What Amazon SEO actually is — A9, A10, and the three things the algorithm cares about.

Amazon SEO is the discipline of getting a listing to appear, in the right slot, for the searches its target shoppers actually run on Amazon's search bar. That sentence sounds obvious until you start unpacking which searches, which slot, which signals, and most importantly, which algorithm.

A9 and A10 — Amazon's ranking systems

A9 was Amazon's original ranking algorithm — built to optimise for one thing: which listing is most likely to convert this query into a sale? A10 is the iterative successor that re-weights toward organic signals (sessions from off-Amazon traffic, conversion rate on detail-page sessions, account health) and weights pure keyword-stuffing less heavily. In practice, modern Amazon SEO is "A10-aware" but most of the levers — title structure, bullet keyword placement, backend search terms — still match what A9 documented.

The two together optimise for what Amazon's bottom line cares about: revenue per shopper session. Listings that earn revenue per session faster than alternatives climb; listings that don't, fall.

Three things the algorithm cares about

1. Findability. Is the listing in the index for the search term at all? A keyword that never appears on the listing in any indexed field is invisible to A10 for that query — no rank, no impressions, no clicks. This is binary, and it's the most fixable layer.
2. Relevance. Of all indexed listings, how strongly does Amazon think this listing matches the query? Weighted by which field the keyword appears in (title carries the most weight, then bullets, then backend search terms, then A+ alt text), how early in the field the keyword appears, and whether the listing's category matches the dominant category for that query.
3. Performance. Of all relevant listings, which one earns revenue from this query fastest? Click-through rate from the SERP, conversion rate on the detail page, review count, average rating, Buy Box ownership, FBA / Prime eligibility, and price competitiveness vs the SERP cohort.

Field weights — where keywords actually count

• Title — the highest-weight field. The first 60–80 characters carry the most signal; the last third counts but is less impactful. A keyword in the title is treated as a strong topicality signal.
• Bullets — second-highest weight. Indexed in full, with stronger weight on the first bullet. Bullets are also the field shoppers actually read, so they have to balance SEO with conversion.
• Backend search terms (generic keywords) — indexed but invisible to shoppers. The 250-byte limit per field is enforced strictly; anything past byte 250 silently isn't indexed. Use for misspellings, regional synonyms, and long-tail variants.
• A+ content alt-text — indexed since 2023. The alt-text on each A+ image is a small but growing slot for keyword coverage.
• Brand name and product name fields — indexed for exact-match queries that include the brand.
• Product description — indexed in some categories, ignored in many. Treat as low-priority.
• Category and attribute fields — indexed for filter-style queries ("blue running shoe size 9"). The right category and complete attribute fields are essential for filter visibility.

What Amazon SEO is not

• Not Google SEO. No external backlinks, no domain authority, no schema markup. The index rebuilds inside Amazon's own catalogue, not from a web crawl.
• Not Sponsored Ads keyword research. Sponsored campaign data feeds Amazon SEO (the Search Term Report is one of the best synonym sources — see Episode 07), but choosing campaign keywords for paid placement and choosing organic keywords for the listing are different decisions with different evaluation criteria.
• Not a one-time exercise. Amazon's index re-evaluates listings continuously. A keyword set that ranked well 18 months ago will have drifted; re-harvesting synonyms every 6–12 months is the discipline that keeps rank stable.

The implication for the rest of Module 6

If findability is binary and relevance is weighted, then the work splits cleanly: harvest every legitimate synonym (Episodes 03–13), evaluate which ones deserve a slot (Episode 14), and structure them so the writing team can place each one in the highest-weight field it earns (Episodes 15–18). That's the entire module in one sentence.