Back to Resources

How Amazon ranks and converts product listings in 2026, and how to optimize every part of the detail page for both.

If your Amazon rankings have stalled, more keywords probably won’t fix them.

We’ve watched the same pattern across hundreds of accounts. A listing ranks for plenty of terms and impressions look fine, but click-through rate is mediocre and conversion rate is worse. The brand decides the answer is broader keyword coverage, so it stuffs the title, packs the backend, and adds bullets nobody reads. Three months later, ranking is flat or down.

Amazon listing optimization and Amazon SEO are the same job. Relevance gets your listing into the results. Performance (clicks, conversions, sales) decides where it stays. And in 2026 the ground moved under both: Amazon capped product titles at 75 characters on July 27, 2026, replaced Rufus with Alexa for Shopping on May 13, 2026, and moved customer Q&A off the main detail page. A listing built on the old playbook can lose rank while every field still looks “complete.”

This guide covers how Amazon’s search works now, how to optimize each element of a listing for ranking and conversion at the same time, and where AI shopping is changing the rules.

What Do Amazon Listing Optimization and SEO Cover?

Amazon SEO is the work of earning organic visibility in Amazon’s search results. Amazon listing optimization is the work of improving the product detail page so it ranks for the right searches and converts the shoppers who land on it. Same page, same shopper, same algorithm, which is why we treat them as one practice.

The elements you control:

• Visible content: title, Item Highlights, bullet points, description or A+ Content, Brand Story, images, and video
• Hidden content: backend search terms, category, and product attributes
• Offer signals: price, reviews, in-stock status, and fulfillment method

Each element feeds two kinds of signals in different proportions. Relevance signals tell Amazon whether your product matches a search. Performance signals tell Amazon whether shoppers who see it click and buy. A listing optimized only for keyword coverage indexes broadly and converts poorly. A listing written only as persuasion ranks for too few terms to generate the traffic that would prove the copy works. Both failures show up in the data, and neither corrects itself.

How Does Amazon’s Search Algorithm Rank Products in 2026?

Amazon ranks products by how relevant they are to a search and how likely they are to convert, then watches what shoppers do to confirm or correct that ranking. Keywords get your listing considered. Behavior gets it ranked.

The algorithmic model, often called A9, matched keywords and rewarded sales velocity. That foundation still exists: the words in your listing still determine which searches you’re eligible for. What changed is the layer on top.

COSMO and intent-based matching

COSMO is Amazon’s common-sense knowledge system, which Amazon’s researchers described in a paper published at SIGMOD 2024. It mines shopper behavior to build knowledge graphs of how products, uses, and needs relate to each other. In practice, Amazon’s search now interprets the intent behind a query as well as its words.

A search for “shoes for a beach wedding” doesn’t only match titles containing those words. The system infers that the shopper wants dressy footwear that can handle sand and heat, and ranks accordingly. For sellers, keyword density matters less than whether your listing clearly states what the product does, who it’s for, and how people use it. We break down the current ranking signals in our guide to the Amazon algorithm in 2026.

Relevance signals and performance signals

Relevance signals come from listing content: keywords in the title, Item Highlights, bullets, and backend fields, plus category placement and product attributes.

Performance signals come from shopper behavior: click-through rate, conversion rate, sales velocity, price competitiveness, review count and rating, and in-stock status.

For established products, performance carries more weight. A perfectly keyworded listing that doesn’t sell will rank below a less keyword-rich listing that converts well. That’s why “add more keywords” so rarely fixes a stalled product. We’ve audited accounts where a seller spent six months adding keywords to a listing converting at about a third of its category’s rate. Three more keywords don’t close a gap like that. The main image does, or the price, or the bullets.

Alexa for Shopping changes how shoppers search

On May 13, 2026, Amazon replaced Rufus with Alexa for Shopping, an assistant built into the Amazon app, the website, and Echo Show devices. Shoppers ask it questions in the search bar, and it compares features, prices, and reviews across Amazon and the web before it recommends anything.

That changes the query your listing has to answer. A shopper who asks “which water bottle keeps ice overnight on a camping trip” isn’t typing your keywords. If your listing doesn’t state the answer plainly, the assistant has nothing to recommend you for. Our Alexa for Shopping guide for sellers covers what changed and what to do about it.

Keyword Research That Drives Sales

Good Amazon keyword research finds the terms shoppers use when they’re ready to buy what you sell, then assigns each one to the listing field where it does the most work. The target is purchase intent, and search volume is a poor proxy for it.

Intent beats volume

“Coffee maker” is about as broad as intent gets. The shopper could want a $20 drip machine or a $1,000 espresso setup. “Stainless steel pour-over coffee maker for one person” has a fraction of the volume and a far higher purchase rate for the listing that matches it. The strongest Amazon SEO strategies don’t chase everything. They find the query clusters where the product genuinely fits, invest in ranking for those, and use PPC data to confirm which terms convert before committing organic space to them.

Where the real keyword data lives

Your Sponsored Products search term report is the highest-signal source you have. It shows the queries real shoppers typed before they bought (or didn’t) your specific product, in your category, against your competitors. No third-party tool matches that precision.

If you’re brand registered, Brand Analytics adds Search Query Performance data: impressions, clicks, add-to-carts, and purchases for each query, with your brand’s share of each. It shows where you’re visible but losing the sale, and where competitors take purchases you’re missing. We cover the reports worth pulling in our Amazon Brand Analytics guide.

Reverse-ASIN research on the listings outranking you fills in the rest. Cross-reference their ranking terms against your own conversion data and you’ll find gaps where the product fits but the listing doesn’t say so. Our 10-step guide to Amazon keyword research walks through the full process.

Give every keyword a home

Tier your terms by intent, then map each tier to a field:

• Primary keywords describe the product directly and carry the highest purchase intent. They belong in the title.
• Secondary keywords add the attributes shoppers filter by (material, size, use case, audience). They belong in Item Highlights and the first bullets.
• Long-tail keywords capture specific, conversational searches. They fit naturally in later bullets and backend search terms, and they’re the phrasing an AI assistant is most likely to match against.

Our breakdown of keyword research that drives sales shows how to build that list around intent instead of volume.

Titles and Item Highlights Under the 75-Character Rule

Since July 27, 2026, Amazon caps product titles at 75 characters, including spaces, in every category except media. Amazon’s stated reason is that 75 characters displays in full on mobile. A separate Item Highlights field of up to 125 characters now carries what used to crowd the back half of the title, and Amazon confirms that field is searchable and shown with the title in search results and on the detail page.

The title still carries more weight in Amazon’s relevance evaluation than any other field, and it’s the first text a shopper reads after the main image. It now has 75 characters to do both jobs.

Split the work between two fields

The title wins the click. Lead with your brand, then the core product noun, then the one or two attributes a shopper needs to recognize the product as theirs (size, count, or the key feature). Put your primary keyword as early as that structure allows.

Item Highlights carries everything that helps you get found and chosen but doesn’t need to be in the first glance: materials, use cases, compatibility, and secondary keywords. Write it as a short, readable phrase set rather than a keyword string, because shoppers see it too. Our breakdown of Amazon’s 75-character title limit covers the rule and edge cases in detail.

What Amazon will rewrite for you

Promotional language (“best seller,” “free shipping,” “sale”), repeated words, special characters, and all-caps already broke Amazon’s title style rules before July 2026. The July change added enforcement. Amazon’s Seller Central announcement says titles that don’t comply will be “updated to the AI recommendation gradually.” An automated rewrite rarely keeps the terms you’d have fought for, so rewrite over-length titles yourself: decide what stays in the title, move the rest into Item Highlights, and change one element at a time so you can see what moved ranking.

Write for shoppers first. We see it every time we test a stuffed title against the same listing rewritten cleanly: indexing barely changes, and click-through rate moves.

Bullet Points Are Where Persuasion Happens

Bullet points are the main persuasion surface on the detail page. Shoppers who clicked through from search read them to decide whether the product fits their need, so each bullet has to earn its place for both the shopper and the algorithm.

Benefit first, feature as proof

Lead each bullet with the outcome the shopper cares about, then back it with the feature that delivers it. “Premium stainless steel construction” makes the shopper work out the benefit, and most won’t. “Keeps drinks cold for 24 hours, with double-wall stainless steel that stops condensation through a full day outdoors” states the outcome first and proves it second. That structure persuades, and it gives Amazon natural-language coverage for the words shoppers use when they search for that outcome.

Order bullets by what decides the sale

On mobile, shoppers see the top of your bullet list before they have to tap to expand it. Put your three decision-makers first: your strongest benefit, the answer to the most common objection, and your clearest differentiator. Save later bullets for secondary benefits and additional keyword coverage. A shopper who decides against you after three bullets never reads the fourth.

Product Description, A+ Content, and Brand Story

For sellers without Brand Registry, the product description is the only long-form copy field. Write it for the shopper who is almost convinced and needs one more answer: compatibility, sizing, care, or a use case the bullets didn’t cover.

For brand-registered sellers, A+ Content replaces the plain-text description with image-and-text modules, comparison charts, and richer layouts. Brand Registry requires a pending or registered trademark from a government trademark office in a country with an Amazon store. It’s worth the effort: Amazon states that Basic A+ Content can increase sales by up to 8% and Premium A+ by up to 20%, citing its own internal data without a published methodology. Premium A+ is also free, and in 2026 it’s broadly available to brand-registered sellers, though Amazon adjusts the qualifying criteria, so check A+ Content Manager for your account.

The lift isn’t automatic. A+ that restates the bullets in a prettier format adds little. The modules that earn their space in our experience:

• Comparison charts that answer “which one should I buy” across your own catalog
• Feature-with-proof modules that pair a specific claim with an image that demonstrates it
• Lifestyle modules that show the product in real use

Publish a Brand Story as well. It sits above your A+ modules, carries across every ASIN under your brand, and cross-sells your catalog from inside a single product page. Our guide to A+ Content in 2026 covers module selection and what separates A+ that converts from A+ that fills space.

Backend Search Terms

Backend search terms are the hidden keyword field in Seller Central. Shoppers never see it, and Amazon limits it to less than 250 bytes. Bytes, not characters, which matters once you use accented letters or symbols.

Use it for terms that don’t fit naturally in visible copy: synonyms, alternate spellings, common misspellings, and adjacent terms. Don’t repeat words already in your title or Item Highlights, since that spends bytes on terms already indexed.

Review backend terms at least quarterly and before peak season. Search behavior shifts, seasonal queries come and go, and a backend field untouched for two years almost certainly describes how shoppers searched two years ago. Our guide to Amazon backend keywords covers field structure and the mistakes we see most often.

Listing Images and Video: Where the Click and the Sale Are Won

Images move click-through and conversion more than any other part of the listing. The main image wins the click in search results. Secondary images and video close the sale on the detail page. Both feed ranking through the performance signals they create.

The main image has one job

Earn the click. Amazon requires the main image on a pure white background (RGB 255, 255, 255) with the product filling 85% to 100% of the frame, and at least 1,000 pixels on the longest side to enable zoom. Aim for 1,600 pixels or more.

Compliance only gets you in the door. Pull up your listing in mobile search results and ask whether it stands out from the listings around it and communicates what the product is at a glance. If it doesn’t, it’s capping your click-through rate, and your click-through rate is capping your ranking. We’ve seen brands spend heavily on lifestyle photography and A+ Content while the main image loses the click before a shopper sees any of it.

Give every secondary image a job

Treat secondary images as a planned sequence:

• Feature callout: the one or two features that justify the price
• Scale or size reference: so the shopper knows exactly what arrives
• Lifestyle: the product in realistic use, which often drives the most lift in home, apparel, fitness, and personal care
• Objection infographic: the one question that kills the sale, answered visually
• What’s in the box: closes the gap between expectation and delivery and cuts returns

Comparison images work when your advantage can be shown visually. Be careful about depicting named competitor products, which Amazon’s content policies restrict. Check that each image reads on a phone screen, where most shoppers see it; our guide to optimizing Amazon listings for mobile covers what changes on the small screen.

Add video

Video appears in the image gallery and gives shoppers the closest thing to handling the product. Three types earn their place: a highlight video for standout features, a lifestyle video that shows the product in use, and an unboxing video that previews exactly what arrives. Video uploads come with Brand Registry, and video pays off most for complex or unfamiliar products, where hesitation comes from not knowing how the thing works.

Reviews Shape Both Ranking and Conversion

Review count, rating, and recency affect ranking, mostly through conversion rate, which is the performance signal that matters most. In our experience, a listing with 200 reviews at 4.7 stars usually outperforms one with 20 reviews at 4.9 in a competitive category, because the larger review set removes more doubt.

Amazon gives you two compliant ways to build review velocity. Amazon Vine sends products to Amazon’s reviewer network; to enroll, a product needs fewer than 30 published reviews and an active FBA offer, and the seller needs a role on the brand in Brand Registry. The Request a Review button in Seller Central is available for eligible orders. Anything that offers compensation for reviews, asks specifically for positive ones, or routes buyers around Amazon’s process violates policy and puts the account at risk.

Reviews also feed the AI layer. Alexa for Shopping compares reviews when it recommends products, so the language shoppers use in reviews becomes part of how your product is described back to the next shopper.

Testing With Manage Your Experiments

Manage Your Experiments is Amazon’s native A/B testing tool. It lets brand-registered sellers with a Professional account test product images, titles, bullet points, descriptions, and A+ Content (including Brand Story) against each other and measure which version sells more.

What it takes

The ASIN needs enough recent traffic to produce valid results, which rules out many low-volume products. Amazon recommends running experiments for 8 to 10 weeks to reach statistical significance, and says pre-selected settings can produce results in as little as four weeks. Ending a test early gives you a guess dressed up as data.

What to test first

Test the elements that decide the click and the purchase: main image, title, and A+ comparison modules. Minor wording changes in a single bullet rarely produce a difference large enough to clear significance. Run one experiment per ASIN at a time, since two simultaneous changes make the result impossible to attribute. And write down what you learn. Knowing that a benefit-led title beats a feature-led title on one product gives you a starting point for every listing in that category.

The Conversion Rate Diagnostic

Conversion rate on Amazon is units ordered divided by sessions. Amazon reports it as Unit Session Percentage in Business Reports, under Detail Page Sales and Traffic by ASIN. Measure it against your competitive set, the listings ranking around you for your main keywords. Industry averages hide too much variation to be useful.

When conversion falls below what your category supports, diagnose in this order:

1. Main image. The most common culprit and usually the biggest fix.
2. Price. Strong listings still underperform when priced well above comparable options without a visible reason.
3. Reviews. If your count or rating sits well below the listings around you, that gap is working against every other fix.
4. Bullets. Do the first three answer the shopper’s most pressing questions, benefit first?
5. A+ Content. For brand-registered sellers, weak A+ is worth fixing once the simpler levers are in place.
6. Everything else. Wrong category, suppressed attributes, and mobile display problems are real issues, but rarely the place to start.

The temptation when conversion drops is to adjust bids and pause campaigns. Fix the listing first. A listing that converts poorly makes every click more expensive than bid strategy can fix, and the weak performance signals it sends hold down organic rank at the same time. For a structured audit, start with these seven mistakes that quietly kill Amazon rankings.

The PPC and Organic Flywheel

On Amazon, PPC and organic ranking feed each other. Advertising buys the sales velocity and conversion data that improve organic rank, and better organic rank reduces how much advertising you need.

Search terms that convert in paid campaigns build relevance for those terms organically. Velocity from paid traffic strengthens the performance signals the algorithm uses to place you. TACoS (total advertising cost of sale) makes the flywheel visible: as organic sales grow, TACoS falls even with flat ad spend. A falling TACoS over time is the clearest sign the organic work is paying off. We explain how to read both metrics together in our guide to ACoS and TACoS.

The practical implication: launch-phase PPC buys ranking data along with sales. Brand registration also opens formats like Sponsored Brands that support this flywheel; our comparison of Sponsored Products vs Sponsored Brands covers when each earns its budget. Our guide to integrating Amazon SEO and PPC covers how to run the two as one program.

Where AI Shopping Changes the Work

Alexa for Shopping changes who reads your listing. Every shopper who asks the assistant a question instead of typing keywords gets an answer assembled from product information, and your listing is only part of what it weighs.

Three shifts matter for sellers:

• Answers beat keywords. Listings that state use cases, fit, and specifics in plain language give the assistant something to match. Vague benefit copy gives it nothing.
• Q&A moved off the main page. Amazon moved customer Q&A off the main detail page and replaced the browsable block with a search field. Amazon hasn’t published seller guidance on the change. Our read, covered in Amazon Q&A optimization in the age of generative search, is that the answers still matter as input even though fewer shoppers see them.
• Reviews carry more weight in the answer. The assistant compares reviews across products, so review language shapes how your product gets described.

Traditional Amazon SEO is the foundation the AI layer builds on. The listings that perform in AI shopping are the ones with clear titles, complete attributes, specific bullets, and well-built A+ Content. Our generative search optimization guide for Amazon and our voice search optimization guide go deeper on the mechanics.

What Consistent Optimization Produces

Disciplined listing work compounds, and the gains change character over time.

In the first months, gains come from structural fixes. Titles get rebuilt for the 75-character rule, bullets get rewritten benefit-first, main images get tested and replaced, backend terms get cleaned of redundancy, and A+ Content gets rebuilt around the questions shoppers ask. Our competitor analysis guide shows how to judge those fixes against the specific listings you’re competing with rather than a generic checklist.

After that, conversion data accumulates. The listing ranks for more terms, which brings more impressions, sales, reviews, and conversion signal. The business depends less on paid traffic, TACoS falls, and margin rises because a growing share of sales carries no ad cost.

Two of our partners show what this looks like. Snow Cosmetics saw a 27% increase in conversion rate and a 43% increase in sales in three months after we rebuilt the listings for their refill wands with keyword research, new copy, product photography, and custom A+ Content. Zugu Case grew sales 161% and conversion rate 23% in three months by building listing content around a premium position in a category crowded with cheap copies.

How Canopy Approaches Listing Optimization and SEO

We run listing optimization, SEO, and advertising as one program, because they share the same data and the same shopper. Every partner has a dedicated brand manager who owns the account for the life of the engagement, so listing decisions get made with full sight of PPC performance, organic ranking trends, and the conversion signals that connect them. That approach is a big part of why Canopy’s partners see an average 84% year-over-year profit increase.

Our listing optimization services and product photography and video teams build every listing from competitive data and real keyword performance.

Frequently Asked Questions