Amazon bullets are where the buy decision actually happens, not the title and not the images. The five bullets above the fold are the only block a shopper reads before deciding. A buyer should answer three questions in five seconds: what is it, why this one, what is the catch. Most listings fail because the bullets paraphrase the title five times and skip the catch entirely. The prompts below force feature-to-benefit conversion, mobile-truncation safety, and marketplace-specific phrasing.
TL;DR
- Amazon gives you 5 bullets. The conversion sweet spot is 150–250 characters each (Brand Registry can technically stretch higher, but display and readability cap the payoff). Keep all five combined under ~1,000 characters.
- About two-thirds of Amazon shoppers buy on mobile, where each bullet truncates at roughly 80 characters before “Read more.” Put the decisive benefit first.
- Since the AI shopping assistant Rufus was renamed “Alexa for Shopping” on May 13, 2026 (US; logic unchanged), bullets and A+ Content are now read as a semantic knowledge source. Write explicit use cases (“for runners who pronate inward”), not keyword piles. Products under 4.0 stars are typically not surfaced by the assistant.
- Run these prompts in any current model (GPT-5.5, Claude Sonnet 4.6, Gemini 3.1 Pro all handle them well). For long review-mining pastes, the 1M-token context on Sonnet 4.6 or Gemini 3.1 Pro lets you drop in hundreds of reviews at once.
Specs that constrain every bullet (as of June 2026)
| Constraint | Value | Why it matters |
|---|---|---|
| Bullets shown | 5 | Amazon caps the list; bullet 6+ is ignored |
| Per-bullet limit | ~200 chars standard, up to ~250 with Brand Registry | Going long hurts mobile readability more than it helps SEO |
| Mobile cut-off | ~80 chars before “Read more” | ~2/3 of buyers shop mobile; the front half is what sells |
| Prohibited | superlatives (“best,” “#1”), competitor brand names, price/promo claims, unverifiable health/legal claims, emojis, ALL-CAPS run-ons | Triggers suppression or A+ rejection |
| AI assistant signal | use-case and problem language | Alexa for Shopping ranks on intent match, not keyword density |
Best for
- Amazon listings (Seller Central + Vendor Central)
- A+ Content modules (Basic, Premium, Brand Story)
- Re-launching slow listings
- Multi-marketplace SKU rollout
- Brand registry catalog rewrites
1. 5-bullet structure
Product: [description + specs]. Write 5 Amazon bullets, each opens with a CAPS benefit phrase (max 6 words), then a colon, then up to 25 words of feature to benefit. Vary themes across bullets: durability, comfort/fit, primary use case, edge use case, package contents. No "high quality" anywhere. Keep each bullet under 200 characters.2. Competitor bullet gap analysis
Below are my current bullets vs the top 3 competitors on the same keyword. Identify: (a) gaps in my bullets, (b) 1 thing each competitor does better, (c) 1 claim every competitor makes that I should answer pre-emptively. Output a 5-row table.
[paste all]3. Bullets for problem-aware buyers
Audience already knows the problem ([problem]). Write 5 bullets, each addressing one buyer doubt. Open each bullet with the doubt in CAPS ("TIRED OF X?") then resolve with benefit + feature in up to 25 words. Doubts should not repeat between bullets.4. Bullets for solution-aware buyers
Audience knows alternatives exist and is comparing. Write 5 bullets, each emphasizing a differentiator vs the most common alternatives, but never name competitors. Use phrases like "unlike clip-on versions" or "without the usual battery swap" so buyers connect the dots themselves.5. A+ Content block titles + body
For 4 A+ Content image+text blocks, write block titles (max 8 words) and 50-word descriptions, each on a different angle: durability, design language, primary use case, brand origin story. Block 1 must work as a standalone hero, assume the buyer scrolls past the rest. Use plain, legible sentences so the AI shopping assistant can parse them.6. Amazon-policy and SEO audit
Below are my current bullets. Flag every Amazon policy or SEO mistake: superlatives ("best", "#1"), brand mentions of competitors, claims that require FDA/legal evidence, price or promo language, length issues (>250 chars per bullet), keyword stuffing, all-caps run-on, emojis. Rewrite each flagged bullet beneath the flag.
[paste]7. Internationalize bullets
My bullets work for the US marketplace. Adapt for UK / DE / JP buyers: units (oz to g, inch to cm), idioms, voltage references, cultural references, formality level (DE is more direct, JP more deferential). Flag any claim that may fail each marketplace's rules.
[paste US bullets]8. Negative-review-mining bullets
Below are 20 negative reviews of my product. Cluster them by root complaint, count each cluster, then for the top 3 clusters write a bullet that pre-empts the complaint ("FITS TRUE TO SIZE: see size chart in image 3, runs identical to industry standard").
[paste reviews]9. Mobile-truncation safe rewrite
On Amazon mobile, only the first ~80 characters of each bullet show before "Read more". Rewrite the 5 bullets so the most important benefit is fully readable in the first 80 characters. Mark a "[CUT]" at the 80-char mark of each so I can verify.
[paste current bullets]10. Bullets for gift-buyer search intent
Many buyers of [product] are giving it as a gift. Rewrite the 5 bullets so they speak to the gift-giver, not the end user: who it suits, presentation, returns, gift packaging. Keep the spec content; change the frame. Avoid "perfect gift", show why.11. Voice-search and Alexa-for-Shopping bullets
Some buyers reach this product through voice search and Amazon's AI shopping assistant (Alexa for Shopping, formerly Rufus). Rewrite my bullets in natural-spoken language that reads aloud cleanly: no CAPS, no symbols, no arrow characters. Lead each sentence with one explicit use case or problem solved, 1 benefit per sentence, conversational rhythm. Keep keyword coverage intact.
[paste current bullets]12. Subscribe-and-save / replenishment-cycle angle
This is a consumable that buyers reorder every [N] weeks. Rewrite bullets 3 to 5 to surface the subscribe-and-save value: cost per week, no-reorder convenience, easy cancel. Bullets 1 and 2 stay product-focused. Output all 5 in order.Common mistakes
- Generic adjectives (“high quality”, “premium”, “best in class”) that Amazon’s filters can flag and buyers ignore
- Skipping size, fit, voltage, or compatibility specifics, the #1 negative-review trigger
- Repeating the brand name in every bullet, eating character budget
- One bullet set shipped to all marketplaces without unit or idiom adaptation
- Writing for desktop preview when ~2/3 of buyers see only the first 80 mobile characters
- Treating bullets as keyword bins instead of intent statements the AI assistant can read
FAQ
How many bullet points does Amazon allow, and how long can each be? Five bullets per listing. Standard sellers get roughly 200 characters per bullet; Brand Registry accounts can stretch to about 250. Best practice is 150–250 characters each, with all five combined under ~1,000 characters. Longer bullets rarely lift SEO and almost always hurt mobile readability.
Why do the prompts push the benefit into the first 80 characters?
On Amazon’s mobile app, each bullet truncates at roughly 80 characters behind a “Read more” link, and about two-thirds of shoppers buy on mobile. If the decisive benefit lives in the back half, most buyers never see it. Prompt 9 inserts a [CUT] marker at the 80-character mark so you can verify before publishing.
Do bullets matter for Amazon’s AI shopping assistant? Yes. Amazon renamed Rufus to “Alexa for Shopping” on May 13, 2026 (US), keeping the recommendation logic unchanged. The assistant reads bullets and A+ Content as a semantic knowledge source and ranks on intent match, not keyword density. Phrasing like “for runners who pronate inward” outperforms a keyword pile, and products under 4.0 stars are typically not recommended at all. Prompt 11 rewrites bullets in that explicit use-case voice.
Which AI model should I run these prompts in? Any current frontier model handles them: GPT-5.5, Claude Sonnet 4.6, or Gemini 3.1 Pro. For Prompt 8 (mining dozens or hundreds of reviews at once), Sonnet 4.6 and Gemini 3.1 Pro both carry a 1M-token context window, so you can paste the full review dump without trimming.
Can I use the same bullets across US, UK, DE, and JP? No. Beyond translation, you must convert units (oz to g, inch to cm), swap idioms, adjust voltage references, and shift formality (German buyers prefer direct claims, Japanese buyers more deferential phrasing). Each marketplace also enforces its own restricted-claim rules. Prompt 7 handles the adaptation and flags claims that may fail per market.
Related
- Product description prompts
- Review reply prompts
- Write Amazon Bullet Points With AI: 5 Bullets That Win on Benefit
Tags: #Prompt #E-commerce #Amazon