Amazon Listing Image Audit: The 13-Point Checklist

Every listing audit template gives images the same treatment: one checkbox — “images high quality?” — wedged between keyword coverage and restricted terms. Meanwhile the gallery decides your click-through rate in search and carries most of your conversion on the page. The most important creative asset on the listing gets 3% of the audit.

This is the missing piece: a dedicated image audit. Thirteen checkpoints, a severity score for each finding, and a fix order that puts traffic before conversion and conversion before polish. Run it quarterly; it takes about an hour per parent ASIN once you have the files.

Before You Start: Audit the Files, Not the Page

Auditing by scrolling your own listing is how image problems survive for years. The product page shows you a compressed 500-700px derivative inside a viewer built to flatter it. What you need on your desk is what Amazon actually serves:

1. Export your live gallery — every slot, every variant, A+ modules included — as one ZIP with ASINCrate . The files come back at original resolution, named by slot (MAIN, PT01…), foldered by variant. (How Amazon’s image delivery actually works is covered in our image download guide .)
2. Check what came back against what you think you uploaded. The first finding of almost every audit happens right here: a variant still carrying last year’s packaging, a slot you forgot existed, a hero image that was “temporarily” swapped during Prime Day and never swapped back.

Auditing the exported files has a second benefit: it is exactly the view a competitor running an image analysis on you sees.

The 13-Point Checklist

Traffic layer — the main image (checkpoints 1-3)

1. Main image compliance. Pure white background (RGB 255,255,255), product only — no props, no badges, no text overlays, no packaging unless the product is the packaging. Compliance failures here risk suppression, which is why this is checkpoint #1: a suppressed listing makes every other checkpoint irrelevant.

2. Thumbnail impact at 160px. Resize your MAIN to 160px — the size it actually competes at in search results — and place it next to your top five competitors’ mains at the same size. Does yours fill the frame (85%+ product coverage)? Is the product orientation instantly readable? Is there any visual reason to click yours? This single test predicts CTR better than any full-size review.

3. Resolution and zoom. At least 1000px on the longest side to activate zoom; 1500-2000px is the working standard (the full Amazon image size requirements — formats, the 85% main-image rule, rejection causes — are broken down here). Zoom is one of the highest-converting interactions on the page and low-resolution uploads disable it silently — no warning, no error, just a dead feature.

Conversion layer — the gallery (checkpoints 4-8)

4. Gallery depth. All 7 visible slots filled (9 where the category allows). Empty slots are free conversion surface handed back to Amazon. But depth without purpose is noise, so:

5. Slot logic. Each image must have a distinct job. The proven sequence: MAIN → key benefit infographic → scale/size reference → lifestyle in use → feature close-up → comparison or how-to → social proof or guarantee. If two images do the same job, one of them is wasted inventory.

6. Mobile-first legibility. 70%+ of Amazon traffic is mobile. Check every infographic on a phone: text must survive a 375px-wide screen without pinching. The working rule is max 6 words per callout, minimum ~28px equivalent type at export size. Desktop-designed infographics with paragraph text are the most common conversion leak in the entire audit.

7. Scale reference. “Smaller than expected” is one of the most repeated phrases in negative reviews across categories. One image must anchor size unambiguously — in-hand, next to a common object, or on-body. Dimensions printed in text do not count; nobody visualizes centimeters.

8. Lifestyle presence. At least two images showing the product in real use, with the target customer visible. Product-only galleries read as catalog pages; lifestyle images let the shopper cast themselves as the user — and they double as your defense when competitors’ galleries are all studio shots.

Trust layer — beyond the gallery (checkpoints 9-13)

9. Variant image parity. Open every child variant and compare galleries. Each color must show that color; each size variant’s scale shot must match. The default variant almost always has the full treatment while variant #7 carries three stale images — invisible from the default view, and a direct driver of “item not as described” returns.

10. A+ imagery quality. A+ Content modules should use custom-shot imagery, not recycled gallery files at different crops. Include a comparison chart (against your own range — never name competitors) and check that module images survive mobile stacking, where Premium A+ layouts collapse into a single column.

11. Video slot. Is the video slot filled? A 15-60 second demo outperforms an empty slot in essentially every category. If you have review videos but no seller video, that is a gap a competitor’s video workflow will spot before you do.

12. Review-photo contradiction check. Read your most recent 20 customer photos against your gallery. If customer photos consistently show a different color tone, a smaller-looking product, or a flimsier texture than your gallery promises, the gallery is writing checks the product doesn’t cash — that is a returns problem and an ammunition source for any competitor analyzing you.

13. AI-generated people disclosure. If a gallery image, A+ module or video shows a photorealistic person generated entirely by AI, Amazon’s July 2026 rule expects the file to carry the keyword contains-synthetic-performer in its XMP dc:subject field before upload — a real model retouched with AI is out of scope. Audit the exported copies, not the masters: compression and asset-manager round-trips strip metadata silently, and the keyword only reaches a live image by re-uploading it. Reading the field back takes ExifTool or a browser-side checker such as DiscloseTag , which writes and re-reads the standard XMP field without the file leaving the browser.

Scoring: Turn Findings Into a Fix Order

Rate every failed checkpoint:

Fix order is mechanical: all Criticals this week (traffic and compliance first — checkpoint 2 alone can move rank), Highs this month, Mediums when the next photo shoot is scheduled anyway. Resist the temptation to polish a Medium while a Critical sits open; a beautiful infographic changes nothing if the main image loses the click.

After the Audit: Benchmark It

A self-audit tells you what is broken against best practice. It cannot tell you what is broken against your category — whether your 4/5 gallery is competing with 5/5 galleries or 3/5 ones. That is the job of the companion workflow: run the competitor image analysis on your top 5-10 rivals, score them on the same dimensions, and read your audit findings against their scores. The overlap — checkpoints where you fail and the category leads — is your priority list, pre-sorted.

Frequently Asked Questions

How is this different from a general listing audit?

General audits cover indexing, keywords, compliance and pricing; images get one checkbox. This checklist spends the full hour on the layer that drives CTR and most of conversion.

Why audit exported files instead of the live page?

Because the page shows flattering derivatives. The exported files show true resolution, dead zoom, stale variants and forgotten slots — the same raw view competitors see when they analyze you.

How often should I run it?

Quarterly, plus immediately after a CTR drop, a conversion drop with stable traffic, a new visually-aggressive competitor, or any packaging change.

What resolution do my images need?

1000px minimum on the longest side to keep zoom alive; 1500-2000px as the working standard.

What is the most common critical finding?

Variant image gaps — full gallery on the default child, two stale images on everything else. Invisible from the default view, expensive in returns.

Conclusion

An image audit is the cheapest conversion project on Amazon: the assets already exist, the problems are objective, and half the fixes are re-uploads rather than reshoots. The discipline is simply looking at your actual files — at thumbnail size, on a phone, variant by variant — instead of admiring the desktop product page.

Export your gallery, run the thirteen checkpoints, score the findings, fix Criticals first. Then benchmark the result with the competitor image analysis workflow — because a clean audit is table stakes, and the next rank position belongs to whoever reads the category best.