Amazon listing optimisation in 2018 meant keyword-stuffing the title. In 2022 it meant A+ Content + lifestyle imagery. In 2026 it means architecture: a catalog discipline + a documented optimisation workflow + a measurement loop that actually attributes lift to specific changes.
This is the reference architecture we apply on Amazon listing engagements at Rs 25 crore+ Amazon ARR. It covers title structure, A+ Content design, image strategy, backend keywords, the testing methodology, and the operating cadence.
Why most listing optimisation under-delivers
Three failure modes recur:
"We optimised the listing 8 months ago". and haven't touched it since. Amazon's algorithm + competitive landscape + customer language all shift. A static listing decays.
"We A/B tested the title and saw no lift". but the test ran for 2 weeks with insufficient traffic + no proper control. The lift was statistically invisible, not actually zero.
"We hired an Amazon agency and they sent us a 47-page audit". that recommended every possible change with no prioritisation, no implementation plan, no measurement framework.
The fix is architectural, not creative.
The five layers of a listing
A listing is not a single artefact. It's five layers, each with its own optimisation surface + measurement model.
| Layer | What it is | Optimization surface |
|---|---|---|
| Title | The product title (200-char limit) | Search relevance + click-through |
| Bullets | The 5 bullet points | Pre-purchase information + conversion |
| Description | Long-form text or A+ Content | Deep engagement + objection handling |
| Images | Hero + gallery + lifestyle + infographics | Click-through + conversion |
| Backend keywords | Search terms field + attributes | Search visibility for non-obvious queries |
Each layer is optimised independently, with its own success metric. Treating the listing as one undifferentiated artefact loses the ability to measure what's working.
Title structure: where most lift hides
The title is the single highest-use field. It drives search relevance (Amazon's A10 algorithm reads it heavily), click-through from search results (the title is the headline), and Sponsored Ads ad copy (Sponsored Products + Sponsored Brands often use parts of the title).
The Dcrayons title formula
For most India D2C product categories:
[Brand] [Product Type] [Key Attribute 1] [Key Attribute 2] - [Use Case] - [Size or Pack]
Example for a skincare serum:
Dcrayons Vitamin C Serum 20% with Hyaluronic Acid - Brightening + Anti-Aging - 30ml
Breakdown: - Brand first (Amazon prefers this; counts toward brand recognition signal) - Product type immediately after (the most-searched term) - Key attributes in order of importance - Use case (helps Amazon match to search intent) - Size or pack at the end
What doesn't work
- All-caps anywhere (Amazon penalises)
- Promotional language ("Best Selling", "Top Rated", "Featured")
- Punctuation overload (pipes + bullets + ampersands clutter)
- Special characters that don't display correctly on mobile
- Titles over 150 characters (the back half gets truncated on mobile)
Category-specific patterns
Some categories have category conventions Amazon enforces:
- Books: Title + Author + Edition
- Fashion: Brand + Item + Colour + Size
- Electronics: Brand + Model + Spec + Variant
- Grocery: Brand + Product + Pack Size + Variant
Check the category style guide before fighting the convention.
Bullets: the conversion layer
Once a customer clicks through, the 5 bullets decide whether they read further or bounce.
Bullet 1: the value proposition
What this product does + who it's for in one tight sentence. Specific outcomes, not generic benefits.
Bad: "Premium quality skincare for radiant skin"
Good: "20% Vitamin C serum that visibly brightens dull skin + reduces fine lines in 4 weeks, dermatologist-tested for Indian skin types"
Bullets 2-4: the substantiation
Each bullet handles one objection or one differentiator:
- Ingredient story (what's in it; what each thing does)
- Use case (when + how to use it; for which skin types or contexts)
- Comparison or trust signal (vs alternatives, certifications, ratings, awards)
Bullet 5: the brand or social proof
Either: - Brand story (why this brand made this product; what they stand for) - Customer proof (X customers, Y star average, top review excerpt)
Formatting rules
- 200-250 characters per bullet (mobile-readable density)
- Start each bullet with a capital letter; no period at the end (Amazon convention)
- Use ALL CAPS for the first few words ONLY for the most important benefit (sparingly)
- No HTML; Amazon ignores it
A+ Content: the deep-engagement layer
A+ Content (Brand Registry holders only) replaces the standard product description with rich modules. It typically lifts conversion 5-12 percent on a well-designed listing.
The seven A+ modules that work
| Module | Purpose | When to use |
|---|---|---|
| Hero image with text overlay | Brand voice + key benefit | Every listing |
| Comparison table | Differentiate vs your other products or generic alternatives | Multi-SKU brands |
| 4-image quadrant | Use cases + key benefits | Lifestyle + product education |
| Ingredient / spec module | Detailed breakdown | Beauty + supplements + electronics |
| Brand story | Founder + brand voice | Build trust + brand recognition |
| FAQ module | Pre-empt objections | Categories with high customer questions |
| Lifestyle photo + caption | Use context + emotional hook | Fashion + home + lifestyle |
A+ Content best practices
- Mobile-first design (over 70% of Amazon India traffic is mobile)
- Image dimensions per Amazon's specs (1464x600 for hero; 300x300 for quadrant cells)
- Text overlay sized for mobile readability
- Brand voice consistent with off-Amazon assets
- No medical / regulatory claims that don't have documented substantiation
Premium A+ (when eligible)
For brands meeting Amazon's usage criteria (top sellers + high A+ adoption), Premium A+ adds: - Video modules - Interactive carousels - Comparison tables with up to 10 products - Brand-story modules with custom layouts
Premium A+ typically lifts conversion an additional 3-7 percent on top of standard A+.
Image strategy: the click-through driver
Hero image drives click-through from search; gallery drives conversion on the product page.
The 7-image standard
| Position | Image type | Purpose |
|---|---|---|
| Main image | Product on white background, fills 85% of frame | Amazon mandated; drives search click-through |
| Image 2 | Product with size + scale reference | Conversion |
| Image 3 | Product in use (lifestyle) | Emotional hook |
| Image 4 | Infographic with key benefits | Quick value summary |
| Image 5 | Ingredient / spec breakdown | Information density |
| Image 6 | Trust badge / certification | Skeptical-buyer reassurance |
| Image 7 | Comparison or before-after (where applicable) | Differentiation |
Image technical specs
- Minimum 1000x1000 pixels (zoom enabled at this resolution)
- Recommended 2000x2000 pixels (better display on retina screens)
- JPEG or PNG, sRGB colour space
- No watermarks, no promotional text overlay on the main image, no model with face cropped above shoulder
Mobile-first image design
Most images get viewed on mobile screens around 4 inches wide. Text must be readable at that size. Test every image on a phone before publishing.
Backend keywords: the underused field
The Search Terms field in Seller Central + the attribute fields are crawled by Amazon's search algorithm but never shown to customers. This is where you handle queries the front-end doesn't address.
What to put in backend keywords
- Synonyms for your product (people search "vitamin c serum" + "vit c serum" + "ascorbic acid serum")
- Common misspellings (your audience genuinely searches "vitamine c")
- Regional language transliterations (where applicable)
- Customer language that doesn't fit gracefully in the title or bullets ("for oily skin", "for sensitive skin")
- Use-case keywords ("anti-aging", "morning skincare", "summer skincare")
What NOT to put
- Brand names (yours or competitors')
- Misleading terms ("organic" if not certified organic)
- Promotional language
- Restricted-category claims
The 250-byte limit
Backend keywords field is limited to 250 bytes (NOT characters; UTF-8 multi-byte characters count as 2+ bytes). Indian language characters often double or triple the byte cost. Plan accordingly.
Testing methodology: how to know what actually worked
Most Amazon listing changes get evaluated by "did sales go up?" That ignores seasonality, competing changes, Amazon algorithm shifts, ad spend changes, and external factors. Real attribution requires methodology.
The Manage Your Experiments (MYE) tool
Amazon's built-in A/B testing for Brand Registry holders. Tests: - Titles - Main images - A+ Content variants - Bullets
MYE runs proper randomised splits + reports statistical significance. It's the right tool for testing individual elements.
The before / after approach (when MYE isn't available)
For non-MYE-eligible tests: - Define the success metric upfront (conversion rate, units per session, click-through) - Set a baseline window of 4-6 weeks pre-change - Run the change for 4-6 weeks - Compare like-for-like (same days of week, control for seasonality, control for ad spend changes) - Account for known confounders
Statistical discipline
- Don't test more than one variable at a time on the same listing (you cannot disentangle the effects)
- Don't peek at daily numbers + bail early; pre-register the test duration
- 95 percent significance threshold; don't accept 80 percent as conclusive
- Watch for novelty effects (a change that wins in week 1 may regress in week 4)
The operating cadence
Sustained listing performance requires an operating cadence, not one-off projects.
Monthly
- Review listing-quality score per priority ASIN (Amazon publishes this internally)
- Identify listings that fell off Buy Box or F-Assured / Prime
- Refresh A+ Content where seasonal or competitive moves require it
- Review backend keyword performance via Brand Analytics
Quarterly
- Full title + bullets + image refresh per priority ASIN (top 20 by revenue)
- MYE tests on at least 3-5 priority ASINs
- Backend keyword audit + refresh
- Competitor listing audit (what's changed; what to respond to)
Annually
- Full catalog quality review (every ASIN)
- A+ Content refresh on all active listings
- Image library refresh (new photoshoot if needed)
- Backend keyword + search trend review
Production checklist
For an enterprise Amazon listing programme at Rs 25 crore+ ARR:
- Title formula documented + applied per category
- Bullet template per category + customised per product
- A+ Content active on 90+ percent of revenue-generating ASINs
- 7-image standard met per priority ASIN
- Backend keywords populated + refreshed quarterly
- MYE tests running on 3-5 priority ASINs at any time
- Listing-quality monitoring active per ASIN
- Buy Box + F-Assured / Prime monitoring per ASIN
- Monthly + quarterly + annual operating cadence with named owners
- Competitor listing audit cadence
- Brand Registry maintained + counterfeit-defence workflow active
References + linked context
- Dcrayons glossary: a-plus-content, brand-registry, buy-box, tacos
- Dcrayons reference architectures: Enterprise Amazon Presence, Enterprise Amazon Ads Portfolio Strategy, Enterprise Amazon Influencer Programme
Listing optimisation is the conversion floor under everything else on Amazon. If your programme has hit a listing-quality plateau, an A+ Content gap, or a testing-methodology question, reach out via the contact form for a 30-minute review.
