Research dispatches
Field notes from the marketplaces.
Long-form research on algorithm shifts, category dynamics, and seller tactics — measured against our panel of 17,932 listings. Published weekly, free to read.
Amazon listing images: the small lever everyone overweights.
Image optimization gets more consultant ink than almost any other listing topic. The actual A10 weight is small. We measured 11 image variables across 6,800 listings; here's where image investment pays back and where it doesn't.
Read dispatch→Amazon review velocity, explained: why steady beats burst (and what Vine is doing wrong).
Review velocity is the third-largest weight in the post-May A10 model and the one most sellers think they understand but don't. We analyzed 14,200 Vine campaigns and a control group of organic-only launches. The timing pattern is counterintuitive.
Read dispatch→We ran 540 Amazon bullet A/B tests. Outcome-led bullets beat feature bullets by 2.1×.
Bullet copy is the most-rewritten and least-measured part of an Amazon listing. We ran 540 controlled bullet swaps across nine subcategories. The patterns that won, the patterns that lost, and a template you can steal.
Read dispatch→Does PPC actually move organic rank? A 7,418-listing natural experiment.
Amazon says no. Most sellers act like yes. We ran the experiment that settles it. The short answer: PPC works for organic, but only if your listing is already converting.
Read dispatch→Stop writing 200-character Amazon titles. We measured 9,184 listings and the sweet spot is way shorter.
The 200-character ceiling is no longer the optimal target. We measured CTR, CVR, and rank for 9,184 listings across three length buckets. Here's where the sweet spot actually lives.
Read dispatch→Amazon's A10 just got rewritten. Here's what 17,932 listings told us.
Amazon shipped a quiet A10 update sometime in early May. We measured the damage (and the openings) across 17,932 listings, and the story isn't what most consultants are telling you.
Read dispatch→Also available: 中文研究简报