Amazon SEO Services: Listing and Catalog Optimization Based on What Amazon Documents

Updated September 2026 · Written and maintained by the Progression Agency strategy team

Amazon SEO services improve how a product listing, and the catalog data behind it, matches the searches shoppers type and converts once they arrive: titles and item highlights, bullet points, backend search terms, browse-node classification, images, A+ content and the reviews Amazon’s rules allow. We do this for brand owners, FBA sellers, vendors and KDP authors who want more of their sales to come from search and less from ad spend. Progression Agency is based in New York City and works with clients across the United States and worldwide.

What do Amazon SEO services include?

They cover every field that decides whether an ASIN appears for a search and whether the shopper buys: the title, item highlights, bullets and description, the hidden search terms, the category and attributes, images and video, A+ content and the review program. Advertising is a separate service, Amazon Ads management, run from the same keyword map.

Sellers search for this work as Amazon product SEO, Amazon seller SEO or simply Amazon SEO; the scope above is the same under every name. Demand is modest and specific: “amazon seo” and “amazon seo services” each draw about 880 US searches a month and “amazon seo agency” about 210, while advertisers pay $83.35 for a click on “amazon seo agency” (Ubersuggest, September 2026).

What Amazon documents about how its search works

Amazon publishes how a product becomes findable, not how it ranks. Its Seller Central help pages describe search terms as keywords used for lexical matching between an ASIN and a query, say search does not do partial matching, and state plainly that no placement is guaranteed.

That list is short for a reason. It explains how to be eligible and relevant; it says nothing about how relevance, sales, price or reviews are weighed against one another.

What makes an ASIN searchable?

Four conditions come before any keyword work: a live, buyable offer; assignment to at least one specific, relevant browse node; no suppression for listing-quality problems; and a launch date in the past. Our audits test all four for every ASIN, because a listing that fails one of them cannot rank for anything, however good its copy.

Why does your position look different on every device?

Amazon’s troubleshooting guidance says positions shift with the shopper’s location, the type of account they are signed in to and whether they shop on desktop or mobile, and that results past a set number of pages are not shown at all. Checking your rank from one browser is a weak measure; query-level data from Brand Analytics is a far better one.

A9, A10 and ranking factors: what is documented and what is guesswork?

“A9” and “A10” are names that sellers and software vendors use for Amazon’s ranking system. Amazon’s seller documentation does not describe a ranking formula under either name, so any list of “A10 ranking factors” is inference, and we label it that way.

Why we do not sell ranking hacks

Tactics built on unproven theories, such as stuffed keyword fields, schemes that fake searches and purchases for a keyword, or bought reviews, either do nothing or breach Amazon’s policies, and several put the account at risk. Our work is limited to what Amazon documents, what Amazon’s own data can measure, and experiments we can read.

Titles and item highlights under Amazon’s 75-character limit

Amazon now limits product titles to 75 characters including spaces for all product types except media, in every store except Saudi Arabia, Egypt, Türkiye and the United Arab Emirates, and it may correct titles automatically or keep non-compliant ones out of search (Seller Central: product title requirements).

Item highlights: 125 more characters

Once a title fits 75 characters, Amazon opens a second field, item highlights, with 125 more characters for details such as material or recommended use, written as comma-separated phrases rather than sentences. Highlights appear below the title in search results and on the detail page, so the detail that no longer fits the title still reaches the shopper.

The order Amazon asks for

Amazon’s guidance orders a title as brand, then flavor or style, product type, key attribute, color, size or pack count and model number. It asks for numerals rather than spelled-out numbers, capitals only at the start of words, no subjective phrases such as “hot item”, and size and color only in child titles, never in the parent’s.

A title rewrite, worked through

Before: “BEST Insulated Water Bottle!! Stainless Steel Water Bottle for Gym Hiking Travel – Free Shipping” breaks four rules at once, with a subjective claim in capitals, a banned exclamation mark, a promotional phrase and a length of 96 characters. After: “[Brand] Insulated Water Bottle, Stainless Steel, 32 oz, Slate Blue” is 66 characters, and “double-wall vacuum insulation, leak-proof lid, fits most cup holders” moves into item highlights.

Bullet points and descriptions that answer the buying question

Bullets carry the questions a shopper asks before buying. Amazon asks for at least three, allows up to five, keeps all of them together under 1,000 characters and says they are not always indexed for search, so we write them for the decision first and for keyword coverage second (Seller Central: bullet point requirements).

• Start each bullet with a short header and a colon, then the detail, written as a fragment with no closing punctuation.
• Spell out the numbers one to nine except in names, model numbers and measurements, with a space between a number and its unit.
• Keep claims to what the packaging can verify; Amazon removes unsubstantiated performance, comparative and award claims.
• Leave out the claims Amazon lists as prohibited in bullets, such as “eco-friendly”, “anti-bacterial” or “made from bamboo”.
• No guarantees, contact details, links, emojis, trademark symbols or phrases in capitals.
• Keep fabric content, care instructions and country of origin in their own fields rather than in bullets.

Descriptions still matter

Every ASIN must have a product description, even when A+ content fills the description section for brand owners. We write it to cover usage, specifications and fit, because it remains part of the catalog data and some placements show the text rather than the modules.

Backend search terms: getting the 250-byte field right

The search terms field holds the words shoppers use that do not belong in visible copy: synonyms, abbreviations and alternate names. Amazon limits it to less than 250 bytes and counts bytes rather than characters, so an accented letter can use two or more (Seller Central: product discoverability).

• Lowercase words separated by spaces, with no commas, semicolons or dashes.
• No repetition, and either the singular or the plural form of a word, not both.
• Short words such as a, an, and, by, for, of, the and with left out.
• Spelling variants and abbreviations included; common misspellings excluded.
• No brand names, including your own, no ASINs, no profanity, no temporary words such as “new” or “on sale”, and no subjective words such as “best”.

Amazon may ignore the field when it breaks those rules, and violations can suppress the ASIN and affect account health (Seller Central: search terms). Spaces and punctuation do not count toward the limit. Since December 6, 2023, Amazon has required the field for 23 product types, from board games and blankets to balloons and beauty products.

Where our backend terms come from

We build the field from three sources: the search term report from any Sponsored Products campaigns, which shows queries that produced clicks; the Search Query Performance dashboard in Brand Analytics; and the words customers use in reviews and questions. Amazon asks sellers to avoid repetition, so terms already in the title or bullets stay out, and the bytes go to words the listing does not yet contain.

Browse nodes, item type keywords and attributes

Classification decides which shelf a product sits on, and Amazon says a product must be assigned to at least one specific, relevant browse node to be searchable at all. In the US store, item type keywords are how an ASIN is assigned to a node (Seller Central: search glossary).

Attributes matter as much as words. Amazon describes searches with built-in constraints: a search for red dresses returns products assigned to a dress node and a red color refinement, and a gluten-free food search returns only products that met a verification requirement. A listing with the right words but an empty or invalid color, size or material value can fall outside those results, which is why we fill every attribute with Amazon’s valid values.

Browse tree guides

Amazon’s category-specific browse tree guides list the exact item type keyword values, which must be entered with the same spelling and formatting, at the most specific subcategory available. “Women’s road running shoes” is the level Amazon wants; “shoes” is too broad.

Parent and child variations

Variations share one detail page under a parent ASIN, which cannot itself be bought. Amazon displays the parent’s title on the detail page and shows a child’s title only once it is in the cart, so size and color belong in the child titles.

Images and video that pass Amazon’s main-image rules

The main image wins or loses the click in search results, and if no compliant main image exists Amazon may take the listing out of search until one is provided. It needs a pure white background (RGB 255, 255, 255), the product filling 85% of the frame, a single unit shown once, and nothing that is not included in the box (Seller Central: product image guide).

Additional images and video

Amazon recommends at least six more images and one video. Images must be at least 500 pixels on the longest side, and 1,000 pixels or more enables zoom. A new image can take up to 24 hours to appear, and Amazon chooses which images to show when several sellers share a listing, so compliance also decides whose images shoppers see.

Images that break the rules

No review screenshots, star graphics or claims such as free shipping in any image, no text or prices except on swatch images, and no mannequins for clothing. Adult clothing is shown on a standing model, and footwear as a single shoe facing left at a 45-degree angle.

AI-generated people

If an image or video shows photorealistic people generated entirely by AI, Amazon requires the file to carry the IPTC keyword contains-synthetic-performer, and Amazon then adds a disclosure where applicable. Real people edited with AI tools do not need the tag.

A+ content and Brand Story

A+ content fills the product description section with modules: images, text, comparison charts and, in Premium A+, video, hotspots and carousels. Both tiers are free, and Amazon’s A+ guide says there are no additional eligibility criteria for sellers to use Premium A+ (Seller Central: A+ Content guide).

Brand Story sits in a separate “From the brand” section: one carousel module of up to 19 cards, with links to other products and the Brand Store. Submissions take up to seven business days to review and publish within 24 hours of approval. Amazon rejects A+ with pricing, promotional language, warranty or shipping details, competitor comparisons, QR codes or outside links, and A+ is not yet supported for books, music, video and DVD sold through Seller Central.

Amazon states that basic A+ content can increase sales by up to 8% and well-implemented Premium A+ by up to 20% (Sell on Amazon: A+ Content). Those are Amazon’s own figures, not an independent study, so we test a layout with Manage Your Experiments before rolling it across a range.

Reviews: what Amazon allows and what the FTC prohibits

Reviews carry legal risk as well as ranking hopes. Amazon’s anti-manipulation policy lets it suspend or terminate selling privileges, remove reviews, delist products and withhold payments (Amazon: anti-manipulation policy for customer reviews), and the FTC’s Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465, makes fake reviews, sentiment-conditioned incentives, undisclosed insider reviews and review suppression unlawful (16 CFR Part 465).

The FTC announced the rule on August 14, 2024 (FTC announcement). Amazon’s Community Guidelines add the practical detail: a seller may ask for an honest, un-incentivized review, while inserts promising gift cards for five stars, refunds offered for changing a review and warranty sign-ups that demand a review are all listed as violations (Amazon Community Guidelines).

Amazon Vine: eligibility and fees

Vine lets a brand enroll up to 30 units of a parent ASIN for invited reviewers. The seller needs a Professional account, a Brand Representative or Reseller role in Brand Registry and an active FBA listing, and the product must have fewer than 30 reviews, an image and a description, and must not be adult or digital (Seller Central: Amazon Vine). Fees are charged once per parent ASIN by units enrolled: $0 for one or two units, $75 for three to ten and $200 for 11 to 30, with no charge if no review arrives within 90 days and no refund for a review you dislike (Seller Central: Vine enrollment fee).

Why we will not buy reviews for you

Sellers looking for an agency for Amazon reviews are often offered exactly what these rules forbid: rebates for five-star reviews, review swaps, product inserts promising gift cards, or warranty sign-ups that require a review. We decline that work. Our review management and reputation management services cover the legitimate side: replying to reviews, fixing the product or listing problems they point to, and asking buyers honestly. This is how we work, not legal advice; brands in regulated categories should have claims reviewed by counsel.

Does Alexa for Shopping change Amazon SEO?

It raises the stakes on accurate catalog data. Amazon has combined its Rufus shopping assistant with Alexa; in the Amazon store the result is called Alexa for Shopping, available to US customers in the Amazon Shopping app and on Amazon.com without a Prime membership (Seller Central: Alexa for shopping).

Shoppers use it to get recommendations, compare products, check price history over 30, 90 or 365 days, understand product details and read what other customers say. For product questions such as dimensions or washability, Amazon says the assistant answers from the product’s detail page, and it asks sellers who see a wrong answer to correct the detail page first. That is the most concrete thing Amazon has said about optimizing for its assistant: the facts have to be on the page, complete and correct.

What we change for AI-assisted shopping

Nothing Amazon publishes describes a separate ranking system for the assistant, so “AI SEO for Amazon” in practice means catalog completeness:

• Every relevant attribute filled with a valid value, including dimensions, materials, capacity and compatibility.
• Bullets that answer the questions shoppers raise in reviews and in the questions section.
• Premium A+ Q&A modules for questions that need more room than a bullet.
• Claims limited to what the packaging and evidence support, since an assistant repeats what the page says.
• A monthly read of new reviews, because the assistant summarizes what other customers report.

How do sellers and shoppers use ChatGPT, Claude, Perplexity, Gemini and Copilot around Amazon?

In two different ways. Sellers ask assistants to recommend an Amazon SEO agency or expert, and shoppers ask them which product to buy before they open Amazon at all; both answers are assembled from pages the assistant can read.

When a seller asks for an Amazon SEO agency

Prompts look like “best Amazon SEO agency for a small kitchenware brand” or “is it worth paying for Amazon SEO?”, and the pages that rank for the equivalent searches are the pages assistants find. For “amazon seo services” in September 2026, Google showed an AI Overview above results that mixed agency service pages with two “best Amazon SEO agencies” roundups and threads on Reddit and Quora, and a domain with an authority score of 1 held position seven (Ubersuggest SERP data). A clear, specific service page can be cited without a large site behind it, and the same holds in ChatGPT, Claude, Perplexity, Gemini and Microsoft Copilot when they search the web.

When a shopper asks which product to buy

Shopping prompts such as “quietest blender for a small apartment” send assistants to brand sites, retailer pages, review sites and forums. A brand whose own product pages carry the same specifications as its Amazon listing gives the assistant one consistent answer to repeat. Our AEO for ecommerce and LLM SEO work covers those pages.

Using AI tools to write listings

ChatGPT and Claude draft titles and bullets quickly, but they cannot see Amazon’s search data, may not know current rules such as the 75-character title limit, and will happily write claims Amazon prohibits. We use them for first drafts and variations, then check every line against Amazon’s requirements and the product facts before anything is published.

What we publish so a brand gets named

Product pages on the brand’s own site with full specifications, a plain-text FAQ, identical facts on Amazon, the site and social profiles, and mentions in reviews and publications that are earned rather than bought. Those are the sources an assistant can cite by name.

Measuring Amazon SEO in Amazon’s own data

We measure listing work with Amazon’s reports, not screenshots of a rank. For brand representatives in Brand Registry, the Search Query Performance dashboard reports query volume, impressions, clicks, add-to-carts and purchases for a brand’s top queries, up to 1,000 of them in the brand view, covering organic and Sponsored Products results (Seller Central: Search Query Performance).

Manage Your Experiments

Brand representatives can A/B test titles, main images, bullet points, descriptions, A+ content and Brand Story on products with enough recent traffic. Amazon says a test run to significance can return a result in as little as four weeks, and suggests 8 to 10 weeks when you choose the duration yourself (Sell on Amazon: Manage Your Experiments).

Suppressed and unsearchable listings

Listings removed from search for quality problems appear as search suppressed in inventory, and Amazon’s troubleshooting tool diagnoses why a specific ASIN does not appear for a search. We check both every month, because one invalid attribute can hide a product that was selling the week before.

How much do Amazon SEO services cost?

We price the work from the SEO planning ranges we publish. Catalog size, the number of ASINs that need new content and whether A+ design is included drive the quote, which always follows a written scope.

Product photography and video are quoted separately through our video production team, and the ranges match our SEO pricing page. Advertisers bid $83.35 a click for “amazon seo agency” and $34.61 for “amazon seo services” on Google (Ubersuggest, September 2026), which is what firms pay to be found by sellers who want this work.

What moves the price

Five things account for most of the difference between the low and high end of each range:

• The number of parent ASINs and variations that need new content.
• The number of marketplaces and languages.
• How many A+ modules and Brand Story cards need design.
• The state of the catalog: suppressed, misclassified or duplicate listings take longer to fix.
• Whether the program also feeds an advertising account’s keyword map.

Our Amazon SEO process

Eight steps, run product line by product line, with eligibility fixed before any copy is written.

1. Eligibility audit: offers, browse nodes, suppression and launch dates for every ASIN.
2. Query research: search term reports, Brand Analytics and customer language, mapped to ASINs.
3. Titles and item highlights rewritten inside the 75-character rule.
4. Bullets and descriptions written for the buying decision.
5. Search terms and attributes rebuilt with valid values.
6. Images, video and A+ content brought up to Amazon’s standards.
7. Reviews: Vine enrollment for eligible new ASINs and neutral request practices.
8. Testing and reporting: experiments where traffic allows, and a monthly read of query performance.

How do you choose an Amazon SEO agency or expert?

Choose the provider that separates what Amazon documents from what sellers believe. Use the table, and check each answer before you sign.

An Amazon SEO expert working alone can suit a small catalog in one marketplace. A larger catalog, several marketplaces or A+ production across dozens of parent ASINs needs an Amazon SEO agency with writers, designers and someone reading the data.

FBA sellers, vendors and KDP authors

The rules above apply to everyone; each account type adds its own details.

An Amazon sellers SEO checklist for FBA brands

For FBA brands the order is: confirm every ASIN is searchable, fix classification and attributes, rewrite titles and highlights, rebuild search terms, bring images and A+ up to standard, then enroll new ASINs in Vine, which requires an active FBA listing. Amazon FBA SEO is the same discipline; fulfillment only changes which programs you can use.

Vendors in Vendor Central

Vendors publish A+ through A+ Content Manager in Vendor Central, and vendor-supplied content lands on the same detail page shoppers see, so we audit it against the same title, bullet and image requirements and the same latency expectations.

Amazon KDP SEO for books

KDP lets authors add up to seven keywords or short phrases, and asks them to avoid subjective claims, time-sensitive words, program names such as Kindle Unlimited, unrelated author names and brands they do not own (KDP: keywords). Amazon’s glossary reserves subject keywords for media products. Book advertising sits with our Amazon Ads team, and launch work with our book marketing service.

Related services for listings, stores and search

• Amazon Ads management: Sponsored Products, Sponsored Brands, DVA+ and Brand Stores.
• Ecommerce SEO: Google visibility for your own online store.
• Shopify SEO: the same discipline on a Shopify storefront.
• Product and brand copywriting: listing, A+ and site copy from our copywriters.
• Etsy shop optimization: search and ads for Etsy sellers.
• Review management: replying to and learning from reviews.
• AEO for ecommerce: product data AI assistants can quote.
• Conversion rate optimization: testing on the store you own.
• Video production: product video for listings and A+ modules.
• Marketing analytics: one view of Amazon and non-Amazon sales.
• Book marketing: for KDP authors and publishers.

Frequently asked questions