Amazon’s 75-Character Title Rule: What Happened After July 27

The deadline has passed. We’re seeing rankings drop for brands that let Amazon’s AI rewrite their listings. Here’s what happened and how to recover.

What changed on July 27

On July 27, 2026, Amazon began enforcing a new format for product listings. Titles in all categories except media must be 75 characters or less, including spaces. That is down from the 200 to 250 characters most brands used before the deadline.

Amazon also introduced a new field called Item Highlights, which provides 125 additional characters for materials, recommended use cases, and comparison details. Item Highlights are searchable and appear alongside titles in search results and on product detail pages.

The enforcement mechanism is aggressive. Amazon is using AI to generate compliant title and Item Highlights recommendations for every listing over 75 characters. Brand owners have a 14-day window to review, modify, or approve those recommendations before they go live automatically. Listings still over the limit after July 27 get updated to Amazon’s AI-generated recommendation with no further action required from the seller.

What we’re seeing from brands that did nothing

The mechanics of the drop are predictable in hindsight. Amazon’s AI has no visibility into which of your keywords actually drive conversions. It optimizes for what looks like a clean, compliant title from Amazon’s perspective, not for what preserves your discoverability. Long-tail modifiers, use-case language, and category-defining terms that took years of testing to identify get replaced with generic descriptors. The listing becomes technically compliant and functionally invisible.

The compounding damage is worse than the initial ranking drop. When titles no longer align with the terms brands bid on, ad relevance drops. Return on ad spend declines. Sponsored placements that were profitable become marginal. The entire performance stack degrades from the top down.

Why the title was your highest-stakes field

Three things happened when brands lost control of their titles.

Discoverability collapsed. The title is the single strongest relevance signal for Amazon search, for both human shoppers and AI agents. A generic AI rewrite stripped the specific terms these listings ranked for. Rankings built over years evaporated in a single automated update.

Advertising performance dropped. Ad relevance and return on ad spend depend on tight alignment between bid keywords and listing content. When the title no longer matched the terms these brands bid on, every campaign behind that listing lost efficiency. Brands kept paying for clicks, but the pages converted fewer of them.

Brand control disappeared. The title is a brand’s promise on the digital shelf. When Amazon’s AI decided it, positioning, voice, and differentiators became whatever the model thought was generic-sounding enough to be safe.

Losing the title was not a formatting problem. It was losing control of how brands are found, ranked, and advertised.

Why 75 characters was never really the story

Here is the frame most sellers missed in the run-up to July 27. Amazon compressed titles to 75 tight characters precisely so machines could parse them cleanly, at the exact moment machines were becoming the primary readers of product listings.

This is the argument I’ll be developing at my eTail East keynote, “Feed the Agent: Content Is the New Frontier in Agentic Commerce.” The July 27 deadline was not really about titles. It was a preview of how retail is going to work from now on. Compliance was the surface story. The deeper story is that discovery itself has shifted from human browsing to machine reading, and Amazon just made the shift official.

Shopping has become agentic

Increasingly, an AI agent, not a person scrolling, reads your listing and decides whether to surface, recommend, or buy your product. Three forces are driving this shift.

Conversational discovery. Shoppers are asking full-sentence, intent-rich questions instead of typing two-word queries. “Best wireless earbuds” is becoming “wireless earbuds that stay in during running and have at least 6 hours of battery.” Alexa+ and Amazon Rufus answer those questions using your listing content as source material.

AI intermediaries. Alexa for Shopping, Rufus, ChatGPT, and Gemini increasingly sit between the shopper and the shelf, filtering options on the shopper’s behalf. When a shopper asks any of these agents for a recommendation, the agent parses listings, matches attributes to the query, and returns a curated shortlist. Your listing does not get considered unless the agent can understand it.

Machine-readable content wins. Agents do not scan images or absorb brand vibes. They parse text and structured data. Your listing is the API that AI agents read. Every empty attribute field is a missed connection. Every generic phrase is a lost recommendation.

The recovery and compliance playbook

Whether you’re recovering from a bad AI rewrite or getting ahead of future updates, the discipline is the same.

1. Audit which ASINs took the hit

Pull organic ranking data for the 30 days before and 30 days after July 27 for your top-revenue ASINs. Flag any listing that lost first-page ranking on high-intent terms. These are your recovery priorities. The temptation is to treat this as a catalog-wide problem, but the damage is concentrated on the ASINs where AI rewrites stripped high-value keywords.

2. Write for the question, not just the keyword

Old model: stuff every possible keyword into the title, hoping something matches. New model: write content that specifically answers the questions shoppers ask. If someone asks Rufus “which sun hat is packable and has UPF 50+,” the listings that get surfaced are the ones that literally answer both parts of the question in their title, bullets, and attributes.

3. Complete every structured data field

Attributes like material, size, age range, use case, and features are how AI maps your product to a query. These fields feel unglamorous compared to title and hero image, but they are load-bearing for machine discovery. Every empty attribute is a query you cannot appear for.

4. Find your best keywords in your own data

Two data sources you already own tell you exactly what keywords deserve a spot in your 75 characters.

Search Query Performance reports (Amazon SQP for Amazon, Search Insights for Walmart) show the exact queries shoppers used before buying your product. That is a ready-made list of buyer-intent terms. If a query drives purchases, it belongs in your listing.

Your advertising data is already paying to test terms across broad matches and discover what resonates. When a keyword drives a sale, that is a customer confirming it is relevant to your product. Move those winning terms into your organic listing content and you unlock the organic ranking signal without paying for it a second time.

5. Reject keyword stuffing, embrace intentional relevance

The shotgun approach, stuffing every plausible keyword and hoping something ranks, is actively counterproductive in an agentic commerce world. AI agents are trained to recognize genuine relevance and dismiss noise. Target the terms that are genuinely relevant to your product and audience, support them with real use cases and benefits, and reinforce them consistently across title, bullets, description, and attributes.

The listing and advertising flywheel

When listing content and advertising work together, both improve. Marketplace and ad signals like SQP data and paid performance reveal what converts. That intelligence feeds AI-optimized listing content. Better content lifts both paid and organic performance, which feeds the loop again. Better ads produce better listings, which produce better ads.

The brands that used July 27 to align their listings with their actual conversion data are pulling further ahead every week. The brands that let Amazon’s AI decide are spending August trying to recover what they lost.

How ARI Catalog solves both problems

Teikametrics built ARI Catalog, part of our ARI (Artificial Retail Intelligence) platform, to solve exactly this problem. ARI Catalog Smart Pages is our patented algorithm that rewrites catalogs for the 75/125 format while protecting the keywords brands cannot afford to lose.

The system uses each seller’s own Search Query Performance and advertising data to decide what stays in the title. It generates compliant 75-character titles and 125-character Item Highlights automatically. It scales from 50 ASINs to 50,000 ASINs in a single pass. And you review and approve every word before it goes live, so nothing publishes without human oversight.

Here’s what recovery looks like when the flywheel runs. On a recent customer engagement using ARI Catalog Smart Pages to rebuild a set of underperforming listings, the results after one content optimization cycle:

One content upgrade. Compounding returns across both organic and paid.

Your recovery options

You have three paths, all of which we support for existing ARI Catalog customers at no additional cost.

Review and Approve is best for teams that want full oversight. Teikametrics generates recommendations, you review and approve every change before publication, and we prioritize the ASINs where you lost the most ranking first.

Fully Managed is the fastest path back to compliance and recovered rankings. Teikametrics executes all updates, provides a complete audit of changes, and delivers a 30-day performance summary showing what recovered.

Collaborative works for teams that prefer weekly meeting cadence, prioritizing the catalog together and implementing agreed optimizations progressively.

All three deliver the same outcome: fully compliant 75/125 listings, recovered listing quality scores, and keyword protection driven by your own performance data.

The bigger picture

July 27 was not really about titles. It was about who reads your listings from now on. Amazon compressed the title to 75 characters because machines need clean, parseable content, and machines are now the primary readers.

Every discovery surface, from standard search to sponsored placements to Alexa for Shopping to third-party AI agents like ChatGPT and Gemini, pulls from the same source: your listing content. Fix the content, and every surface improves. Neglect the content, and every surface degrades.

This is what agentic commerce looks like in practice. July 27 was your first mandatory encounter with it. The brands treating this as a compliance task are going to spend the next twelve months trying to recover. The brands treating it as the start of a new operating discipline are going to compound advantage.