Amazon Video Ads: Sponsored Brands Video, DSP Specs, and UGC-Style Creative That Converts

Amazon is no longer just a search engine for products. It is a full-funnel advertising platform where video has become one of the highest-intent placements in digital media. Shoppers on Amazon are already in a buying mindset, which makes video ads unusually effective compared to social platforms where users are scrolling for entertainment.

For DTC brands and performance marketers, two formats matter most: Sponsored Brands Video and Amazon DSP video. Understanding how each one works, what technical specs they require, and what creative style actually converts is essential before spending a dollar on either.

The Two Core Amazon Video Ad Formats

Amazon video advertising splits into two distinct systems with different targeting logic, pricing models, and creative requirements. Choosing the wrong one for your objective is one of the most common budget mistakes brands make early in their Amazon advertising journey.

Sponsored Brands Video

Sponsored Brands Video (SBV) is a self-serve, keyword-triggered format available through Seller Central and Vendor Central. Ads appear in search results, typically in the middle of the organic listings page, and autoplay silently on scroll. The ad links directly to a product detail page or a custom brand store.

Key characteristics of Sponsored Brands Video:

• Bidding model: cost-per-click, same auction logic as Sponsored Products
• Targeting: keyword-based (exact, phrase, broad match) and product targeting
• Placement: Amazon search results pages, both desktop and mobile
• Minimum daily budget: $1, though meaningful data requires higher spend
• Attribution window: 14 days, click-based

Because the format autoplays without sound, the first two seconds of your creative are doing almost all the work. A viewer reads what is on screen before they decide to tap the product image or scroll past. This rewards text-heavy, benefit-forward visuals rather than cinematic brand storytelling.

Amazon DSP Video

Amazon DSP (Demand-Side Platform) is a managed or self-serve programmatic buying system that places video ads both on Amazon properties, including Fire TV, Twitch, IMDb, and Amazon-owned apps, and across third-party publisher sites. Unlike SBV, DSP does not require a product to be sold on Amazon, though most DTC advertisers use it for retargeting or upper-funnel reach.

Key characteristics of Amazon DSP Video:

• Pricing model: CPM (cost per thousand impressions), not CPC
• Targeting: first-party Amazon audience data covering purchase history, browsing signals, and lifestyle segments
• Placements: streaming video (OTT/CTV), online video (OLV), and display
• Minimum spend: managed accounts carry meaningful monthly minimums; self-serve thresholds vary by region
• Attribution: Amazon's closed-loop system ties DSP exposure to on-Amazon purchases, including view-through conversions

DSP is typically used once a brand has an established Amazon presence and wants to retarget shoppers who viewed a product detail page without purchasing, or to reach lookalike audiences built from existing customer data.

Amazon Video Ad Specs at a Glance

Getting specs wrong means your creative gets rejected before it ever runs. The table below covers the most common requirements for both formats as of mid-2026. Always verify against Amazon's current advertising console, as specs update periodically.

Spec	Sponsored Brands Video	Amazon DSP (OLV)	Amazon DSP (OTT/CTV)
Aspect ratio	16:9	16:9 or 1:1	16:9
Resolution	1920x1080 minimum	1280x720 minimum	1920x1080 minimum
Duration	6 to 45 seconds	6 to 30 seconds	15 or 30 seconds
File format	MP4 or MOV	MP4 or MOV	MP4 or MOV
Max file size	500 MB	1 GB	1 GB
Audio	Required (muted on autoplay)	Required	Required
Captions	Strongly recommended	Optional	Optional
Frame rate	23.976, 24, 25, 29.97, or 30 fps	Same	Same
Bitrate	1 Mbps minimum	2 Mbps minimum	4 Mbps minimum

Two things stand out from this table. First, SBV autoplays muted, so captions are not optional in practice even if Amazon marks them as recommended. Viewers will not unmute unless your visual storytelling gives them a clear reason to. Second, the 45-second cap for SBV is rarely the right call. Most high-performing SBV creative runs between 15 and 30 seconds, with 15-second cuts consistently delivering stronger completion rates.

Why UGC-Style Creative Outperforms Polished Video on Amazon

The conventional assumption is that Amazon shoppers respond to clean product photography and studio video. That assumption is increasingly wrong. Authentic, creator-style video content generates higher engagement and purchase intent than traditional brand creative, a pattern that carries directly into Amazon's ad placements.

According to short-form video statistics for 2026, viewers retain significantly more information from short video than from static images, and authenticity signals in video, such as handheld camera movement, natural lighting, and direct-to-camera delivery, correlate with higher trust scores among younger demographics, which now represent a growing share of Amazon's most active shoppers.

On Amazon specifically, UGC-style video ads work for three structural reasons:

1. They mirror organic review content. Amazon shoppers are trained to look for social proof. A video ad that feels like a genuine product review fits naturally into that mental frame.
2. They front-load benefits. UGC scripts lead with a specific problem or result rather than a brand statement, which aligns with how Amazon search intent works. Someone searching "wireless earbuds under $50" wants to know immediately whether your product solves their problem.
3. They hold attention without sound. Creator-style videos with on-screen text, reaction shots, and product demonstrations communicate value even when muted, which is the default state for SBV.

The challenge has always been production volume. Testing creative on Amazon requires multiple hooks and multiple cut lengths. Running five versions of the same 15-second SBV ad with different opening hooks is standard practice among sophisticated advertisers, and tracking hook rate benchmarks helps you understand which creative elements are actually moving the needle before you scale spend.

Scaling Amazon Video Creative With AI

Traditional UGC production does not scale cheaply. A single creator collaboration typically costs several hundred dollars per deliverable, and turnaround time for revisions adds days or weeks to a testing cycle. For DTC brands running aggressive Amazon advertising programs, that pace creates a bottleneck between creative insight and creative output.

AI UGC generation closes that gap. Tools that take a product photo and a short brief and output multiple video ads with different hooks, AI voiceover, captions, and music allow brands to produce a full batch of creative for a single campaign in hours rather than weeks.

For Amazon specifically, this workflow has concrete advantages:

• Hook variation at scale. Instead of paying for three creator takes, generate eight different opening hooks from the same product brief and test them simultaneously against the same keyword set.
• Rapid iteration on duration. A 30-second master can be cut to 15 and 6-second versions, each with caption timing adjusted to match.
• Consistent caption quality. AI-generated captions are synchronized and formatted correctly from the start, removing manual QA before upload.
• Lower cost per creative. Pay-as-you-go credit models mean you spend in proportion to what you produce, not on a retainer that idles between campaign launches.

Platforms like Sepia use models including Seedance, Veo, Kling, and ElevenLabs to generate footage, voice, and audio from a product photo and brief, producing ready-to-post video without a shoot. The output maps directly to Amazon's SBV spec requirements when exported at the correct resolution and duration. For benchmarking how many variations to generate per campaign, ad creative volume benchmarks provide useful reference points based on what competitive Amazon advertisers are actually testing.

FAQ

What is the difference between Sponsored Brands Video and Sponsored Products?

Sponsored Products is a static or carousel image format that appears in search results and product detail pages, using cost-per-click bidding. Sponsored Brands Video is a video format that appears in search results, also on a cost-per-click model. SBV gives you more screen real estate and the ability to communicate multiple benefits before a click, while Sponsored Products is simpler to set up and typically carries lower CPCs. Many Amazon advertisers run both simultaneously, using SBV for new-to-brand acquisition and Sponsored Products for high-intent, bottom-funnel targeting.

Do Amazon video ads need to be 16:9, or can I use vertical video?

Sponsored Brands Video requires a 16:9 aspect ratio. Amazon DSP Online Video accepts 1:1 square formats in some placements, but true 9:16 vertical video is not currently supported for Amazon's in-search placements. If you produce 9:16 content for TikTok or Meta, you will need to reformat it to 16:9 for SBV use. Many AI video generation tools let you export the same underlying footage in multiple aspect ratios to cover both placements from one production pass.

Can I run Amazon video ads without selling on Amazon?

Yes, but only through Amazon DSP. DSP can drive traffic to any destination, including a Shopify store or a direct-to-consumer site, though you lose access to Amazon's closed-loop purchase attribution when traffic exits the platform. If you are selling exclusively DTC and are not listed on Amazon, DSP is your only entry point into Amazon's advertising ecosystem. Sponsored Brands Video requires an active product listing in Seller Central or Vendor Central.

How many creative variations should I test for Sponsored Brands Video?

Experienced Amazon advertisers typically launch at least three to five creative variations per ad group, isolating one variable at a time, usually the opening hook. Once one variation shows a consistently lower cost-per-click and higher click-through rate over seven to fourteen days, you pause underperformers and shift budget to the winner before introducing a new challenger. Running too many variations simultaneously with insufficient budget dilutes statistical significance and slows down the learning cycle considerably.