Incrementality Benchmarks by Channel: What the Data Shows in 2026

• Incrementality benchmarks answer two different questions: how much incremental return a channel produces (Incremental ROAS), and how much of the platform’s reported performance is real (the adjustment factor). Both vary enormously by channel.
• Among video channels, CTV delivers the highest median Incremental ROAS at $1.38, ahead of YouTube ($1.26) and Linear TV ($0.75), based on Measured’s published analysis of 148 video advertisers. Among enterprise brands, CTV’s median rises to $3.18.
• Benchmark ranges are wide by nature: the middle 50% of CTV advertisers span $0.50 to $3.07. A median is a reference point, not a target.
• Adjustment factors follow a consistent pattern by funnel role: demand-capture tactics (branded search, retargeting, email, affiliate) measure far below their platform-reported numbers, while demand-creation tactics (prospecting social, non-brand search, video, audio) often measure at or above them. The specific factors by tactic are part of Measured’s benchmark dataset.
• Benchmarks are for orientation; experiments are for decisions. Use a benchmark to pick which channel to test first, then measure your own number.

Every marketer eventually asks some version of the same question: what is a good incremental ROAS for this channel? It is a fair question with a genuinely useful answer, as long as the answer comes with the honesty most benchmark content skips. Incrementality benchmarks describe distributions, not verdicts, and your brand’s number is knowable only by measurement. This guide shares what Measured’s published data shows for video channels, explains the adjustment factor pattern that repeats across thousands of experiments, and covers how to use benchmarks without being misled by them.

Measured’s benchmarks are built differently from survey-based industry reports. They aggregate incrementality measurement, not platform-reported performance, drawn from 25,000+ incrementality experiments and always-on measurement across 160+ enterprise brands and $35B+ in optimized ad spend. Two metrics matter throughout this guide:

Incremental ROAS (iROAS): revenue that happened because of the media, divided by spend. Unlike platform ROAS, it excludes conversions that would have occurred anyway. This is why incremental figures run well below the numbers platforms report, and why they can be compared across channels while platform figures cannot.

Adjustment Factor: the share of a platform’s reported conversions that measurement shows to be genuinely incremental. An adjustment factor below 100% means the platform over-reported; above 100% means the platform under-reported, usually because the tactic’s impact surfaced through other touchpoints that collected the attribution credit. The adjustment factor is a correction to platform reporting, not a grade on the tactic itself: a heavily over-credited tactic can still earn a solid incremental return once corrected.

The most complete published benchmark set covers video, from Measured’s Q1 2026 analysis of 148 brands investing in video channels.

Three findings worth internalizing:

1. CTV leads the video mix at every scale. It delivers the highest median incremental return among video channels overall, and among enterprise brands its lead widens: roughly 40% above YouTube and nearly double Linear TV.
2. Scale correlates with efficiency, but not through budget. Enterprise brands earn more than double the overall CTV median while spending only modestly more. The gap is structure and measurement discipline, not spend.
3. The range is the finding. A middle-50% spread of $0.50 to $3.07 means the same channel is a top performer for one brand and below breakeven for another. Channel choice matters less than how deliberately the channel is deployed and measured.

The full analysis, including adoption rates and the CTV maturity framework, is in the report CTV is Mainstream. Optimization Just Hasn’t Caught Up. For the measurement methodologies behind these numbers, see How to Measure TV and CTV Incrementality.

Beneath channel-level benchmarks sits the most consistent pattern in incrementality measurement: adjustment factors sort by funnel role, not by channel name. Tactics that sit closest to the conversion intercept credit for sales that were already happening. Tactics that create demand upstream donate their credit to whatever touchpoint converts later. The direction of this pattern repeats across Measured’s experiment library; the size of each tactic’s correction is what the benchmark dataset quantifies.

Here is the pattern, with the specific factors gated:

Marketers who fund channels according to platform-reported performance therefore systematically over-invest in capture and under-invest in creation. This is the central finding explored in Upper Funnel vs. Lower Funnel Marketing and Fixing the Funnel with Incrementality Measurement.

Get the full benchmark data

Measured maintains tactic-level incrementality benchmarks, including adjustment factors and Incremental ROAS by channel and tactic, drawn from live cross-client measurement and refreshed continuously. The specific figures behind the pattern above—matched to brands that resemble yours by conversion type, industry, and order value—are available in a conversation with our team.

Talk to us about your benchmarks

Paid search. The brand versus non-brand split matters more than any cross-channel comparison. Branded search consistently produces the lowest adjustment factors in measurement because the searcher has already decided; non-brand prospecting frequently measures above its platform-reported numbers because its wins convert later through brand search and direct visits. Benchmark your two search halves separately or the blended number will mislead in both directions.

Paid social. The widest tactic-level spread of any channel: within a single platform, prospecting and remarketing routinely sit far apart in measured incrementality while showing near-identical platform-reported ROAS. Social benchmarks are only meaningful at the tactic level.

CTV and TV. The best-documented public benchmark set (table above). Note that view-through over-crediting and genuine incremental value coexist: a streaming TV line can be meaningfully over-credited by view-through attribution and still deliver a positive incremental return after correction.

Display and retargeting. Retargeting shows the strongest over-crediting in any category. Display prospecting behaves differently and should be benchmarked separately.

Email, SMS, and owned channels. Low adjustment factors are structural: the audience is by definition your existing customer base. Known Audience Testing, not platform reporting, is the right measurement instrument here.

Affiliate. Last-click affiliate attribution overstates caused conversions substantially, and affiliate incrementality varies heavily by partner type (loyalty and coupon versus content), which is where testing pays off fastest.

Audio and podcast. Small spend bases and no click path make platform reporting nearly meaningless; measured results frequently exceed reported ones.

Four rules keep benchmark data useful:

1. Benchmarks orient; experiments decide. The median tells you what other brands achieved with different creative, offers, audiences, and spend levels. Your number exists and is measurable; a geo incrementality test retrieves it in four to eight weeks.
2. Compare your gap, not your level. The productive use of a benchmark is prioritization: the channel where your measured (or suspected) performance deviates most from the benchmark distribution is your highest-value test candidate.
3. Respect the range. Any channel with a wide interquartile range is telling you that execution dominates channel selection. Chasing the median of a wide distribution is chasing noise.
4. Never benchmark platform-reported numbers against incremental ones. A platform ROAS and an Incremental ROAS benchmark are different units. Correct your numbers with an adjustment factor first, or the comparison flatters every channel you own.

Measured customers get this comparison built in: the Benchmarks product compares each brand’s tactic-level performance against peer brands over the most recent six months, matched on conversion type and optionally industry, AOV band, and customer mix, so the reference set is brands that actually resemble yours rather than a market-wide average.

See your benchmarks against matched peers

What is a good incremental ROAS?
It depends on the channel, your margins, and your goals, but published medians give orientation: among video channels, CTV’s median Incremental ROAS is $1.38, YouTube’s is $1.26, and Linear TV’s is $0.75 across 148 measured advertisers, with enterprise brands roughly doubling those figures. A “good” number for your brand is one above your breakeven threshold given contribution margin, which is why the benchmark question should always be followed by a measurement plan.

What is a typical adjustment factor by channel?
Adjustment factors sort by funnel role rather than by channel name. Demand-capture tactics such as branded search, retargeting, email, and affiliate measure well below their platform-reported numbers, because platforms claim conversions that would have happened anyway. Demand-creation tactics such as non-brand search prospecting, upper-funnel social, and podcast frequently measure at or above platform-reported numbers. The specific medians and ranges by tactic are part of Measured’s benchmark dataset, available to customers and in a conversation with our team.

Why is my incremental ROAS so much lower than my platform ROAS?
Because platform ROAS includes conversions that would have happened without the ad. Platforms attribute using clicks, views, and device matching, all of which capture correlation rather than causation. The gap between the two numbers is the adjustment factor, and for lower-funnel tactics that gap is routinely large. A lower number that is real beats a higher number that is not, particularly when defending budget to finance.

Are incrementality benchmarks reliable enough to plan against?
They are reliable for orientation and prioritization, not for target-setting. Benchmark distributions are wide because incrementality responds to creative, audience, offer, spend level, and measurement configuration. Use benchmarks to identify which channels to test first and to sanity-check measured results, then plan against your own experimental data.

How often do incrementality benchmarks change?
Measured’s peer benchmarks aggregate the most recent six months of customer data and refresh continuously, so they track market shifts like platform algorithm changes and privacy-driven signal loss. Published industry figures like the video benchmarks above are point-in-time analyses; treat any benchmark older than a year as historical context rather than current reference.

Where can I benchmark my own channels?
Measured customers compare tactic-level performance against matched peer brands directly in the platform’s Benchmarks product. The reliable path for any brand is running a geo holdout on the largest or most-questioned line item and building the benchmark comparison from measured results. To see how your channels compare against the benchmark dataset, talk to our team.