Amazon SEO guide: how the A9/A10 algorithm ranks you

TL;DR: Amazon’s ranking system (commonly called A9, though Amazon has iterated it well past that name and sellers now shorthand the current version A10) ranks products on two axes: relevance (does your title, bullets, and backend keywords match the query) and performance (does your listing convert once it’s shown). Fix relevance first with exact-limit keyword placement, then fix performance with price, images, and reviews, because Amazon will stop showing a relevant listing that doesn’t sell.

---

Table of contents

1. What Amazon SEO actually means
2. How the ranking algorithm works: relevance and performance
3. Product title optimization
4. Backend search terms: the field nobody gets right
5. Bullet points and product description
6. Images and A+ content
7. Reviews and the trust signal Amazon weighs hardest
8. Sales velocity, conversion rate, and the PPC-to-organic flywheel
9. Amazon SEO vs Google SEO
10. The Amazon SEO optimization decision flow
11. A step-by-step Amazon SEO checklist
12. Measuring Amazon SEO: the metrics that matter
13. Common Amazon SEO mistakes
14. Frequently asked questions
15. Key takeaways

What Amazon SEO actually means

Amazon SEO is the practice of optimizing a product listing, title, bullet points, backend search terms, images, and price, so Amazon’s search engine surfaces it for relevant queries and shoppers buy once they see it. It is a different discipline from web SEO. Google ranks pages for informational, navigational, and transactional intent. Marc Schenker, writing for WordStream, points out that Amazon’s algorithm relies on far fewer signals than Google’s roughly 200 ranking factors because “pretty much every Amazon search is transactional” (WordStream). Every searcher on Amazon has their card ready. That single fact changes what you optimize for.

Amazon calls its ranking engine A9, the name of the Amazon subsidiary that built it. Jordi Ordóñez, writing for Helium 10, traces the origin: A9 was “built on top of Google’s code,” with Amazon licensing Google’s search index in the engine’s early years before building its own retrieval layer on top of it (Helium 10). Sellers now use “A10” as informal shorthand for the current, heavily-iterated version of that engine, since Amazon has never given the later releases an official new name. Whichever label you use, the mechanics below are what the current algorithm actually rewards.

Amazon is not a secondary channel to optimize for. Amazon accounted for 37.6% of the US ecommerce market in 2023, according to Statista data cited by Jungle Scout, more than Walmart, Apple, eBay, and Target combined (Jungle Scout). The same Jungle Scout data shows 56% of US consumers start their product search on Amazon rather than a search engine. If your product doesn’t rank inside Amazon’s own search box, you’re invisible to more than half the people looking to buy it that day. Amazon’s own seller data backs the scale: more than 75,000 independent sellers surpassed $1 million in sales in 2025, and over 11,000 US-based independent sellers grew sales more than 10x that year, according to Amazon’s official seller hub (Amazon).

If you’re weighing Amazon against a smaller marketplace for a handmade or niche product, the ranking mechanics diverge enough that the same playbook won’t transfer directly; see the Etsy SEO guide for how tag and title mechanics differ there.

How the ranking algorithm works: relevance and performance

Amazon’s algorithm scores every listing against a query on two separate axes, and you need both to rank. WordStream’s breakdown names them directly: relevance (does the listing match the words in the search) and performance (does the listing sell once Amazon shows it) (WordStream). Jungle Scout’s Brian Connolly lists the concrete factors Amazon has confirmed feed the relevance and performance score: “degree of text match, price, availability, selection, and sales history” (Jungle Scout).

Relevance is the gate. If your title, bullets, and backend keywords don’t contain the words a shopper typed, Amazon won’t retrieve your listing for that query at all, regardless of how well it converts once found. Performance is the ranking mechanism inside that gate. Among the listings that match, Amazon orders by which ones actually turn views into purchases, because a sale is Amazon’s own success metric, not yours.

Helium 10’s research adds a structural detail most sellers miss: A9 doesn’t weigh every field equally when it indexes your listing for keyword match. The field-weight hierarchy, from heaviest to lightest, runs title, then bullet points, then description, then backend keywords, then reviews (Helium 10). A keyword buried only in your description does less for indexing than the same keyword in your title. That hierarchy should decide where you spend your character budget when a keyword can only fit in one field.

Amazon’s query interpretation also isn’t a simple string match. Helium 10 describes A9 using probabilistic context-free grammar to parse a search query, identifying product type, likely user intent, and price sensitivity, while generally discounting specific brand names, gender terms, and color words unless the shopper explicitly searched for them (Helium 10). That’s why stuffing your title with every color variant you sell rarely helps: Amazon already knows to route “blue phone case” queries to the blue variant if your listing has a working variation structure, not because the word “blue” sits in your title 4 times.

Product title optimization

Your title is the single highest-weighted field in Amazon’s index, per Helium 10’s field hierarchy above, and it’s also the first thing a shopper reads in search results. Semrush’s Boris Mustapic lists Amazon’s formatting rules directly: titles should run up to 80 characters, use title case rather than all caps, and exclude promotional phrases, size, and color when those are captured by dedicated variation fields instead (Semrush).

Structure the title as: brand, then core product descriptor, then the 2 to 3 attributes shoppers actually search on (material, quantity, size, compatibility). Skip filler adjectives Amazon can’t index for intent, like “premium” or “best.” WordStream’s guidance matches this: title recommendations should include the brand name, a clear product description, and the specific attributes (ingredients, color, size, quantity) that separate your listing from a near-identical competitor, while avoiding keyword stuffing that breaks readability (WordStream).

A title that reads naturally to a human and hits your top 2 to 3 keywords in the first 60 characters, since some category templates truncate titles in mobile search results, beats a title crammed with every keyword variant you found in research. Amazon’s field-weight system already rewards title placement heavily; you don’t need to exploit it past the point of readability, because a title a shopper can’t parse in one glance hurts your click-through rate, which is itself a performance signal that feeds back into ranking.

Backend search terms: the field nobody gets right

Backend search terms are the hidden keyword field shoppers never see, and it’s the field with the most-violated rules in every Amazon SEO article reviewed for this guide. WordStream specifies the current limit precisely: 5 backend keyword fields, each capped at 50 characters, for roughly 250 characters of hidden keyword space total (WordStream).

The rules for what goes in that space, per WordStream’s breakdown, are strict and mechanical:

• Don’t repeat a word already used in your title or bullets; Amazon has already indexed it there.
• Skip quotation marks; they add characters without adding matching power.
• Commas are ignored by Amazon’s parser, so don’t waste characters on them, use spaces instead.
• Include spelling variations and true synonyms a shopper might type, not more brand or product-type repetition.
• Skip excessive word-form variations of the same root word (plural, past tense) since Amazon’s stemming already handles most of that.

If you sell the same product into a UK or EU Amazon marketplace, this is also where regional spelling variants belong. That’s a narrower version of the problem covered in international SEO and hreflang without breaking your site: different markets search in different words for the same product, and the fix here is a backend field, not a new page.

Never put a competitor’s brand name in backend keywords. It’s against Amazon’s terms of service and, per the rules above, Amazon’s parser is built to catch keyword abuse rather than reward it. The highest-value use of your 250 characters is the search terms real shoppers use that don’t fit naturally in a title or bullet: regional spelling (“colour” vs “color”), abbreviations, and use-case phrases you found in structured Amazon and Google keyword research or by mining your own search-term reports in Amazon Advertising.

Bullet points and product description

Bullet points are the second-heaviest field in Amazon’s index, per the Helium 10 hierarchy, and they’re also the section shoppers actually read before adding to cart. Semrush’s rules: sellers get up to 5 bullet points, vendors get up to 10, and the entire bullet section should stay under roughly 1,000 characters combined, about 200 characters per bullet (Semrush). Semrush also flags formatting Amazon strips or penalizes: hyphens, periods, special symbols, and exclamation points inside bullets.

Write each bullet around one benefit, backed by one concrete detail: a dimension, a material spec, a compatibility list, a use case. Put your highest-priority remaining keywords (the ones that didn’t fit in the title) in the first 2 bullets, since some category displays truncate the bullet list on mobile. The product description sits third in Amazon’s field-weight hierarchy, below bullets and above backend keywords, so it’s still worth targeting long-tail keyword variants there, but treat it as supporting evidence for a shopper who scrolled past the bullets, not your primary keyword field.

If you’re eligible for A+ Content or Premium A+ Content (Brand Registry required), that space replaces the plain-text description with modules, but it doesn’t add new indexed keyword text in most cases. Use it for conversion, comparison charts, lifestyle imagery, size guides, not as a keyword field.

Images and A+ content

Images don’t carry keyword weight in Amazon’s index, but they carry outsized weight in the performance half of the ranking equation, because they are the single biggest lever on click-through rate and conversion rate. Semrush’s technical requirements: up to 9 images total (7 typically visible on the listing), a minimum of 500x500 pixels with 1,600+ pixels on the longest side recommended, at least 72 dpi, the product filling a minimum of 85% of the frame, and a pure white background on the main image with the entire product visible outside its packaging (Semrush).

Marc Schenker at WordStream adds the mechanical reason image size matters beyond the visual: Amazon recommends images at 1,000 pixels or larger specifically because that resolution enables the pinch-to-zoom feature on product images, which “has been proven to enhance sales” (WordStream). A listing with a compliant but low-resolution main image is leaving a built-in conversion feature unused.

Order your secondary images by the objections a shopper has right before buying: scale/size reference, what’s included in the box, the product in use, a comparison chart against variants, and any certification or compatibility badges relevant to the category. Each image should answer one question a bullet point already raised, not repeat the same angle from a different distance.

Reviews and the trust signal Amazon weighs hardest

Reviews sit at the bottom of Amazon’s keyword field-weight hierarchy but at the top of its trust signal for performance ranking. WordStream cites the scale of that trust: 85% of customers say they trust online reviews as much as a personal recommendation (WordStream). Semrush’s figure is close: 41% of consumers say they always read product reviews before buying on Amazon (Semrush). Either number makes the same point: review volume and rating are not a vanity metric, they’re a conversion input Amazon’s algorithm can observe directly through your listing’s actual sell-through rate.

You cannot buy or incentivize reviews under Amazon’s terms of service, and doing so risks suspension faster than almost any other listing violation. The compliant levers are Amazon’s built-in Request a Review button (automated, one tap, inside Seller Central), the Vine program for new listings with fewer than 30 reviews, and simply shipping a product that matches its listing description closely enough that organic reviews stay positive. A 4.6-star listing with 40 reviews will consistently outrank a 4.1-star listing with 400 reviews for the same query once Amazon’s performance score kicks in, because rating quality is closer to a real trust proxy than raw count. This is the same trust-signal logic behind E-E-A-T for small sites with no brand: a smaller, cleaner trust footprint beats a larger, noisier one on almost every platform that scores trust at all.

Sales velocity, conversion rate, and the PPC-to-organic flywheel

Sales velocity is the performance metric that ties everything above together, and Jungle Scout gives the exact formula: (impressions × price × conversion rate) / number of days, illustrated with a worked example of 3,000 impressions, a $20 price point, and a 10% conversion rate producing roughly $200 per day in velocity (Jungle Scout). Amazon tracks this per-ASIN over a trailing window, and a rising velocity trend is one of the clearest performance signals feeding into where your listing ranks tomorrow.

Conversion rate itself is measurable directly in Seller Central. WordStream points sellers to Unit Session Percentage on the Detail Sales Page, the ratio of units bought to sessions, as the practical dashboard number to watch week over week (WordStream). Jungle Scout frames conversion rate the same way structurally, total order items divided by sessions, and notes Amazon favors higher-converting listings because a high conversion rate signals relevance the algorithm can trust (Jungle Scout).

This is why PPC and organic ranking aren’t separate workstreams on Amazon the way paid and organic search are more separable on Google. A well-targeted Sponsored Products campaign buys impressions on exactly the queries you want to rank for organically. If your listing converts well on those impressions, the resulting sales velocity and Unit Session Percentage feed the organic ranking algorithm directly. Sellers run this deliberately as a flywheel: launch with PPC on your priority keywords, let real conversion data accumulate for 2 to 4 weeks, then watch organic rank climb on the same terms as ad spend can be pulled back. It’s the same logic behind tracking content ROI with a defined formula rather than judging spend by impressions alone: the number that matters is what converts, not what merely gets shown.

Amazon SEO vs Google SEO

Amazon and Google SEO share vocabulary (keywords, ranking, indexing) but reward almost opposite behaviors in several places. Sellers who bring Google-SEO instincts to Amazon usually over-invest in the wrong field.

Factor	Amazon SEO	Google SEO
Primary intent	✓ Nearly 100% transactional (WordStream)	✗ Mixed: informational, navigational, transactional
Ranking signal count	✓ A small, largely confirmed set (relevance + performance)	✗ Roughly 200 known and unknown signals
Backlinks	✗ Not a direct ranking factor	✓ Core ranking factor
Content length	✗ Character limits are hard caps (80-char title, ~250-char backend)	✓ No hard cap, longer often ranks for competitive terms
Reviews/UGC	✓ Directly observable performance input	✗ Indirect, via engagement and E-E-A-T signals
Off-platform traffic	✓ Rewarded (external traffic is part of the modern algorithm’s signal set)	✓ Rewarded (backlinks, brand search)
Paid-organic interaction	✓ Tight: PPC conversions feed organic rank via sales velocity	✗ Loose: Google Ads spend does not affect organic rank
Update cadence visibility	✗ Opaque, no public update log	✗ Opaque, though Google publishes named core updates

The row worth internalizing is backlinks. There is no Amazon-native equivalent to a backlink profile, and unlike the shift toward brand mentions as the new backlinks for AI search on the open web, Amazon doesn’t credit unlinked brand mentions either. Every ranking lever instead runs through the listing fields themselves and your sell-through performance, which is why an Amazon SEO project is closer to structured data work than to link-building. If you’re used to running an ecommerce category page strategy on your own site with faceted navigation and internal links, none of that transfers; Amazon’s browse-node placement is Amazon’s decision, driven by your chosen category and item-type keywords, not your architecture.

The Amazon SEO optimization decision flow

Use this to decide where to spend the next hour of listing work, in order.

flowchart TD
    A[Is the listing ranking for your target query at all?] -->|No| B[Fix relevance first]
    A -->|Yes, but rank is low| C[Fix performance]
    B --> B1[Add missing keyword to title if room exists]
    B1 --> B2[Move overflow keyword to bullets]
    B2 --> B3[Fill remaining backend search term space]
    B3 --> D[Re-check indexing in 24-48 hours]
    C --> C1{Is click-through rate low?}
    C1 -->|Yes| C2[Fix main image and title clarity]
    C1 -->|No| C3{Is conversion rate low?}
    C3 -->|Yes| C4[Fix price, reviews, or A+ content]
    C3 -->|No| C5[Increase sales velocity with targeted PPC]
    C5 --> E[Monitor Unit Session Percentage weekly]
    C4 --> E
    C2 --> E

If a listing isn’t indexed for a query at all, relevance is broken and no amount of ad spend fixes it, because Amazon can’t rank what it can’t match. If it’s indexed but ranks low, the problem has moved to performance, and the fix depends on which specific metric is weak: a low click-through rate points to the image and title, a low conversion rate points to price and reviews, and a healthy conversion rate with low overall volume points to sales velocity, which PPC can accelerate directly.

A step-by-step Amazon SEO checklist

1. Run keyword research inside Amazon (Brand Analytics search-term reports if you have Brand Registry, or the autocomplete/related-search data on the live site) rather than importing Google keyword volume wholesale; Amazon query patterns diverge from web search patterns the same way community-driven keyword signals diverge from search-console data on the web.
2. Draft a title under 80 characters: brand, product type, then your top 2-3 differentiating attributes, in title case, no promotional language.
3. Write 5 bullets (10 for vendors), roughly 200 characters each, one benefit and one concrete spec per bullet, no hyphens/periods/exclamation points, priority keywords in the first two bullets.
4. Fill all 5 backend search term fields (50 characters each, ~250 total) with synonyms, spelling variants, and use-case phrases not already used in title or bullets. No competitor brand names.
5. Shoot or source a main image on pure white background, product filling at least 85% of frame, minimum 1,000px on the longest side.
6. Build out the remaining 6-8 secondary images to answer scale, contents, in-use, and comparison questions in that order.
7. Enable A+ Content if Brand Registry-eligible; use it for conversion detail, not keyword stuffing.
8. Set up Amazon’s automated Request a Review flow and, for new listings, evaluate Vine eligibility.
9. Launch a tightly-scoped Sponsored Products campaign against your priority keywords to seed sales velocity.
10. Check Unit Session Percentage and sales velocity weekly for the first 4-6 weeks and reallocate ad spend toward the keywords converting best.
11. Re-run steps 2-4 any time you add a genuinely new attribute (a new size, a new certification) rather than leaving the listing static once it ranks; the same refresh-over-new-content logic that applies to web pages applies to a listing that’s decaying in rank.
12. Once the checklist above is running smoothly on your top listings, decide what to hand off to rules and automation versus what still needs a human judgment call; see what to automate and what will burn you for the general framework.

Measuring Amazon SEO: the metrics that matter

Track these inside Seller Central rather than relying on external rank trackers alone, since Amazon’s own performance data is what the algorithm itself is reading.

• Unit Session Percentage (conversion rate): the ratio Amazon itself uses as a performance signal, per WordStream’s Detail Sales Page guidance (WordStream).
• Sales velocity: impressions × price × conversion rate, divided by days, calculated per the Jungle Scout formula above (Jungle Scout).
• Organic rank for priority keywords: track weekly for your top 10-15 target terms, separate from paid placement.
• Search term impression share from Brand Analytics, to see whether you’re being retrieved for a query at all before worrying about position.
• Review count and rating trend, watched for sudden drops that flag a product or fulfillment issue before it drags conversion down.

This mirrors the shift happening in the 9 SEO KPIs that still matter when clicks are disappearing on the web SEO side: vanity traffic numbers matter less than the metrics tied directly to what the algorithm is optimizing for. On Amazon, that’s conversion and velocity, not impressions. If you’re running the web side of your brand alongside Amazon, the same discipline of GA4 reporting that answers real questions instead of vanity sessions applies to whichever storefront or listing page sends Amazon-bound traffic.

Common Amazon SEO mistakes

Keyword stuffing the title past readability. Once your top attributes are in, extra keywords crammed into the title hurt click-through rate more than they help match rate, since Helium 10’s field-weight data shows the title is already the heaviest-indexed field; you don’t need to exploit it further.

Repeating keywords across title, bullets, and backend fields. WordStream’s backend rule is explicit: don’t repeat a word already used elsewhere in the listing, since Amazon has already indexed it. Repetition wastes your scarce 250 backend characters.

Ignoring the description and A+ content as keyword-dead zones. They’re lower in the field hierarchy than title and bullets, per Helium 10, but not zero-weight; leaving them generic wastes indexable text.

Chasing reviews with incentives. It’s a terms-of-service violation and risks account suspension; use Amazon’s own Request a Review automation and the Vine program instead.

Treating PPC and organic as separate budgets. Given how tightly sales velocity from ad-driven conversions feeds organic rank, cutting PPC the moment a listing starts ranking organically often stalls the velocity that got it there.

Copying a Google-SEO content strategy onto Amazon. There’s no backlink equivalent on Amazon, and content length caps are hard limits, not soft guidelines, so tactics built for programmatic SEO or long-form web content don’t transfer directly.

Letting stale listings sit unreviewed. A listing that hasn’t been touched since launch accumulates outdated pricing context, missing keywords from newer search behavior, and images that no longer match current buyer expectations; the same content pruning discipline that clears dead weight from a web site applies to a catalog full of listings nobody has revisited in a year.

Ignoring where AI shopping assistants are heading. Amazon’s Rufus assistant and similar AI shopping tools read the same listing fields the ranking algorithm does, so a listing optimized for human scanning is already most of the way toward being optimized for AI agents doing the shopping on a buyer’s behalf; don’t treat this as a separate future project.

Frequently asked questions

What is Amazon’s A9 algorithm?

A9 is the name of Amazon’s product search ranking engine, built by the Amazon subsidiary of the same name. Jordi Ordóñez at Helium 10 traces its origin to Amazon licensing Google’s search index before building its own retrieval layer on top of it. It ranks listings on relevance (keyword match) and performance (conversion and sales history) (Helium 10).

Is A10 a real, separate algorithm from A9?

Amazon has never officially named a successor to A9. “A10” is seller shorthand for the current, iterated state of the same engine, reflecting years of updates that shifted more weight toward performance signals like conversion rate and off-platform traffic. Treat A9 and A10 as the same system at different points in time, not two competing engines.

How long can an Amazon product title be?

Semrush’s guidance puts the practical limit at up to 80 characters, formatted in title case, excluding promotional phrases and any size or color already captured in dedicated variation fields (Semrush). Category-specific limits can vary, so check your specific category’s style guide in Seller Central before finalizing.

How many backend search term characters does Amazon allow?

Five backend keyword fields at 50 characters each, roughly 250 characters total, per WordStream’s breakdown of current field limits (WordStream). Commas are ignored by the parser, so separate terms with spaces, and never repeat a word already used in your title or bullets.

Can I put a competitor’s brand name in my backend keywords?

No. It violates Amazon’s terms of service and can trigger a listing suspension. Amazon’s search parser is built to detect this kind of keyword abuse rather than reward it, so the risk outweighs any short-term matching benefit.

Does Amazon rank listings based on backlinks?

No. Unlike Google, Amazon has no backlink-based ranking signal. Off-platform traffic can help, as external visits are part of the modern algorithm’s broader signal set, but the mechanism is different from a link authority model; it’s closer to a conversion and demand signal than to a citation graph.

What is sales velocity and why does it matter for ranking?

Sales velocity is impressions multiplied by price multiplied by conversion rate, divided by the number of days measured. Jungle Scout’s worked example: 3,000 impressions, a $20 price, and a 10% conversion rate produce roughly $200 per day in velocity (Jungle Scout). Rising velocity is one of the clearest performance signals feeding Amazon’s ranking decision.

Does running Amazon PPC ads help organic ranking?

Indirectly, yes. Sponsored Products campaigns targeted at your priority keywords generate conversions on those exact terms. If the listing converts well, the resulting sales velocity and Unit Session Percentage feed directly into the same performance score used for organic ranking, which is why sellers treat PPC and organic as one connected system rather than separate budgets.

How many product images can I upload on Amazon?

Up to 9 total, with 7 typically visible directly on the listing page. Semrush specifies a minimum size of 500x500 pixels, with 1,600 or more pixels on the longest side recommended, at least 72 dpi resolution, and the product filling a minimum of 85% of the image frame (Semrush).

Why does Amazon recommend 1,000-pixel images specifically?

That resolution enables the pinch-to-zoom feature on the product image, which WordStream cites as a feature “proven to enhance sales” (WordStream). A compliant but lower-resolution main image forfeits that built-in conversion lever.

How much do reviews actually affect Amazon ranking?

Heavily, through the performance side of the algorithm. WordStream cites that 85% of customers trust online reviews as much as a personal recommendation, and Semrush’s figure shows 41% of shoppers always read reviews before buying on Amazon specifically. Rating quality and recency function as a directly observable trust signal Amazon’s ranking system can weigh.

Can I pay for or incentivize Amazon reviews?

No, this is against Amazon’s terms of service and risks account suspension. Use Amazon’s built-in Request a Review automation inside Seller Central, and evaluate Vine program eligibility for new listings with fewer than 30 reviews, as the compliant paths to more reviews.

What is Unit Session Percentage and where do I find it?

It’s Amazon’s name for conversion rate on a listing: units ordered divided by sessions. WordStream points sellers to the Detail Sales Page inside Seller Central as the practical place to track it weekly, since it’s a direct performance input to ranking, not just a reporting metric.

How many bullet points does Amazon allow?

Up to 5 for third-party sellers and up to 10 for vendors, per Semrush’s listing rules, with the entire bullet section recommended to stay under roughly 1,000 characters combined, about 200 characters per bullet (Semrush).

What formatting is not allowed in Amazon bullet points?

Semrush’s guidance flags hyphens, periods, special symbols, and exclamation points as formatting Amazon strips or penalizes inside bullet point fields. Write clean, symbol-free sentences instead, each centered on one benefit backed by one concrete spec.

Does A+ Content help Amazon SEO directly?

A+ Content, available to Brand Registry-enrolled sellers, mainly affects the performance side of ranking by improving conversion rate through richer visual and comparison modules, rather than adding significant new indexed keyword text. Use it to answer buying objections visually, not as a keyword-stuffing opportunity.

What percentage of US ecommerce sales happen on Amazon?

Amazon accounted for 37.6% of the US ecommerce market in 2023, per Statista data cited by Jungle Scout, well ahead of Walmart at 6.4% and every other named competitor in that dataset (Jungle Scout).

What share of shoppers start their product search on Amazon instead of Google?

56% of US consumers start their product search on Amazon rather than a traditional search engine, according to Jungle Scout’s Consumer Trends Report data (Jungle Scout).

Is Amazon SEO different for third-party sellers vs vendors?

Mostly at the field-limit level: vendors get up to 10 bullet points versus 5 for third-party sellers, per Semrush, and vendor accounts have access to different content tools (like Premium A+ Content) than standard Seller Central accounts. The core relevance-and-performance ranking logic applies to both account types the same way.

How often does Amazon update its ranking algorithm?

Amazon does not publish a public update log the way Google names core updates. Sellers infer changes from shifts in what correlates with rank, most notably an increasing weight toward external traffic and conversion metrics over the raw keyword-matching that dominated the original A9 model. Treat the algorithm as continuously iterating rather than updated on a fixed schedule.

Key takeaways

• Amazon’s ranking system scores every listing on relevance (keyword match in title, bullets, backend terms) and performance (conversion rate, sales velocity, reviews), and both have to clear the bar to rank.
• Title (80 characters), bullets (5 for sellers, 10 for vendors, ~1,000 characters total), and backend search terms (5 fields, 50 characters each, ~250 total) are hard field limits, not style suggestions.
• Field weight for keyword indexing runs title, then bullets, then description, then backend terms, then reviews, so put your highest-priority keyword in the heaviest field it fits.
• Sales velocity (impressions × price × conversion rate / days) and Unit Session Percentage are the performance metrics Amazon itself tracks; watch those weekly, not just search rank.
• PPC and organic rank are connected through sales velocity, not separate systems, so pulling ad spend the moment organic rank improves often stalls the momentum that produced it.
• There is no backlink equivalent on Amazon; every lever runs through the listing fields and your own conversion performance.

Run the checklist in this guide against one listing this week: fix the title first, then backend terms, then check whether your images hit the 1,000-pixel zoom threshold. Track Unit Session Percentage for 30 days before judging whether the changes worked, since Amazon needs real sales data to re-score the listing.