Amazon Full-Funnel Campaigns: the video, display and audio kit
Amazon's unBoxed 2026 Full-Funnel Campaigns span video, display and audio. What a brand should have ready, and which pieces Sume can make: video, images, music.

If you want to be ready for Amazon's Full-Funnel Campaigns, have three kinds of asset on hand: video, display images and audio. Sume can make the video, the images and a music bed; it does not make Amazon's own audio ads or place any of it in Amazon Ads.
The facts about Amazon here come from its unBoxed 2026 announcements page. Sume facts come from its docs.
What did Amazon announce about Full-Funnel Campaigns?
The page describes Full-Funnel Campaigns as AI-optimized campaigns across video, display and audio, and says the beta results were "67% higher long-term return on ad spend and 29% lower cost to acquire new-to-brand customers." Those are Amazon's own beta numbers, and the page lists the product as in beta with no expansion timeline. It also lists DVA+ (one objective, one budget and one campaign across display, video and audio) and an Amazon Ads Agent with a conversational experience in expanded closed beta.
The page gives no upload specs and no launch dates for most items, so check Amazon's own specs before you export anything.
The same page announces an AI creative tool that generates display ads and TV-quality video from product assets, so part of this kit may be made inside Amazon. The reason to keep your own set is control: files you made can be reused on other channels, while files made inside a platform tool may only live there.
Which assets does one campaign across three media need?
The announcement does not list asset requirements, so this table is a planning list, not Amazon's. It maps each medium to the Sume surface that can produce a first version.
| Medium | What to prepare | Sume surface |
|---|---|---|
| Video | Vertical and landscape product clips | POST /v1/videos |
| Display | Product images in several aspect ratios | POST /v1/images |
| Audio | A music bed or sting without vocals | POST /v1/music-router/generate |
| Voice | Narration if your plan includes spoken ads | Not covered here; check the Sume docs |
How do I make one product's set in three calls?
Each call returns a job. Use a distinct Idempotency-Key per asset so a retry does not bill twice, and poll the jobs rather than resubmitting. The video call below uses sume/auto, which lets Sume pick the model; the response never says which one ran.
For display images, the image route accepts aspect ratios such as 1:1, 16:9, 9:16 and 4:5, and auto for the provider's choice. For music, put the length in the prompt, because duration fields are rejected.
curl -X POST https://api.sume.com/v1/videos \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: ffc-video-001" \
-d '{
"model": "sume/auto",
"prompt": "A vertical product clip on a kitchen counter, natural light",
"aspect_ratio": "9:16",
"duration": 5
}'What stays outside Sume?
Sume does not buy or run Amazon campaigns, does not read Amazon Ads Agent, and does not generate Amazon's audio ads. It makes files you then upload through Amazon's own tools. If you use Amazon's built-in creative tools, keep Sume for the channels they do not cover, which another post compares.
Treat Amazon's beta figures as Amazon's claims about its beta. They do not tell you what your own account will see.
Be honest about what you will measure. A beta result in a press page is not a forecast for your account, so decide up front what outcome would make you keep spending on creative, such as cost per new-to-brand customer in your own reporting, and compare against a period before the campaign.
How should I stage the work?
Start with one hero product. Make a 9:16 clip, a 16:9 clip, two or three images and a short music bed, then open Amazon's specs and trim to them. Only after that fits do you extend to the catalog. Because DVA+ is announced without a date on the page I read, the rollout post tells you what to watch for next.
Sources
Related posts
More in Use cases
- Animate a locally generated image with Sume video: first frame
Made a still with a local open-weights model? Host it at a public HTTPS URL, send it as first_frame to POST /v1/videos, and poll. Plus the licence check first.
- Appcues flow images: 5 MB cap, 650 px modals, Sume hero art
Appcues accepts JPG, PNG, SVG and GIF up to 5 MB, with 650 px minimum modal width. Generate onboarding hero images to size with Sume Image Router.
- Appcues can't host video: trim a 20-second onboarding clip
Appcues only embeds video from YouTube, Vimeo or Wistia, never uploads. Trim a long screen recording with Sume video trim, then host the cut elsewhere.
- Does Ask YouTube search include Shorts? What to make findable
YouTube's interactive search compiles long-form videos and Shorts and was Premium-only in the US at launch. What a Short's spoken text can and cannot do.
Written by Sume