Amazon announced Full-Funnel Campaigns on September 29, combining sponsored ads, display, video, and streaming TV inside a single campaign. They are live for US advertisers now. A separate product, DVA+, begins rolling out in late October, consolidating capabilities previously managed through Sponsored Display, Sponsored TV, and programmatic campaigns.

Manual tools and advanced controls remain available. The question is what changes when you stop using them.

What Full-Funnel Campaigns Actually Combine

Amazon's announcement describes one campaign spanning formats that have historically been managed separately, with distinct budgets, bids, and success measures.

That consolidation is the product. Amazon's system decides how much budget flows to a Sponsored Products placement versus a streaming TV impression, based on its own view of where the spend performs best.

Those formats do not carry the same purchase intent. Someone searching a product keyword is closer to buying than someone watching streaming content. Both can generate attributed sales, and both can appear in the same campaign report.

Why Attributed Sales Can Rise While Contribution Falls

This is the specific risk worth modeling before you move budget.

A campaign reporting more attributed sales has not necessarily produced more profit. Upper-funnel formats typically carry longer attribution windows and looser causal links between impression and purchase. When those sit in the same campaign as search placements, the blended report can look strong while the incremental contribution from the added spend is thin.

Run the comparison at the account level rather than inside the campaign. Track total ad spend, total Amazon revenue, and contribution after product cost, fees, returns, and advertising. If a Full-Funnel campaign increases attributed sales while total ACoS and account contribution move the wrong way, the campaign is reallocating credit rather than creating demand.

Keep a control. Leave a comparable product group in your existing campaign structure for the test window, so you have something to measure against.

The Controls to Check Before You Enable It

Before launching, confirm what you can still constrain.

Set a budget ceiling that reflects an experiment, not a channel migration. Check whether you can cap or exclude specific formats, and whether format-level reporting shows spend and attributed sales separately rather than only in aggregate.

Review creative approvals. Video and streaming placements may use assets generated or selected automatically, and your brand team should see what runs before it runs.

Confirm who on your team can enable the campaign type. Standard PPC bid and budget discipline still applies, and an agency or employee building Sponsored Products should not assume a new campaign format falls inside existing authorization.

What DVA+ Changes in Late October

DVA+ consolidates Sponsored Display, Sponsored TV, and programmatic capabilities. If you currently run those as separate line items with separate owners and separate performance targets, that structure is what gets merged.

Document your current baselines before the rollout reaches your account: spend, impressions, attributed sales, and new-to-brand percentage for each format individually. Once they are combined, reconstructing those baselines becomes considerably harder.

Treat Amazon's Performance Claims as Hypotheses

Amazon's stated results for new ad products come from its own testing, usually without published sample sizes, control groups, or category breakdowns.

That does not make them wrong. It makes them untested against your catalog. Pick one product family, set a defined test window and loss limit, and measure contribution per dollar of total ad spend rather than ACoS inside the new campaign.

The useful disclosures to watch for are format-level reporting detail, exclusion controls, and whether Amazon publishes incrementality data rather than attribution data. Until then, the honest position is that Full-Funnel Campaigns reduce your work and reduce your visibility at the same time.