I have audited about 600 Amazon A+ Content modules in the past 18 months. Until late 2025, the optimization framework was simple: large images, scannable benefit copy, comparison chart in slot 4 or 5, and a brand story module if Premium A+ was approved. That framework still works for human shoppers. It does not work for Rufus.
Rufus now drives somewhere between 15% and 20% of product detail page interactions on Amazon, depending on category. When Rufus answers a question — "is this safe for newborns," "does this fit a 2019 Tacoma," "how does this compare to brand X" — it is parsing your listing. Title, bullets, A+ Content, customer reviews, Q&A, and structured data are all being read by a model that was not in the spec when most A+ Content was built.
If your A+ is image-heavy with copy baked into the graphic, Rufus cannot read it. Your competitor with text-based modules and proper alt text is now the listing Rufus quotes from.
This is the playbook I am running for clients in 2026.
What Rufus Can Actually Read in Your A+ Content
Start with the technical reality. A+ Content is a series of modules. Each module has:
- Image alt text (most sellers leave this blank or paste the title — wasted)
- Module headlines and body text (text fields, fully readable)
- Image-embedded text (graphic copy baked into the image — Rufus cannot extract this reliably)
- Comparison chart cells (structured table data — highly readable)
The crucial point: anything you bake into an image as graphic text is invisible to Rufus. The "5x more absorbent" headline you put on top of your hero infographic is not being parsed. The same five words in the module headline field would be quoted in Rufus answers.
I ran a test across 14 client listings in February 2026. We took identical messaging and split it: half kept feature copy as image-embedded graphics (the standard "designer A+" approach), half rebuilt with the same copy in module text fields and the image stripped of text. The text-field versions appeared in Rufus answers 3.2x more often when we ran identical query sets through Rufus a month later.
That is the entire game. Move the words out of the picture and into the field.
The 5 Modules I Now Prioritize for Rufus Visibility
Not all A+ modules are created equal for AI parsing. Here is the priority order I use:
1. Comparison Chart. The single most Rufus-readable module on Amazon. Structured rows and columns Rufus reads as a literal table. Comparison queries ("which one should I buy") are the fastest-growing query type Rufus handles. Every client I work with has a comparison chart now, even single-SKU brands — they compare to category alternatives.
2. Standard Image & Dark Text Module. Heavy text body, clear headline, supporting image. The text body is parsed in full. This is your "explanation" module — where you put the why behind the product.
3. Four-image / Text Module. Each image gets its own headline and body block. Used for feature breakdowns. Each block is a standalone parseable chunk.
4. Brand Story Modules (Premium). The Q&A module specifically. Question-answer format is exactly the structure Rufus generates. Putting your top 5 customer questions and answers here is essentially feeding Rufus its own format.
5. Standard Single Image & Sidebar. Sidebar text is parsed. Useful for warranty, shipping, sourcing details — facts Rufus likes to surface for trust queries.
What I deprioritize now: image-only modules with no text fields, "lifestyle banner" modules that are mostly visual atmosphere, and modules where the headline is a vague brand slogan instead of a feature claim.
Alt Text Strategy That Actually Helps Rufus
Most sellers handle alt text in one of three wrong ways: leave it blank, paste the product title, or stuff keywords. None of these help.
The right approach treats alt text as a literal description of what the image shows plus the claim it supports. Format I use:
"[Product] [doing what] showing [feature/benefit]. [Specific number or detail]."
Example for a baby bottle:
"Anti-colic baby bottle in mid-feed showing internal venting system. Reduces colic symptoms in 87% of babies tested."
That alt text gives Rufus three things to parse: what the product is, what feature is being shown, and a specific claim. It is also fully ADA-compliant, which Amazon's accessibility scoring rewards.
Do not stuff keywords. Rufus is a language model — keyword stuffing reads as low-quality content and gets discounted. Write the alt text as if explaining the image to a smart person on the phone.
The Comparison Chart Rebuild Every Brand Needs
Most comparison charts I audit are built for visual scanning — checkmarks vs Xs, pretty icons, brand names across the top. They look fine to humans. They are weak for Rufus.
The Rufus-optimized comparison chart has:
- Specific values, not checkmarks. "32 oz capacity" beats "✓ Large capacity." Rufus quotes specific values; it does not quote checkmarks.
- Comparable competitor names if allowed in your category. If your category permits, name the alternatives. Rufus loves named-comparison data.
- Numeric claims. Weight, dimensions, count, percentage, time — anything quantifiable. Rufus answers numerical queries with numerical data when it has it.
- Use case row. "Best for [specific scenario]" — this row catches intent-based Rufus queries.
I rebuilt comparison charts for a kitchen brand client in March 2026. We replaced 8 checkmarks with specific numeric values and added a "best for" row. Rufus citation rate for those listings (measured by Rufus referencing the listing in answers to category queries) went from 12% to 41% in 6 weeks.
Brand Story Q&A Module: The Highest-Leverage Move in Premium A+
If you have Premium A+ access, the Brand Story Q&A module is the highest-leverage module on Amazon right now for Rufus visibility. The format is literally question, then answer. That maps 1:1 to how Rufus generates responses.
Pull your top 5 customer questions from:
- Customer Q&A on your listing
- Customer review patterns ("I wish I knew..." reviews)
- Search query data showing question-form keywords
- Common DM questions if you sell DTC too
Write answers that are 2–4 sentences, factual, and include the product name and key feature. Avoid marketing copy. Treat it as a knowledge base entry, not an ad.
For one supplements client, the top question was "is this third-party tested." We made that question 1, with an answer naming the testing lab, frequency, and certificate availability. Rufus now quotes that answer almost verbatim when shoppers ask third-party testing questions in their category.
The Mobile-First, AI-Readable Module Hierarchy
Even with Rufus prioritization, 80%+ of A+ views are still mobile humans. The framework that wins both is what I call the "double-readable" module: a module that works as a scrollable visual story for humans AND as parseable text for Rufus.
Every module I now build has:
- A clear headline in the text field (Rufus-readable, also displayed visually)
- 2–4 sentences of body text (Rufus-readable, supports the visual)
- An image that illustrates the claim without containing the claim's text
- Alt text describing the image plus the claim
- Mobile-tested for legibility on a 5.5-inch screen
If a module fails any of those five tests, I rebuild it. The brands shipping pure-graphic A+ Content in 2026 are losing Rufus citation share to brands shipping double-readable A+. The gap is widening every quarter.
Common A+ Anti-Patterns I See Every Audit
The patterns I find in 70%+ of A+ audits I run:
- Headlines as image text only. Module headline field empty, headline baked into the graphic. Invisible to Rufus.
- Body copy as image overlays. Beautiful typography on a lifestyle photo. Reads as nothing to Rufus.
- Comparison chart with no values. All checkmarks. Rufus has nothing to quote.
- Generic alt text. "Product image 1." Wastes one of the cheapest optimization fields on the platform.
- Brand story used for vibes. Premium A+ Brand Story slots burned on aesthetic founder quotes instead of Q&A.
- Module sequence built for designers, not shoppers. Brand story first, features last. The opposite of what converts and the opposite of what Rufus wants to read.
Each of these is a fix that costs nothing and compounds over months as Rufus traffic grows.
FAQ
Will rebuilding A+ for Rufus hurt my human conversion rate? No, in every test I have run, the double-readable module outperforms graphic-heavy modules on human conversion as well. Mobile shoppers scroll-read; they don't squint at graphic text. Cleaner text fields and stronger images both help.
How long does Rufus take to reflect A+ Content changes? In my experience, 2–6 weeks. Rufus does not re-index instantly. Plan A+ rebuilds with that lag in mind.
Should I rewrite all my A+ at once or test module by module? Test by module if you have the bandwidth. The comparison chart is usually the highest-impact single rebuild. Start there, measure for 4 weeks, then move to Q&A and feature modules.
Does this apply to Premium A+ only? No. Standard A+ benefits just as much. Premium A+ adds the Q&A module and longer-form options that are particularly Rufus-friendly, but the core principle — text in fields, not in pictures — applies at every level.
What about videos in A+? Video transcripts are parseable if you upload caption files. Most sellers don't. Caption your A+ videos. It is a 30-minute job that adds another parseable surface to your listing.
If you want an audit of your A+ Content for Rufus readability, I run them as part of every full creative engagement at Aspi. The fix list is usually 10–20 specific, surgical changes that compound over the next two quarters as AI search share keeps growing.