I've optimized 14,000+ hero images and reviewed 50,000+ listings, and the question I get asked more than almost any other isn't about hero design — it's about stack length.
"Should we use 7 images or 9?"
"If we have the slots, should we fill them?"
"Does Amazon penalize stacks that are too short or too long?"
The honest answer is: most brands are filling slots they shouldn't be filling, and a smaller percentage are leaving slots empty that should be working. After pulling data on 412 image stacks across 11 categories from Q4 2025 through Q1 2026, here's what I'm seeing — and what I'd actually recommend.
What "Stack Length" Means and Why It Matters
For this analysis, stack length = total number of product images in the carousel including the main image. So a "7-image stack" means hero + 6 secondary images. Amazon allows up to 9 product images on most listings (some categories support a 10th video slot, separate from this analysis).
Stack length matters for two reasons most brands don't think about:
-
Slot decay is real. Only about 23% of mobile shoppers swipe to image 4. By image 7, you're talking to single-digit percentages of your traffic. Filling slots 8 and 9 with weak content produces zero CVR lift and can produce a small CVR drag if those late slots violate visual consistency or repeat the hero message.
-
Each slot is a bet against the cost of producing it. A 9th image for a low-AOV product isn't free — it's photo budget, design hours, and review time. If the slot doesn't move CVR, it's worse than empty.
The real question isn't "what's the maximum stack length Amazon allows?" It's "what's the optimal stack length for this category, this AOV, this buyer intent?"
The 412-Stack Dataset: What I Pulled
I pulled stacks I'd either built or audited between October 2025 and March 2026 across 11 categories: supplements, baby, kitchen, outdoor/grill, apparel, beauty, pet, electronics accessories, office, home decor, and tools. For each stack I had:
- Stack length (6, 7, 8, or 9 images)
- Pre/post CVR (when we made stack length changes)
- Mobile vs desktop traffic split
- AOV
- Repeat-purchase rate (S&S participation rate as proxy)
- BSR tier within category
Then I segmented results by category and intent type (impulse vs considered).
The Headline Findings
1. The optimal stack length for impulse-buy categories is 6 or 7 images, not 9.
Across impulse categories — beauty under $30, pet treats, low-AOV kitchen — stacks of 6 images outperformed 9-image stacks by an average of +4.1% CVR. The 9-image stacks weren't worse because the late slots were bad; they were worse because brands stretched their best 6 ideas across 9 slots, weakening slots 2–4 in the process.
2. The optimal stack length for considered purchases is 8 or 9, but only when slots 7–9 carry distinct objection-handling content.
For categories with longer consideration windows — strollers, mattresses, high-AOV electronics, supplements with subscription intent — 8 and 9-image stacks beat 6-image stacks by +7.3% CVR on average. But that lift only showed up when slots 7–9 contained:
- A comparison chart against category alternatives
- Detailed dimension/spec graphic
- Use-case scenarios beyond the primary one shown earlier
When slots 7–9 were "more lifestyle photos that look like slot 3," 9-image stacks performed identically to 6-image stacks.
3. Stack length matters less than slot sequencing, but they interact.
A perfectly sequenced 7-image stack outperformed a poorly sequenced 9-image stack in 78% of head-to-head comparisons in my dataset. The brands that get the most out of a 9-image stack are the ones who already nailed sequencing on a 7-image stack. Adding length to a broken stack doesn't fix the stack — it dilutes the hero handoff and stretches weak ideas thinner.
Slot-by-Slot: What Each Position Should Do at Each Length
6-Image Stack (Impulse / Low Consideration)
- Slot 1 — Hero: Polarizing element + product clarity at thumbnail size
- Slot 2 — Top feature/benefit infographic: The single most important reason to buy
- Slot 3 — In-use lifestyle: Aspirational scenario, real human/environment
- Slot 4 — Comparison or scale: Size/scale or "vs cheap alternative"
- Slot 5 — Secondary benefit/feature: Reinforces purchase logic
- Slot 6 — Trust/social proof: Reviews, awards, certifications, ingredient transparency
Skip the brand story slot. Skip the founder photo. For impulse categories, slot 6 should still be selling, not narrating.
7-Image Stack (Most Categories Default Here)
Add a dedicated objection-handling slot at position 5 between secondary benefit and trust. The most common objection: "will it work for me?" — which is solved with a use-case grid (multiple environments/people/scenarios).
This is the stack length I default to for any brand I don't have a strong category-specific reason to lengthen or shorten.
8-Image Stack (Considered Purchases)
Add a comparison chart at position 7 against the brand's actual SERP competitors. Not a generic "ours vs theirs" — name the competing feature dimensions that drive decision-making in the category.
Comparison charts are the highest-CVR-impact secondary slot in my dataset for considered purchases. They consistently outperform additional lifestyle slots by +9–14% CVR lift when added correctly.
9-Image Stack (High AOV / Long Consideration)
Add a specs/dimensions deep-dive at position 8 and a brand story slot at position 9.
Critical: brand story slot has to come last for a reason. Buyers reach slot 9 already convinced or already gone. Slot 9 reinforces — it doesn't sell. Putting brand story at slot 4 (a common mistake I see) burns the slot that should be doing objection-handling.
Categories Where Shorter Stacks Beat Longer
Three categories where I consistently see 6-image stacks beat 9-image stacks:
- Beauty under $25 — quick decisions, primary visual is the product, less is more
- Pet treats and consumables — visual product + ingredient transparency + size
- Office basics and supplies — utility-first, no aspiration needed
For these, the 9-image stacks I audited were cluttered with "lifestyle" photography that looked staged and reduced the believability of the product.
Categories Where Longer Stacks Beat Shorter
Five categories where 8 or 9-image stacks consistently outperformed:
- Mattresses, beds, sleep accessories — research-driven, comparison-heavy
- Strollers, car seats, baby gear — high stakes, parents read every detail
- Outdoor/grill equipment over $200 — assembly, dimensions, capability proof needed
- Supplements with subscription intent — ingredient deep-dives, lab proof, comparison vs category
- Premium kitchen ($150+) — material, dimensions, use-case range
What "Filling Slots Just Because You Have Them" Costs You
The most common pattern I see in audits: brand has a 9-image stack with slots 7, 8, and 9 filled with content that looks identical to slot 3.
The cost isn't neutral. Three things go wrong:
- Visual consistency drag — repeated lifestyle aesthetics late in the stack make the stack feel padded, not premium.
- Mobile shopper decision fatigue — the few shoppers who reach slots 7–9 are the most engaged, and you're giving them no new information.
- Production budget waste — those slots cost real money to produce. If they don't move CVR, you spent budget that should have gone to a better slot 2.
In my audits, killing slots 7–9 of underperforming 9-image stacks (back to a 6-image stack with the strongest content kept) produced an average CVR increase of 3.8% with zero new content created.
How to Decide Stack Length for Your Listing
Walk the listing through these four questions in order:
- Is this an impulse or considered purchase? Decision time under 60 seconds → 6 or 7 images. Multi-day research → 8 or 9.
- Is mobile traffic over 70% of sessions? If yes, lean shorter — slot decay is steeper on mobile.
- Do I have category-specific objections that need dedicated slots? Comparison charts, dimension specs, use-case grids — each one earns a slot.
- Can I produce slots 8 and 9 at the same quality bar as slot 2? If no, don't add them. A weak slot 9 is worse than no slot 9.
What I'd Test This Quarter
Three high-leverage tests:
- If you're at 9 images and CVR is below category benchmark: kill slots 7–9, run as a 6-image stack for 4 weeks, measure.
- If you're at 6 images and you sell a considered-purchase product: add a comparison chart at slot 7, measure for 4 weeks.
- If you have a brand story in slot 3 or 4: move it to slot 7 or 8 and put your strongest objection-handling content in its place.
All three are reversible in 24 hours. None require new photography. All three move CVR in my dataset.
FAQ
Does Amazon's algorithm prefer longer image stacks? There's no evidence Amazon ranks listings higher for stack length itself. The algorithm cares about CVR, BSR, and CTR. A shorter stack that produces higher CVR will outrank a longer stack that doesn't.
What about main image video — does it count as a slot? Main image video is a separate placement above the image carousel. It doesn't replace a slot — it's an additional surface. If your main image video is performing, keep it; it doesn't change stack length math.
Should I match competitors on stack length? No. Match competitors on category convention only when you have no data. Once you have CVR data, optimize for your numbers, not theirs.
How long should I run a stack length test? Minimum 4 weeks for CVR stabilization. Don't read a stack length test in week 1 — image carousel changes affect later-stage CVR more than CTR, and CVR signal needs time.
If you're trying to figure out whether your stack should be 6, 7, or 9 images — and you've never had someone audit it slot by slot — that's where I'd start before producing any new images.