Amazon Listing Optimization Priority List

Export one month for every listing, with the same dates in each file.

• Business Report (Detail Page Sales and Traffic by Child Item): sessions and unit session percentage per product
• Search Catalog Performance: each product's path from search impression to click, cart and purchase
• Search Query Performance: the same path for each search term, with your brand's share of the total

The Business Report starts counting when a shopper opens your product page, so it misses everyone who scrolled past you in search. Amazon's guide to Brand Analytics says Search Catalog Performance shows impressions, clicks and click rates, cart adds, purchases and conversion rates¹. Both Brand Analytics dashboards need Brand Registry.

Read unit session percentage with care. Amazon calculates it by dividing units ordered by sessions, so one shopper who buys four counts four times. Compare it only across listings with similar units per order.

A sale from search passes two gates: the shopper sees your product in results and clicks, then reads the product page and buys. Work out both rates for each listing: click rate (clicks divided by impressions) and conversion rate.

A listing can convert well and still lose most of its potential sales in search results, which a conversion report alone never shows.

Give each listing a target from your own catalog: the best click rate and conversion rate among similar products at a similar price. Your best listing shows the rate is reachable with your prices and reviews.

Then put a dollar figure on each gap. For the product page, multiply sessions by the conversion gap and your profit per unit. For search results, multiply impressions by the click-rate gap, the listing's conversion rate and profit per unit.

Run two checks before working on any listing. A product pricing template shows whether each sale makes money after fees, and a reorder point template shows whether stock will last. A listing that sells out mid-test proves nothing.

When click rate is the gap, start with the main image. Amazon's seller guide advises plain white backgrounds, with the product filling 85% or more of the frame². Check it at thumbnail size on a phone.

Next, the title. Amazon announced on its Seller Forums that from July 27, 2026, titles in every category except media must be 75 characters or less, including spaces³. A new Item Highlights field adds 125 characters for materials or uses, and shows with the title in search results. Older guides still say 200, including Amazon's own SEO guide from September 2025.

Lead with your brand, what the product is and the one detail shoppers compare, such as size or count. Amazon said it would gradually replace titles still over the limit with its own AI suggestions, so write yours before it does.

Your star rating and review count show in search results too. Northwestern University's Spiegel Research Center found in 2017 that a product with five reviews was 270% more likely to be bought than one with none⁴. The gain tapered quickly after the first five, and purchase likelihood typically peaked at ratings between 4.0 and 4.7, then fell as ratings neared 5.0.

When shoppers click and leave, the page usually left a doubt unanswered. Your 1- to 3-star reviews often say which. Note which complaints repeat, and read rivals' reviews the same way. A competitor analysis counts their complaints by type.

• "Smaller than I expected": a photo of it in someone's hands, and the size in the first bullet
• "Hard to fold": three images showing each step, and the folded size
• "Didn't fit in my car": folded dimensions in an image and in a bullet

Use the images after the main one for what words can't show: scale, what's in the box, how it works. For bullet points, Amazon suggests a one- or two-word description, then an explanation in 100 characters or fewer².

Answer critical reviews with the fix, not an argument. With Brand Registry, Amazon's Customer Reviews tool lets you contact anyone who rates a product below three stars⁵ to offer a refund or help. An ecommerce negative review response template matches each review to its order before you write.

You only learn from an edit if traffic can show its effect. NIST's statistics handbook gives the formula for the minimum sample size to detect a change in a proportion⁶. By that formula, a listing steadily converting at 6% needs about 4,700 sessions after an edit to detect a rise to 7%, at 95% confidence and 80% power.

At 20,000 sessions a month, that takes about a week. At 500 a month, it takes more than nine months, and by then the season and your prices have moved. On quiet listings, make every fix at once and accept you won't know which one worked.

1. 1

   Test the change on Amazon if you can

   Amazon's Manage Your Experiments⁷ tool splits shoppers at random between two versions of a title, image, bullets or A+ content. It needs Brand Registry and enough recent traffic.

2. 2

   Let it finish

   Amazon recommends 8 to 10 weeks if you set the length yourself. Its default runs until there's a winner, sometimes in as little as four weeks.

3. 3

   Otherwise, compare equal periods

   Use the same number of days before and after, and note any price, stock or ad changes.

Rather not join the reports by hand? The tool at the top of this page ranks your listings by what a fix is worth and drafts the edits for the top few.