How Image Search Technology Works
The image search feature available through KakoBuy Spreadsheet is powered by AstroReps, which uses advanced computer vision and machine learning to analyze visual features of uploaded images. Here's what happens when you upload a photo:
Feature Extraction — The AI analyzes the image for visual features including color palette, shape, texture, pattern, and structural elements. For a sneaker, it might identify the silhouette, colorway, sole shape, and any visible branding.
Database Matching — These features are compared against a massive database of product images from Chinese platforms. The system finds products with the most similar visual characteristics.
Ranked Results — Results are ranked by visual similarity, with the closest matches appearing first. You can then browse through results to find exactly what you're looking for.
The technology is particularly effective for items with distinctive visual characteristics — unique colorways, specific patterns, or recognizable silhouettes. It works less well for very generic items or photos with poor lighting and composition.
How to Use Image Search Effectively
Get the best results from image search with these techniques:
- 1
Use high-quality, well-lit photos — blurry or dark images produce poor results
- 2
Crop the image to focus on the specific item you're searching for
- 3
Use photos with a clean background when possible — busy backgrounds confuse the AI
- 4
For shoes, use side-profile photos which show the most distinctive features
- 5
For clothing, use flat lay photos that show the full item clearly
- 6
Try multiple photos of the same item from different angles if the first search doesn't find what you need
- 7
Use the "Search by Image" button on KakoBuy Spreadsheet's homepage or search page
- 8
Browse through multiple pages of results — the exact match might not be on the first page
Best Use Cases for Image Search
Image search is particularly powerful in several specific scenarios:
Social Media Finds — You see someone wearing something on Instagram or TikTok and want to find it. Screenshot the item, crop to focus on it, and upload to image search. This works especially well for sneakers and distinctive clothing pieces.
Celebrity Outfits — Want to cop a celebrity's look? Image search can find similar items to what they're wearing, even if the original is a rare or expensive piece.
Vintage and Archive Pieces — Looking for a specific vintage colorway or archive piece? Image search can find current sellers offering similar items.
Unknown Items — You have a photo of something but don't know what it is. Image search can identify the item and find similar products.
Exact Colorway Matching — When you need a very specific colorway that's hard to describe in words, image search finds it instantly.
Combining Image Search with KakoBuy Spreadsheet
The most powerful workflow combines image search with KakoBuy Spreadsheet's curated database. Here's how to use both effectively:
1. Use image search to identify the item and find similar products on Chinese platforms 2. Note the product name, model number, or any identifying information from the search results 3. Search for that item on KakoBuy Spreadsheet using the keyword search 4. If it's in our database, you'll get a curated listing with QC photos, community reviews, and a direct KakoBuy link 5. If it's not in our database, you can use the image search results to find the item directly on KakoBuy
This combined approach gives you the best of both worlds: the discovery power of image search and the curation and quality verification of KakoBuy Spreadsheet.
Tips for Better Image Search Results
Advanced tips for power users:
Use Google Lens first to identify the item, then use that information for keyword search
For sneakers, use StockX or GOAT to identify the exact model and colorway before searching
Screenshot from video at the clearest frame — pause and screenshot rather than using blurry motion frames
Use image editing to enhance contrast and clarity before uploading
Try searching for the brand + model name after image search identifies the item
Check multiple image search tools — different tools have different databases
