Super Search is an AI-powered image search tool. It analyzes individual images in your library to find the most relevant UGC. You can Super Search for UGC depicting specific objects, activities, or attributes.
In a nutshell, the Super Search AI understands images the same way people understand language. When you type a keyword or phrase into Super Search, the tool will find the UGC in your library that is the best possible match.
Archive's Super Search feature lets you quickly find specific content within your library using advanced AI recognition. This powerful tool helps you locate UGC based on visual elements, actions, and objects without requiring manual tagging of your content.
Find UGC featuring specific objects or products
You can use Super Search to find all your UGC depicting specific objects or products. Try combining multiple descriptive elements (color, material, shape, size, etc.) to improve your results. Here are some examples:
Search for "black leather bag" to find images featuring dark leather handbags
Try "wooden dining table" to locate content showing wooden tables
Use "blue denim jacket" to discover images with blue jean jackets
Search for "white ceramic mug" to find content with white coffee cups
Try "gold hoop earrings" to locate jewelry-focused content
Find UGC featuring specific actions or activities
You can also Super Search for UGC, which shows people taking certain actions or participating in specific activities. You can focus on the action/activity itself or combine it with other physical details. Here are some examples:
Search for "people laughing" to find joyful moments
Try "customers unboxing products" to locate unboxing content
Use "person drinking coffee" to discover coffee-related content
Search for "family cooking together" to find kitchen/cooking content
Try "friends hiking outdoors" to locate adventure-themed UGC
Understanding the 1,250 result count
When you use Super Search in visual search mode, you might notice the content count always shows 1,250 regardless of what you search for. This is not a bug.
Here's what's actually happening: Super Search doesn't work like a keyword search that counts every exact match. Instead, it scans your entire content library using meaning-based matching and returns the best results it finds, up to a maximum of 1,250. That 1,250 is a built-in cap on how many results the search engine will surface at once, not a reflection of how many items technically match your search term.
Think of it like asking someone to pull the most relevant files from a massive archive. At some point, they stop at a reasonable number rather than going through every single item indefinitely.
This only applies to "In Visual Search Only." The other search modes work differently:
In Transcripts -- shows exact match counts
In Captions -- shows exact match counts
So if you're seeing 1,250 in visual search, that's the system working as intended. If you're seeing unexpected counts in Transcripts or Captions, that's worth looking into.
Tips for Effective Super Search:
Be specific with your search terms to narrow down results
Experiment with different combinations of descriptors
Use everyday language rather than technical terms
Try alternative words if your initial search doesn't yield expected results
Refine searches by adding location details (e.g., "people at beach")
Notes: Super Search works best with a diverse content library. Results may vary depending on the visual clarity of your UGC and the specificity of your search terms.