Connected TV as Commerce Media: Turning Streaming Audiences into Revenue

Most brands still treat connected TV as a branding play. They are leaving measurable revenue on the table.

Connected TV ad spend crossed $30 billion in the US in 2025, yet the majority of that budget is still measured by reach and frequency — the same proxies performance marketers abandoned a decade ago on display. That's changing fast. Retailers, publishers, and ad-tech platforms are rebuilding CTV infrastructure around purchase-intent signals, closed-loop attribution, and outcomes that can actually be audited. For brands that think in terms of incremental revenue, the opportunity is real. For those that don't, CTV will remain an expensive awareness channel that never quite justifies its line item.

CTV Has an Attribution Problem — And It's Being Solved

The reason connected TV spent so long as a brand-only medium is straightforward: you couldn't close the loop. A viewer saw your ad, maybe visited your site, maybe bought something — but connecting those dots across device IDs, publisher ecosystems, and walled gardens was technically painful and commercially unappealing to the parties holding the data.

Clean room partnerships between streaming publishers and retail data providers are changing this. Amazon, Roku, Peacock, and Hulu now offer attribution solutions that tie ad exposure directly to purchase behavior — not modeled views, not click proxies, but verifiable transactions linked to logged-in account data. Platforms like DV360 can increasingly serve CTV alongside open-web formats and attribute across a unified path to purchase.

This matters because without attribution, CTV budget is allocated by gut and justified by brand equity arguments that can't be falsified. With proper measurement infrastructure, you can run incrementality tests, set CPA targets, and hold streaming spend to the same standard as paid search.

Purchase-Intent Signals Are Moving into Streaming

Commerce media's core premise is reaching buyers who are already in a purchase mindset — not interrupting someone who just wants to watch a show. CTV has historically failed this test because most intent signals live on search, open-web editorial, and social platforms. That's shifting as streaming services layer commerce-context targeting into their ad products:

• Amazon Prime Video targets based on Amazon purchase and browse history, effectively importing search-level intent into a lean-back format
• Hulu and Peacock are running shoppable ad units with direct checkout or QR-code-to-cart experiences, shortening the path from exposure to conversion
• YouTube CTV inventory now supports TrueView for Action and Performance Max campaigns, bringing Google's intent graph into the living room screen
• First-party retail data onboarding lets brands suppress existing customers, target lapsed buyers, or reach category-intent audiences based on recent purchase signals

None of this makes CTV a lower-funnel channel in the same way search is. But it meaningfully raises the floor on audience quality relative to traditional linear TV — and that changes how performance-focused brands should think about the format.

The question isn't whether your CTV campaign drove purchases. It's whether it drove purchases that wouldn't have happened without the ad. That's an incrementality question, not an attribution one — and the two get conflated constantly.

The Incrementality Trap Is Real

Attribution capabilities can create a new version of an old problem. Brands excited by closed-loop reporting start crediting CTV for purchases that would have happened anyway — organic demand that got a streaming ad stapled to it. The same confusion that makes last-click affiliate attribution misleading exists in CTV commerce media, just with better-looking dashboards.

Running holdout tests — suppressing CTV delivery to a matched control group and measuring the purchase rate delta — is the only reliable way to separate incremental lift from captured demand. Brands that skip this step will report strong ROAS figures and quietly cannibalize their own organic baseline. Brands that build it into their testing cadence can prove what CTV actually contributes and scale with evidence behind them.

Where CTV Fits in a Commerce Media Mix

CTV is not a replacement for high-intent, lower-funnel commerce media. It's a reinforcement layer — most effective for mid-to-upper funnel audiences who are in-market but haven't yet converted. In a well-structured commerce media stack:

• Open-web commerce media (contextual, DV360): reaches buyers in high-intent editorial environments
• Search and shopping: captures explicit, bottom-of-funnel intent
• Social commerce (Meta, TikTok): drives discovery and dynamic remarketing
• CTV: reinforces messaging at scale for in-market audiences, particularly effective for fashion, beauty, and home where consideration cycles are longer

A shopper who has visited your site, encountered your brand in a contextual commerce placement, and then seen your ad while streaming premium content is meaningfully closer to converting than one reached by only one of those touchpoints. CTV doesn't replace the others — it compounds them.

Commerce media has always been about capturing real purchase intent rather than renting eyeballs and hoping something sticks. Connected TV is becoming the first streaming-native channel that can credibly make that claim. The measurement infrastructure is still maturing, but brands that start testing with proper incrementality controls now will carry a meaningful learning advantage into a channel that's growing fast and getting more competitive by the quarter.