I have written about A+ content a lot. Module design. Module sequencing. The quality score. The rejection playbook. What Rufus reads. Which claims belong in the carousel and which belong in A+. Premium A+. Running A+ at catalog scale.
Every one of those is about where things go and what they look like. Not one of them is about the words.
That's not an oversight on my part. It's an accurate reflection of how the work actually happens in most brands. The A+ project gets briefed as a design project, the layout gets chosen first, and then somebody — usually whoever is free that week — writes copy to fill the boxes that already exist. The words are the last thing produced and the first thing nobody reviews.
Which is a problem, because on a surface where the shopper has already clicked, already scrolled past the buy box, already decided you're a candidate, the words are the argument. The images set the tone. The copy closes or doesn't.
So this one is about the text. What to write, how long, in whose vocabulary, and the specific ways A+ copy fails that nobody catches because a design file always looks finished.
The core problem: copy written inside a layout gets written to fit the box
Here is the sequence in almost every A+ project I've reviewed.
Someone picks the modules. A designer builds the layout in Figma or straight into the Content Manager preview. The text boxes exist, with visible dimensions. Then the copy gets written into those boxes, and the writer's actual job — whether they'd describe it this way or not — becomes filling the space attractively.
You can see it in the output. Three modules with body copy that's within ten words of each other, because they're the same module type. Headlines that are all roughly the same length, because a short one would look odd next to the others. A four-icon grid where the fourth benefit is visibly weaker than the first three, because there were four icon slots.
That last one is the tell I look for first. When the number of claims exactly matches the number of slots, the layout wrote the copy.
The fix is boring and it's the whole thing: write the copy in a document, before the layout exists. One line per claim, in plain text, with no idea yet which module it lands in. Then pick modules that fit the claims. If you end up with three strong claims, use three. Nobody has ever bounced off a listing because the A+ had five modules instead of seven.
Who's actually reading this
A+ is a resolution surface, not a qualification surface. Nobody lands on your A+ content by accident. To see it they had to click your listing from a grid of competitors, not bounce, and scroll past the buy box, the bullets and usually a row of sponsored ads.
That's a self-selected reader who has already decided you're a candidate and is now deciding whether to buy. Which changes what the copy is for. The carousel argues consider me. A+ argues this one, not the other one you have open in another tab.
Two consequences that most A+ copy ignores.
They already know what the product is. Restating the category in your first headline is a wasted line to a reader who clicked on the product. "Premium Stainless Steel Water Bottle" as an A+ headline tells a person who is looking at a stainless steel water bottle nothing.
They arrived with a specific worry. People rarely scroll into A+ out of general enthusiasm. They scroll because something is unresolved — will it fit, will it last, is it actually different from the $12 one, what happens if it breaks. Your copy either meets that worry or it doesn't, and no amount of brand voice compensates.
Front-load, because a third of your readers never reach the bottom
Average mobile scroll depth on A+ runs in the low sixties as a percentage. Call it around 62%. Applied to a seven-module layout, roughly a third of the people who reach your A+ never see modules five through seven.
That number should govern your copy order, and in most brands it governs nothing. The most common A+ sequence I see puts brand story first, features in the middle, and the comparison chart — the single highest-CVR-impact module in the whole stack — at position six, where a third of readers never arrive.
Write the decision-changing copy for modules one through three. Whatever answers the objection that most often costs you the sale goes at the top. Brand credentials, warranty language and origin stories are real and worth including, and they belong further down, where the reader who is already convinced goes looking for reassurance.
If you only fix one thing after reading this, fix the order. It costs nothing and doesn't require rewriting a word.
The headline is the copy
Most shoppers do not read A+ body text. They scan headlines, look at images, and read body copy only where a headline made them want more.
That means the headline is not a label for the section. It is the entire message, and the body is the evidence.
Look at what most A+ headlines actually say:
- "Premium Materials"
- "Built to Last"
- "Designed for You"
- "Quality You Can Trust"
Those aren't headlines. They're file names. A reader scanning at speed gets nothing from them, and the body copy underneath — which might be genuinely good — never gets read because the headline gave no reason to.
Now the same four sections written as claims:
- "The same 304 stainless used in commercial kitchens, not the 201 in most sub-$20 bottles"
- "Rated for 10,000 open-close cycles — about eight years of twice-daily use"
- "Fits every standard car cup holder we could find, including the narrow ones in older Hondas"
- "If the seal fails in two years, we replace it, no receipt needed"
Same product, same modules, same layout. The second set is a page a person can decide from at a glance. The first set is decoration.
The test I use: cover the body copy and read only the headlines top to bottom. If that alone doesn't make the case for the product, your A+ is not doing its job for the majority of the people looking at it.
Character limits are ceilings, not targets
A+ module fields have generous limits and brands treat them as briefs. Commonly cited limits across the standard module set:
- Many standard modules allow a headline up to around 160 characters
- Standard Single Image & Sidebar: headline up to ~160, body up to ~1,000
- Standard Single Image & Highlights: three text blocks, roughly 1,000 / 400 / 400
- Image-and-text modules: body typically kept under ~350
- Premium text modules: headline around 80, body up to ~5,000
Confirm the exact numbers for your module set in Content Manager rather than off any blog, including this one — Amazon has adjusted these and they vary by module and program. But the shape is what matters, and the shape is the problem.
A 160-character headline is about 25 words. That's not a headline. That's a sentence and a half, and the field will happily render it at headline size, which is how you get an A+ page that reads as a wall of large text with no hierarchy at all.
Working targets I'd actually brief:
- Headline: 6–12 words. One claim. If you need a comma and an "and," you have two claims and should use two modules or cut one.
- Body: 25–50 words per block. Two or three short sentences. Enough to substantiate the headline and stop.
- Premium body fields: same discipline. A 5,000-character limit is not permission to write 5,000 characters. I have never seen a long A+ body block outperform a short one, and I've never seen anyone read one.
The Premium text module is where this goes worst, because the limit is so large that it invites a brand narrative. If your brand story genuinely needs 900 words, it needs a Brand Story module and a website, not a text field on a product page.
Write in the buyer's vocabulary, not the brand deck's
The single highest-return copy input available to you is free and almost nobody uses it: your own reviews and return reason comments.
Your brand deck says "ergonomic contoured grip." Your reviews say "doesn't kill my wrist." Your A+ copy is almost certainly using the first one, because the person writing it was handed the deck.
The vocabulary gap matters for three separate reasons:
Recognition. A shopper scanning at speed recognises language they've thought in. Marketing language requires a translation step, and translation steps at speed become skips.
Credibility. Copy written in the words customers actually use reads as though somebody has met a customer. Copy written in deck language reads as though somebody has met a deck.
The machine layer. A+ isn't a primary indexing field, but Google indexes the page and the AI shopping layer reads your A+ as context about what this product is and who it's for. Natural language that mirrors how people describe the product is more useful to both than a phrase engineered for a keyword tool.
The extraction is a 20-minute job. Pull your last 50 reviews, both good and bad. Highlight every phrase describing what the product does or fails to do. Cross out the ones that are just adjectives. What's left is your headline vocabulary, and it will be more specific and less flattering than anything in your brand guidelines.
Write against objections, not features
A feature list is what you have. An objection list is what's stopping the sale. They overlap less than brands expect.
Build the objection inventory from three sources, in this priority order:
- Return reason comments. The most expensive feedback you get, and the only feedback from someone who already bought. If buyers keep writing "smaller than I expected," that's a top-three A+ module, not a spec line.
- Customer questions on the listing. Questions asked publicly are questions asked silently by fifty other people who just left.
- Negative reviews on competitors in your category. The objections you haven't earned yet.
Then write one module per objection, headline-first, and stop when you run out of real objections. Four modules that each kill a real objection beat seven that each restate a benefit.
The module most often wasted here is the FAQ module, which brands routinely fill with marketing copy in question form ("Why choose [Brand]?"). Its actual job is edge cases — the compatibility question, the care instruction, the "what if it doesn't fit" — and those are exactly the questions that produce returns when left unanswered.
The comparison chart is copy, and the row labels are the copy
The comparison chart is the highest-leverage module in most A+ stacks and the one where copy quality collapses fastest, because the row labels look like a spec table and get written like one.
Row labels written as features:
- Material
- Capacity
- Weight
- Warranty
Row labels written as buyer questions:
- Will it survive the dishwasher
- Does it fit a 32oz slot
- Light enough for a day pack
- Covered if the seal goes
Same rows, same data, completely different work. The first version asks the shopper to interpret. The second version does the interpreting, which is what you're there for.
Two more rules for chart copy. Only include rows where you win or tie — a chart is an argument, not an audit, and you're not obliged to itemise your weaknesses. And when comparing against your own SKUs rather than competitors, write the row labels around which one is for me, because that's the only question a shopper reading a same-brand comparison chart is actually asking.
The indexing question, answered honestly
Somebody always asks whether A+ text is indexed, usually because they want to know if it's worth stuffing keywords into.
The honest answer is that the claim is contested and nobody outside Amazon can settle it. Amazon's historical stated position has been that A+ text is not indexed for search. A number of experienced practitioners have published tests claiming partial indexing, particularly on image alt text. Both camps are arguing about a system with unpublished weights, and anybody quoting you a percentage on it is making it up.
Here's what's not contested, and it's enough to write from:
Google indexes your Amazon detail page, A+ text included. That's a real long-tail acquisition channel and it rewards readable prose, not comma strings.
The AI shopping layer reads A+ as context. It's deciding whether your product belongs in a consideration set, and natural-language descriptions of use case, fit and material are the most legible signal you can give it. This is also why identical A+ copy across 60 ASINs is actively harmful — you're telling the model in the clearest possible terms that your 60 products are the same product.
Image alt text is a real field and most brands leave it empty. It takes seconds per module, it's the accessible description of an image that is otherwise invisible to every non-human reader, and it's the cheapest per-SKU differentiation available on a family of visually similar modules.
So: write for humans in natural language, fill the alt text, and stop treating a 1,000-character body field like a backend search term box. That posture is correct under either answer to the indexing question, which is the point.
The compliance layer, and why shared modules multiply it
A+ copy gets rejected and — more expensively — gets approved and then pulled in a later sweep. The categories that cause it are predictable:
- Superlatives and unsubstantiated rankings — "best," "#1," "number one selling"
- Guarantee and warranty language that promises something your actual policy doesn't
- Health, safety and efficacy claims, especially in supplements, beauty, baby and anything touching skin
- Price, promotion and shipping references, which age badly and are generally not permitted
- Competitor names and trademarks in comparison content
- Contact information or off-Amazon URLs
None of that is new. What's new — for anyone running A+ across a real catalog — is that a shared module template concentrates the risk. One over-reaching warranty sentence in a bespoke project is a single-listing problem. The same sentence in a chassis applied to 80 ASINs is a catalog event.
Two disciplines that cost nothing. Run the compliance read on the template before it multiplies. And keep a one-line substantiation note for every contestable claim you publish — what backs it, and where. That file does double duty the day somebody files a complaint against your listing.
What good A+ copy actually reads like
Pulling it together, here's the shape of a module that works:
Headline: one claim, 6–12 words, specific enough that a competitor couldn't paste it under their own product. Body: 25–50 words that substantiate the headline with a number, a material, a test, or a use case. Not a restatement. Image: carries part of the argument the copy doesn't have to make. If the image shows the thing fitting in a cup holder, the copy doesn't need to describe the cup holder. Alt text: a plain description of what the image shows, written for someone who can't see it.
And the whole page passes two tests: headlines-only makes the case, and read aloud, it sounds like a person who has used the product rather than a person who has read a positioning document.
The 30-minute A+ copy audit
Run this on your top revenue ASIN.
- Open your live A+ on a phone. Not a desktop preview. The render and the line breaks differ and you will find copy that wraps badly.
- Read only the headlines, top to bottom. Write down the argument they make. If you can't, you've found the biggest available fix in your stack.
- Count the claims per module. More than one means the headline is doing two jobs and both are weaker for it.
- Highlight every phrase that came from the brand deck. Replace each with the closest phrase from your own reviews.
- Check module order against your objection list. The most expensive objection should be resolved in the first three modules. It usually isn't.
- Read every word out loud. Anything that makes you self-conscious to say is copy a shopper won't believe.
- Check the alt text fields. If they're empty, fill them — that's ten minutes for a per-SKU signal you currently aren't sending.
Nothing on that list requires a designer, a photoshoot or a budget. Five of the seven are edits to text fields you already control.
FAQ
How long should A+ content copy be overall? Across the whole page, 250–500 words of body copy is plenty for most products. Technical and high-consideration categories run longer, sub-$30 impulse products run shorter. If you're past 800 words you are almost certainly writing for yourself.
Should I put keywords in my A+ content? Write natural language that happens to use the words your buyers use, and let that be your keyword strategy. The indexing question is contested; the readability cost of comma-stuffing is not. You lose nothing by writing well and you lose real conversion by writing a keyword list.
Do I need different copy for every ASIN in a family? Yes for the product-specific modules, no for the brand modules. A shared chassis with a brand story, a credentials module and a locked visual system is efficient and correct. The objection module, the comparison chart and the FAQ have to be written per SKU or they're describing a product the shopper isn't looking at.
Does Premium A+ need different copy? Not different copy — the same copy, executed on better modules. Premium gives you more space and richer formats, which is exactly why it produces the worst over-writing I see. The discipline gets more important as the canvas gets bigger, not less.
Who should write A+ copy — the designer, the agency, or us? Whoever has read your reviews. That's the only qualification that reliably predicts good A+ copy. A brilliant copywriter working from a brand deck will produce something polished that argues for the wrong things; a competent one working from return reason comments will produce something that sells.
Every A+ project I've audited had a design review. Somebody looked at the layout, argued about the colour of the icon bar, and signed it off. Almost none of them had a copy review — a meeting where somebody read the words alone, without the pictures, and asked whether they made the case.
That meeting takes twenty minutes and it's the one that changes the number.
If you want the rest of the system around this, start with the division of labor between your image stack and your A+, then the module sequencing and, if you're running more than 20 SKUs, the catalog-scale module system.