Image Quality
How Vantage scores product images — count, alt text, variety, and tips to improve
Last updated: 6th June 2026
Image Quality
Weight: 8.9% — Product images are critical for shoppers, and the metadata around them (especially alt text) is critical for AI agents. AI systems can't "see" your images — they rely on alt text and image metadata to understand what's shown.
What Vantage checks
Vantage evaluates four aspects of your product imagery:
Image count (40 points)
More images give AI agents (and shoppers) a more complete picture of your product.
| Image count | Points |
|---|---|
| 5 or more | 40 |
| 3-4 images | 30 |
| 1-2 images | 15 |
| No images | 0 |
Primary image (20 points)
Does your product have a primary (featured) image set? This is the image that represents the product in search results and listings.
- Primary image exists: 20 points
- No primary image: 0 points
Alt text coverage (25 points)
Vantage calculates what percentage of your images have descriptive alt text, then multiplies that ratio by 25. If 4 out of 5 images have alt text, you earn 20 out of 25 points.
Image variety (15 points)
Products with multiple images showing different aspects (different angles, details, lifestyle shots) score higher. Variety is only assessed when a product has 3 or more images, and Vantage detects it by counting distinct alt text values across images.
- 3+ images with 2 or more distinct alt texts: 15 points
- 3+ images with fewer than 2 distinct alt texts: 8 points
- Fewer than 3 images: 0 points
Tips to improve
- Upload at least 5 images per product. Show the product from multiple angles, in context, and with close-ups of details.
- Write descriptive alt text for every image. "Nike Air Max 90 — side profile showing mesh upper and visible Air unit" is far better than "product image" or leaving it blank.
- Vary your alt text. Each image should have unique alt text that describes what's actually shown in that specific photo.
- Always set a primary image. Make sure your best product shot is marked as the featured image.
- Describe what's unique in each image. Front view, back view, close-up of fabric texture, size comparison — call it out.
Before and after
Poor example:
1 image, no alt text
- Only one image, no alt text, no variety
- Score: 35/100 (15 for having 1 image + 20 for the primary image being set)
Good example:
5 images:
- "Hydro Flask 32oz Wide Mouth — front view, Pacific blue"
- "Hydro Flask 32oz — lid detail showing leak-proof seal"
- "Hydro Flask 32oz — size comparison with standard water bottle"
- "Hydro Flask 32oz — in-use during hiking"
- "Hydro Flask 32oz — bottom showing non-slip pad"
- 5 images (40pts) + primary set (20pts) + all have alt text (25pts) + varied alt text (15pts)
- Score: 100/100
Common mistakes
- No alt text at all. This is the most common miss. AI agents can't interpret images without alt text.
- Generic alt text. "Image 1," "Product photo," or the product title copy-pasted on every image — none of these help.
- Only one image. Even a great single photo can't score above 35 on this dimension.
Quick win: go through your top 10 products and add descriptive alt text to every image. It takes 30 seconds per image and can boost your Image Quality score significantly.