Sponsored Brands Video vs DSP Video: Two Different Jobs

Where each one actually sits

Sponsored Brands video is keyword-triggered: a shopper searches, the ad appears inside search results or on a product detail page, muted until the viewer opts to unmute, and Amazon bills per click or per 1,000 viewable impressions with no set minimum spend. It's a bottom-of-funnel format wearing a video wrapper — the audience is already searching for something related, and the job is to win the click inside that existing intent.

DSP video is a completely different mechanism. It's audience-based rather than keyword-triggered, reaches viewers on Prime Video, Fire TV Channels, Twitch and third-party publishers rather than inside Amazon's search results, plays sound-on by default under current in-stream guidelines, and bills on CPM rather than per click. There's no search query triggering it — the job is to put a brand in front of a relevant audience before they've searched for anything at all.

The minimum-spend gap between the two says a lot about their roles. Sponsored Brands video accepts daily budgets from $10 with no formal minimum; DSP video carries Amazon's own recommended $10,000 self-serve floor. A search-triggered, click-billed format can run at almost any scale because it only spends when intent already exists; an audience-based, CPM-billed reach format needs real volume to produce a meaningful signal, which is exactly why the two products sit at such different points on the spend spectrum.

Eligibility differs too, in a way worth knowing before assuming both are available to every seller. Sponsored Brands video requires Brand Registry, which most established sellers already hold, but it's a real gate for a brand new to Amazon. DSP video carries no Brand Registry requirement of its own, but it does carry Amazon's recommended $10,000 self-serve spend threshold — a brand can clear one gate and not the other, and the two products are worth checking against a brand's actual current status rather than assumed together.

Why the funnel position, not the fact that both are video, is what matters

Both formats get filed under "video" in a media plan, which invites treating them as interchangeable or competing for the same budget. They're not — one is a search-adjacent, intent-capture tool, and the other is a reach-and-discovery tool. A brand deciding between them is usually asking the wrong question; the more useful question is what each is doing for a specific product at a specific stage, and whether both are needed together rather than one instead of the other.

A worked example: running both, and reconciling them

A brand running $10,000 on Sponsored Brands video and $15,000 on DSP video in the same month is running two funnels that both touch the same shoppers, at different points. Sponsored Brands video's own reporting will show its click and conversion data; DSP video's non-clickable inventory will show mostly completion rate and, if the brand is checking it, branded search lift. The real question — whether the DSP spend actually contributed anything the Sponsored Brands spend wasn't already going to capture — needs Amazon Marketing Cloud to answer, because last-click reporting in each console separately will let both channels claim credit for the same eventual sale.

This is the same double-counting problem that shows up wherever DSP and Sponsored Ads overlap, not something specific to the video formats — the video wrapper just makes it easy to forget, since both look similar on the surface. Treat the two consoles' separate ROAS numbers as each telling a partial, non-additive story, not two figures that should sum to a combined total.

What to do when the two seem to be fighting each other

If Sponsored Brands video's ACOS looks like it's climbing right as a DSP video campaign scales up, that's worth investigating in AMC rather than assuming DSP is cannibalizing search performance — it's equally possible DSP is creating new demand that's showing up as more branded search volume, which raises Sponsored Brands' delivery and cost together without actually hurting efficiency. The two channels showing different last-click numbers isn't evidence they're competing; it's evidence they're measuring different parts of the same customer journey with different tools.

A simpler first check, before reaching for Amazon Marketing Cloud, is timing: did the DSP campaign's scale-up and the Sponsored Brands ACOS shift happen in the same week, or did one clearly precede the other by enough time to suggest cause and effect rather than coincidence. A rough timeline comparison won't prove incrementality on its own, but it's a fast, no-tooling-required first pass before committing to a full AMC investigation.

The common mistake

The mistake is picking one format over the other because both are labeled video, as if a brand only needs one video channel. They do different jobs at different funnel stages, and our own DSP book shows exactly why treating them as substitutes misreads the data: one in five DSP line items produce zero attributed purchases, but between them they burn only 0.7% of the budget — the opposite waste shape from search, where the losing tail is automatic and can run to a third of spend. The two channels don't just do different jobs, they fail differently too, which is one more reason they need separate waste controls, not one shared budget decision.