How these benchmarks are measured

The category taxonomy

We start from Amazon's own department taxonomy and walk it three levels deep, using the same browse-node IDs that Amazon product pages report. That gives every benchmark a stable category identity — the same IDs our free Amazon audit uses to match your listings to their category benchmark.

The cohort: top 10 best sellers

For each category we take the top 10 ranked products at crawl time. That's a deliberate choice: these benchmarks describe category leaders, not category averages. The premise is that if you want to win a category, the relevant comparison is the listings already winning it.

What we measure per listing

• Gallery image count
• Presence of a brand-uploaded product video
• Count of Amazon Influencer / creator videos on the listing
• Presence of A+ (enhanced brand) content
• Star rating and review count (as context, not a content metric)

Per category we report medians for the counts and cohort shares for the presence metrics (shown as “X of 10 top sellers”). Medians are stored unrounded and rounded up for display — “meeting the benchmark” means having at least that much.

Reading the numbers honestly

• With a cohort of 10, presence shares move in 10-point increments and medians are coarse. Treat them as a clear directional bar, not a decimal-precise statistic.
• This is a snapshot dated August 2026, refreshed roughly quarterly. Every page states its snapshot date.
• Coverage stops at depth 3 of Amazon's taxonomy; Amazon has deeper niche nodes we don't benchmark yet.
• We exclude categories where these content metrics aren't meaningful — digital goods (apps, ebooks, music), physical media (books, CDs, movies), marketplace categories (Handmade, collectibles), and service listings (warranties, plans). The published library covers 1,777 physical-product categories.
• A small number of categories returned fewer than 10 products; those pages are marked with their actual sample size.