:
Overall score: a consolidated score for the asset out of 100.
Checks: a list of passed or failed checks by criterion (for example, aspect ratio, vertical format, logo usage).
Module breakdown: detailed results for each validation (message, platform, brand, legal).
Recommendations: concrete suggestions to improve the asset before publishing it.
Asset statuses
Each creative you upload goes through different statuses within the approval workflow:
Needs Review: the analysis is ready and requires review.
New version: a new version of the asset was uploaded.
Approved: the asset was approved and is ready to publish.
Rejected: the asset did not meet the defined criteria.
"Inspire" feature
In addition to detecting issues, Creative Guard includes the Inspire feature, which generates adjustments and new creative concepts based on the pre-test results, so your team can execute them directly.
