Key Takeaways and Introduction
- CTV out-delivered linear TV. Streaming and CTV buys in Measured’s panel generated roughly 75% higher incremental ROAS than linear TV buys over the same period.
- Prospecting beat retargeting on CTV by roughly 1.9x on measured incremental ROAS.
- YouTube CTV pooled at 1.6x the CTV category benchmark, the strongest combination of scale and measured incrementality in the panel.
- Managed and publisher-direct CTV programs led the category at roughly 1.8x the benchmark.
- Every platform’s own attribution over-credited itself versus incrementality-calibrated results, which is why platform dashboards are not a substitute for geo testing.
Connected TV is now a permanent line item in most enterprise media plans, but most “best CTV platform” lists are built on reviews, reach claims, and vendor marketing. This one is different.
Measured runs incrementality-calibrated measurement for hundreds of consumer brands, which means we observe what CTV platforms actually deliver in causal, incremental revenue, not what their attribution dashboards claim. For this analysis, we looked at the trailing 52 weeks (July 2025 to July 2026) of CTV and streaming performance across a panel of enterprise consumer brands representing more than $25M in measured CTV and streaming spend, all measured on a consistent incremental ROAS (iROAS) basis via Measured’s Cross-Channel Dashboard and calibrated with geo incrementality tests.
Where we have sufficient panel volume, each platform below carries a Measured Incrementality Index: its pooled iROAS relative to the CTV category benchmark in our panel (1.0x = category average). Platforms without enough measured volume in our panel are assessed on scale, targeting, and measurement capability, and marked accordingly.
This is a list, not a leaderboard. The platforms are ordered to help you read it, not to declare a winner, because the right CTV platform is a function of your audience, your budget, and whether you are willing to prove performance with a holdout test.
What the Data Says: CTV Incrementality in 2026
Four findings from the panel that should shape any 2026 CTV plan:
- CTV out-delivers linear. Streaming and CTV buys in our panel generated roughly 75% higher incremental ROAS than linear TV buys measured over the same period.
- Prospecting beats retargeting on CTV. Dedicated CTV prospecting lines delivered roughly 1.9x the iROAS of CTV retargeting lines. On a big screen with no click path, reaching new households is where the incremental money is. Retargeting on CTV frequently pays to re-reach buyers who would have converted anyway.
- Managed and publisher-direct CTV programs led the category at roughly 1.8x the category benchmark, ahead of most self-serve and DSP-routed buying in our panel.
- Platform-reported ROAS remains inflated. Across the panel, CTV platforms’ own attribution consistently over-credited themselves versus incrementality-calibrated results, which is exactly why a list based on measured lift, rather than platform dashboards, matters.
How We Evaluated These Platforms
Our assessment blends four factors: (1) Measured Incrementality Index where panel volume allows, (2) reach and scale, (3) targeting and measurement quality, and (4) buying flexibility and minimums. Indexes are relative multiples only. No client-level data is disclosed.
Overview
YouTube is now the single most-watched streaming source on U.S. televisions, and in our panel it is also the most consistently incremental large-scale CTV buy. Pooled across the largest spend base of any platform we measured, YouTube CTV and video delivered 1.6x the category benchmark iROAS, with top-performing brand programs reaching several multiples of that. Buying via Google Ads (including Demand Gen) keeps it accessible, and DV360 unlocks enterprise controls.
Pros
- Strongest combination of scale and measured incrementality in our panel.
- Top-decile programs in our data ran 3x to 4x the category benchmark.
- Self-serve through Google Ads, with no publisher minimums.
- Granular audience and creative iteration unavailable in most walled-garden CTV.
- Easy to geo-test, so it cleanly supports incrementality experiments.
Cons
- Mixed inventory: TV-screen share must be actively managed versus mobile and desktop video.
- Broader DV360-routed video (non-YouTube CTV) pooled below the category benchmark in our panel. The premium is in YouTube CTV itself, not DSP video generally.
- Platform-reported conversions materially over-credit, so measure independently.
Overview
Amazon’s opt-out ad model made Prime Video an instant giant, with roughly 315 million monthly ad-supported viewers globally as of late 2025. Combined with Fire TV, legacy Freevee inventory, Thursday Night Football, and the Amazon DSP’s retail data, it is the strongest closed-loop commerce story in CTV.
Pros
- Largest ad-supported streaming audience in the market.
- Retail purchase data for targeting and closed-loop reporting via Amazon Marketing Cloud.
- Competitive CPMs relative to premium peers.
Cons
- Attribution is Amazon-centric, so off-Amazon incrementality requires independent measurement.
- DSP learning curve and managed-service thresholds for full capability.
Overview
Disney’s ad-supported footprint across Disney+, Hulu, and ESPN+ reached 164 million monthly active users, and the Disney+ and Hulu app unification in 2026 consolidates that into one of the two or three most valuable premium pools in streaming. Notably, our panel finding that publisher-direct and managed CTV programs led the category at roughly 1.8x benchmark is consistent with what brands buying premium publishers directly are seeing.
Pros
- Premium, fully brand-safe long-form content with strong completion rates.
- Self-serve entry via Disney Campaign Manager (low minimums) up to full upfront deals.
- Robust clean-room measurement options.
Cons
- CPMs at the top of the market, so efficiency depends on rigorous incrementality measurement.
- Frequency management across the merged app footprint is still maturing.
Overview
Roku remains the largest CTV device platform in the U.S., at roughly 37% device share, and its OS position gives it viewing data across every app on the device. The Roku Channel plus network inventory, self-serve Roku Ads Manager ($500 minimums), and a Shopify and commerce integration make it the reach anchor of most CTV plans.
Pros
- Unmatched U.S. household reach and OS-level ACR data.
- Low self-serve minimums, so it is fast to test.
- Strong sports and FAST-channel inventory growth.
Cons
- The Roku Channel’s owned content is less premium than Disney or Netflix environments.
- Measurement leans on Roku’s own graph, so validate independently.
Overview
Netflix’s ad tier passed 94 million monthly active users in mid-2025 and continues to grow steeply, with the majority of new sign-ups now choosing ads. Its in-house ad platform launch cut CPMs from the eye-watering launch era, and live sports and events are adding reach moments linear can no longer deliver.
Pros
- The most engaged premium audience in streaming, with strong incremental reach versus other CTV.
- CPMs down substantially from launch, with programmatic access via major DSPs.
- Culture-moment live events such as sports and specials.
Cons
- Still scaling ad load and measurement partnerships, with less performance tooling than peers.
- Limited self-serve sophistication versus Google and Amazon.
Overview
Universal Ads is Comcast’s self-service platform for premium TV, built on FreeWheel’s ad technology and launched in early 2025 with a cross-publisher roster including NBCUniversal, Fox, Paramount, Warner Bros. Discovery, AMC Networks, A+E, Roku, DirecTV, TelevisaUnivision, and Xumo. The premise is straightforward and, in our view, the most interesting structural development in CTV buying: get social-platform buying ergonomics pointed at publisher-direct premium inventory instead of open-exchange supply.
Two 2026 developments make it materially more relevant to enterprise plans than it was at launch. The Universal Audience Network now spans more than 20 premium publishers, and in April 2026 Universal Ads added Comcast-owned linear inventory to the same self-serve workflow, making it one of the few places a marketer can plan converged linear and streaming in one interface. Comcast reports the platform can reach up to 90% of U.S. households through its national footprint. It also expanded internationally, launching in the UK in June 2026 with ITV, Sky, and Channel 4.
Our panel finding that publisher-direct and managed CTV programs pooled at roughly 1.8x the category benchmark is the relevant context here: Universal Ads is a self-serve route into exactly the kind of inventory that performed best in our data. We do not yet have enough measured volume in the panel to publish an index for it, which makes it a strong candidate for a properly designed geo test rather than an assumption in either direction.
Pros
- Publisher-direct premium supply across 20-plus networks from a single self-serve interface, without a DSP seat or agency intermediary.
- Converged linear and streaming buying in one platform, which is rare outside of managed convergent-TV vendors.
- No upfront commitments or contracts, so the cost of a real test is low.
- Universal Ads Pixel, a Marketing API, and AI-assisted creative generation lower the two biggest practical barriers to TV: measurement plumbing and production cost.
- Publisher-direct inventory sidesteps the supply-path quality problems that dragged open-exchange CTV in our data.
Cons
- No Measured index yet, so treat performance claims as unvalidated until you geo-test it.
- Pixel-based conversion reporting is correlational, and like every platform in this list it will over-credit relative to incrementality testing.
- Younger platform: audience targeting depth and third-party measurement integrations are still maturing versus Google and Amazon.
- Inventory access varies by publisher and flight, so plan reach with the account team rather than assuming full network availability.
Overview
Keynes Digital is a managed-service programmatic CTV shop popular with DTC brands. In our panel its programs pooled just under the category benchmark, but with a sharp internal split: its retargeting and CRM-based lines were among the very best CTV performance we measured anywhere, while broad prospecting lines lagged. With the right audience strategy, Keynes can be exceptional.
Pros
- One of the highest measured-iROAS individual programs in our panel.
- Full-transparency managed service with DTC-native reporting.
- Fast setup, with no self-serve learning curve.
Cons
- Performance is highly strategy-dependent, and broad prospecting pooled well below benchmark in our data.
- Managed-only, so less control than self-serve platforms.
Sources
Overview
Tatari is the best-known convergent TV buying platform, spanning linear and streaming with impression-level measurement. In our panel, Tatari streaming programs pooled below the category benchmark, but with enormous variance: one brand’s Tatari streaming line was among the top CTV programs we measured, while others hovered near breakeven. Its linear buys measured close to parity with its streaming in efficiency for some brands, which is unusual and speaks to strong linear buying.
Pros
- Converged linear and streaming buying and measurement in one platform.
- Impression-level data and genuinely analytical account teams.
- Best-case programs in our panel were top-decile.
Cons
- Pooled streaming performance below category benchmark in our panel, with results varying widely by brand and audience strategy.
- Platform-reported lift methodology still needs independent validation.
Sources
Overview
MNTN (Performance TV) made self-serve, direct-response CTV mainstream, with automated optimization toward site visits and conversions, Verified Visits attribution, and built-in creative via QuickFrame. It is the fastest on-ramp to CTV for teams without an agency, provided you treat its native attribution as directional rather than causal.
Pros
- True self-serve with performance-marketer ergonomics.
- Low barriers: no large minimums and fast creative turnaround.
- Broad premium network inventory access.
Cons
- Verified Visits is a correlational attribution model, so it will over-credit versus incrementality testing.
- Less inventory control and transparency than a DSP.
Overview
The Trade Desk remains the leading independent DSP for CTV, with direct publisher integrations (Disney, NBCU, Paramount, Netflix), UID2 identity, and Kokai’s AI-driven optimization. A caution from our data: DSP-routed video in general (measured via Google’s DV360 in our panel) pooled well below the category benchmark. Open programmatic CTV demands careful supply-path and audience curation to hit publisher-direct levels of incrementality.
Pros
- Broadest independent access to premium CTV supply.
- UID2 and retail data partnerships for addressable targeting.
- Cross-channel dedupe and frequency control.
Cons
- Requires seat access, expertise, and typically agency support.
- Open-exchange CTV quality varies, so supply-path discipline is essential.
Sources
Overview
tvScientific pitches CTV as a pay-for-outcomes channel, with attribution tying exposures to conversions. In our panel, the tvScientific lines we observed were small-budget tests that did not produce measurable incremental sales in incrementality-calibrated modeling. That is a sample-size caveat more than a verdict, but it is also a reminder that CTV generally needs meaningful sustained spend to move a demand needle you can measure.
Pros
- Outcome-based pricing options shift risk toward the platform.
- Performance-marketing UX and granular attribution reporting.
Cons
- Sub-scale budgets may not generate measurable incremental lift, so commit to a real test budget.
- Native attribution over-credits versus causal measurement, so validate with geo testing.
Sources
Also Worth a Look
- Vibe.co (vibe.co): $50/day self-serve CTV for smaller advertisers.
- StackAdapt (stackadapt.com): multichannel DSP with an accessible CTV offering.
- Paramount Ads Manager (paramount.com): self-serve entry into CBS and Paramount+ premium inventory.
- Spectrum Reach (spectrumreach.com): local and MVPD streaming reach, which pooled roughly 0.3x the category benchmark in our panel. Best for geo-targeted local plays rather than national efficiency.
- Samsung Ads and LG Ad Solutions: OS-level ACR data on the two biggest smart-TV brands.
How to Actually Pick (and prove) Your CTV Platform
Every platform above will hand you a dashboard that says it is working. The only way to know is to measure incrementality: run geo holdout tests on each CTV partner, calibrate your media mix model with the results, and reallocate to the platforms that clear your breakeven iROAS. That is precisely the methodology behind the indexes in this article.
If CTV is one of several expansion bets you are weighing, our companion piece covers the full set: Top Media Channels to Test When Google and Meta Are Tapped Out.
Book a demo to see how Measured brands measure and optimize CTV incrementality, or download The Future of Media Mix Modeling.
Frequently Asked Questions
How do I choose a CTV advertising platform?
Start with audience fit and buying model rather than with a leaderboard. Decide whether you need maximum household reach, premium brand-safe content, retail purchase signal, or self-serve control, then shortlist two or three platforms that fit and size a test budget large enough to detect lift. In Measured’s panel, the same platform indexed above benchmark at one brand and below it at another, so the platform is rarely the variable that decides the outcome. Objective, audience strategy, and budget discipline are.
Is CTV advertising more effective than linear TV?
In Measured’s panel over the trailing 52 weeks, CTV and streaming buys generated roughly 75% higher incremental ROAS than linear TV buys. Linear still adds reach among older demographics, but on causal efficiency CTV led.
Should I run prospecting or retargeting on CTV?
Prospecting. In Measured’s data, CTV prospecting lines delivered roughly 1.9x the incremental ROAS of CTV retargeting lines. Retargeting on CTV often re-reaches households that would have converted anyway.
How do I measure CTV advertising incrementality?
Run geo holdout experiments, meaning matched markets where CTV is dark versus live, and use the measured lift to calibrate an always-on model. Platform attribution (pixels, IP matching, “verified visits”) consistently over-credits CTV relative to causal measurement.
How much should I budget for a CTV test?
Enough to move a measurable needle. In Measured’s panel, small-budget CTV tests frequently failed to register statistically measurable incremental sales. Plan a sustained, geo-testable flight sized to your baseline sales volatility rather than a token spend.
Methodology: Platform indexes reflect pooled incremental ROAS (incremental sales divided by media spend) for CTV, streaming, and online-video tactics measured in Measured’s Cross-Channel Dashboard, incrementality-calibrated, across a panel of enterprise consumer brands representing more than $25M in CTV and streaming spend over the trailing 52 weeks ending July 2026, expressed relative to the pooled category benchmark (1.0x). Linear TV lines are excluded from the CTV benchmark. No client-identifiable data is disclosed. Platforms shown without an index lacked sufficient measured volume in this panel for a reliable figure and are assessed on scale, targeting, and measurement capability. Platform order does not represent a ranking.