Comcast Advertising on October 1, 2026 published findings from its Multiscreen Performance Report, stating that campaigns combining traditional television with streaming delivered twice the incremental return on ad spend of campaigns relying on one tactic, and 2.4 times as many website visits, in first-half 2026 data measured with Innovid, Clarivoy and Mastercard.

In Short

Comcast, which sells TV and streaming ads, studied its campaigns from January to June 2026 and says advertisers who bought both regular TV and streaming got about twice as much extra sales value per ad dollar as those who relied on a single approach. That matters to anyone splitting a video budget, but the figures come from the company selling the ads, working with outside measurement firms, and the release does not explain how most of them were calculated. Two of the attention numbers match a streaming-only lab test run for a Comcast-owned company, even though the release presents them as TV results. What changes is the pitch: Comcast's yearly viewing report has become a performance report, built to argue for sales and site visits rather than audience size.

A viewing report becomes a performance report

Comcast Advertising, the advertising division of Comcast, distributed the material through Business Wire from New York at 8:00 AM Eastern Daylight Time on October 1. According to Comcast Advertising, the Multiscreen Performance Report is an evolution of its flagship Multiscreen TV Advertising Report. Its stated purpose is to better prove how advertising affects outcomes across the buyer journey, examining four stages - attention, reach, consideration and website activity - for advertisers "of all sizes."

The renaming reflects a change of subject. Data from the second-half 2025 edition, cited by the Video Advertising Bureau in a June 2026 consumption report, put 73% of traditional TV viewing outside primetime and 65% of it on cable, with households spreading their viewing across an average of 26 networks. Those were statements about audiences; the 2026 edition is about return on spend, site traffic and sales.

According to Comcast Advertising, the metrics cover the first half of 2026. They combine aggregated television and streaming campaign data from billions of impressions with outcome measurement from three attribution partners - Innovid, Clarivoy and Mastercard - plus a viewer study conducted with MediaScience on how attention shapes brand memory. The release gives no count of campaigns or advertisers, no categories, no definition of a campaign "tactic," and no indication of which partner produced which result. The figures below are as Comcast Advertising stated them; the fuller report linked from the release was not reviewed.

Comcast Advertising frames the results as building blocks: attention to TV ads drives memory, after which television builds awareness at scale, sharpens consideration and moves consumers to act.

Dawn Lee Williamson, Chief Revenue Officer for Media Solutions at Comcast Advertising, set out the commercial argument. "Advertisers are under more pressure than ever to prove their media dollars are performing, which is often measured only during conversion. In reality, it starts earlier: when quality access meets the right data to capture true attention," she said. Pairing TV and streaming, she added, "helps brands capture attention and stay top-of-mind to help drive the measurable results they need."

Twice the return - measured against what?

The headline result is a ratio. According to Comcast Advertising, multiscreen TV campaigns delivered a 2X incremental return on ad spend compared with campaigns using only one tactic. The word "incremental" carries weight. A claim about incrementality refers to revenue a campaign caused, judged against what would have happened without it, rather than every sale that followed an exposure. Attributed return, by contrast, also counts purchases that would have occurred anyway.

What the release does not supply is the level. Twice as much could mean $2.00 against $1.00 per dollar of media, or 60 cents against 30 cents. Twice the return, and no figure for the return itself! Nor does Comcast Advertising define "one tactic" - a linear-only flight, a streaming-only buy, a single addressable placement - or say whether the two groups were matched for budget, category and flight length. Without that, the effect of combining channels cannot be separated from that of running larger, longer campaigns.

The website figures raise the same question. According to Comcast Advertising, multiscreen TV campaigns generated 2.4X more website visits than campaigns confined to one tactic and were 105% more efficient at delivering them. "Efficiency" is undefined. If it means visits per dollar, the multiscreen campaigns would have spent roughly 17% more, since 2.4 divided by 2.05 is about 1.17 - arithmetic resting on an assumption, but enough to show the two figures are not one claim stated twice.

What Comcast has claimed before

Comcast Advertising has published outcome numbers in the past. When it partnered with Mastercard in June 2025, the company cited a study of campaigns for five home improvement and furnishing brands showing a 3x incremental return on ad spend. That figure described a return level within one vertical; the new 2X describes a ratio between two campaign types. The two are not comparable.

The partner list hints at a sector skew the release leaves unaddressed. When Comcast's Effectv partnered with Clarivoy in January 2025, the automotive analytics firm said it observed more than 160 million automotive shopping sessions a month; whether the website-visit results draw mainly on car campaigns is not stated. Innovid, which merged with Flashtalking under Mediaocean, added purchase-impact data and control-group measurement in April 2026, though the release does not tie those tools to any result.

Reach: 71, 21 and 8

According to Comcast Advertising, traditional TV delivered 71% of unique multiscreen reach and streaming delivered 21%, with 8% overlap between the two. The company argues that advertisers need both to maximise reach.

The three numbers sum to exactly 100. That suggests they are mutually exclusive shares of the combined audience: households reached only by linear, only by streaming, and by both. Read that way, linear touched 79% of all households the campaigns reached and streaming touched 29%. Streaming therefore added 21 points to a linear base of 79 - roughly 27% more households than linear alone would have delivered. Conversely, about 28% of streaming's audience (8 of 29 points) had also seen the campaign on linear, against roughly 10% of the linear audience that also saw it on streaming.

How was the overlap established? Comcast can match exposures at household level within its subscriber footprint, but streaming impressions outside it must be resolved through identity graphs or modelling, and the release does not explain how the reach was deduplicated. A Video Advertising Bureau guide published on August 1, 2026 listed deduplicated reach among five table-stakes metrics for cross-platform evaluation, naming Comcast Advertising as an example provider of set-top box data and warning that such households skew toward higher incomes. Any household seen on both platforms but not recognised as one household is counted twice as exclusive reach, which shrinks the apparent overlap.

Sports streaming and the calendar

The second reach claim concerns live sport. According to Comcast Advertising, sports streaming campaigns driven by events including the Milano Cortina Winter Olympics, March Madness and the 2026 FIFA World Cup recorded 52% unique reach and captured hard-to-reach households 10 times more effectively than TV alone. Neither term is defined. If unique reach means the share of sports-streaming households no other part of a campaign reached, 52% sits far above the 21% exclusive share for streaming generally, though the release does not confirm a shared definition. "Hard-to-reach" is unspecified, and the 10 times multiple has no unit.

Timing shapes the result. The measurement window closed on June 30. The Winter Games ran from February 6 to 22. The World Cup opened on June 11 and its final was played on July 19, so only 20 of the tournament's 39 days fall inside the period; the late knockout rounds and the final do not.

The rights sit with different companies. NBCUniversal carried the Olympics. Telemundo held Spanish-language World Cup rights, with Peacock streaming that coverage, while FOX held English-language rights to 68 matches. March Madness airs on CBS and the TNT Sports networks. Comcast Advertising sells across much of that range, and NBCUniversal used Universal Ads as the exclusive ads manager for its Milan Cortina coverage, yet the release does not break the sports result down by event, seller or platform.

The six months were unusually rich in sports inventory for Comcast's media business. Peacock's advertising revenue per paid subscriber reached $6.53 a month in the first quarter, which carried the Super Bowl and the Winter Olympics, and domestic advertising revenue in the Media segment reached $5,616 million for the half against $2,863 million a year earlier.

Attention figures that match a FreeWheel study

According to Comcast Advertising, TV exposure produced 36% more time spent viewing an ad than social media did, and 2.2X higher brand recall. The release ties this to the MediaScience viewer study without naming it.

Both numbers appear in research published on September 15, 2026 by FreeWheel, which the release itself describes as Comcast Advertising's media and technology arm. That study, conducted with MediaScience across 436 participants, found that streaming viewers gazed at an ad for 71% of its running time against 52% for social viewers, presented as 36% more time watching, and that two streaming exposures produced 53% unaided brand recall against 24% for two social exposures, a 2.2X gap. The figures, the research partner and the comparison with social media all coincide.

Comcast Advertising does not say the two are the same study. If they are, two qualifications follow. FreeWheel's test environment was an unbranded streaming interface showing a 30-minute programme, and no broadcast or cable condition was tested - so a result the release describes as TV exposure would rest on streaming alone. And 2.2X is the two-exposure ratio. After a single exposure, FreeWheel measured 33% recall for streaming against 12% for social, a ratio of about 2.75 that FreeWheel itself rounded to 2.7X. Comcast Advertising's release uses the smaller number.

Neither figure measures sales. Lab-based attention metrics such as gaze share describe the quality of an exposure, and unaided recall describes memory. When the Media Rating Council and the IAB finalised attention measurement guidelines in November 2025, they cautioned against treating attention as a direct measure of campaign outcomes. The release's chain from attention to action is a model of how advertising works, which the attention data alone does not prove.

A 38% sales lift from addressable

According to Comcast Advertising, addressable advertising delivered a 38% incremental sales lift because it was twice as likely to reach the target audience and increased delivery among priority households. Addressable television allows households watching the same programme to receive different commercials, using a distributor's subscriber data.

The sentence contains a causal claim the release does not test: that the lift came from more accurate targeting. The comparison group - unexposed households, national spots or some other baseline - is unstated. Comcast's collaboration with Mastercard measures advertising's impact on geographically aggregated consumer spending, an area-level approach rather than a household match, and the release does not say whether Mastercard data produced the 38% figure.

A comparable Comcast result exists elsewhere. An automotive dealer case study that Comcast Advertising placed in the Video Advertising Bureau's measurement directory in July 2026 reported a 35% lift in new or used conversion rate against a control group, based on a Polk sales matchback covering measurable Comcast households only. That campaign ran across owned designated market areas between October 1 and December 28, 2025, and coordinated addressable with streaming television - closer to a multiscreen result than a purely addressable one.

Scale is the constraint. Addressable delivery depends on distributor households, and Comcast's own pay TV base is contracting: domestic video customers fell by 280,000 in the second quarter of 2026 to 10.668 million, with the base shrinking by roughly a million households a year. Across the wider market, a VAB guide in June 2026 put 92% of pay TV households as addressable-enabled. A 38% lift on a shrinking footprint is an argument about yield per household, not about reach.

Seller, measurer and publisher

The report's structure merits as much scrutiny as its numbers. Comcast Advertising sells the inventory, assembled the campaign data, chose the partners and decided which results to publish. Its partners are independent companies, but none is described as having audited the study as a whole. When VAB added Comcast Advertising to its measurement directory in July 2026, PPC Land noted that the company both sells premium video inventory and offers attribution tools that report on that inventory's performance.

MediaScience works for more than one seller. Its research also sits behind recall claims from FreeWheel and LG Ad Solutions. Guidance issued by the IAB and IAB Europe in November 2025 placed platform-reported incrementality, which rests on platform attribution rather than an independent control group, in the weakest of four evidence tiers. Where Comcast's figures fall on that ladder depends on how its partners built their control groups - and the release does not describe them.

None of this makes the results wrong. Seller-funded studies are the norm in television measurement, and a household-level view of a cable footprint is data few others hold.

Footprint, FreeWheel and the split

According to Comcast Advertising, its Media Solutions team reaches nearly 125 million households, Comcast and non-Comcast, across all 210 DMAs, while Universal Ads lets brands of any size create, buy and measure ads across premium video directly from publishers with no fees. FreeWheel, the media and technology arm, has carried much of the expansion: Comcast made linear TV inventory biddable through programmatic private marketplaces on October 23, 2025, using FreeWheel's Buyer Cloud and opening access to more than 11 billion monthly impressions. The Business Wire page carrying the report also lists an undated FreeWheel release adding third-party attention, ad experience and engagement signals to its Premium Signals Packaging.

Ownership is the larger open question. Comcast said on June 29, 2026 that it would spin off NBCUniversal and Sky in a tax-free transaction expected to take about 12 months, and the announcement did not specify which company would keep FreeWheel and Universal Ads. PPC Land's analysis of the break-up described Comcast as keeping the ad-serving infrastructure while NBCUniversal takes the content. The October 1 boilerplate still describes Comcast Advertising as part of Comcast Corporation "along with NBCUniversal and Sky." Yet the Olympic and Spanish-language World Cup rights behind part of the sports-streaming result belong to NBCUniversal - the side of the company that is leaving.

Why this matters for the marketing community

Television sellers increasingly compete on outcome claims rather than audience size, and the money at stake is growing: CTV advertising is projected to reach 43% of total TV ad spend in 2026, roughly $38 billion. A 2X claim from a company selling both linear and streaming lands directly in the argument over how buyers split video budgets between those formats, and against social platforms.

The report also joins a run of seller-produced research. FreeWheel published its MediaScience study on September 15; on August 28, VAB headlined a 94% ad completion rate drawn from 2023 FreeWheel data. Comcast Advertising's report now carries figures matching the FreeWheel study under a TV label. Each document is internally consistent, and each favours the inventory its sponsor sells.

For planners, the most usable number in the release may be the least promoted one. The 71, 21 and 8 reach split sums cleanly and describes something measurable within one seller's footprint: how little linear and streaming audiences duplicate on Comcast campaigns. The return, visit and sales-lift multiples are a different matter. Until absolute returns, the definition of a tactic and the partner behind each result are published, can they be set against any other seller's claims?

Timeline

Summary

Who: Comcast Advertising, the advertising division of Comcast Corporation, which includes the Media Solutions sales team, the FreeWheel technology business and the Universal Ads platform. Dawn Lee Williamson, Chief Revenue Officer for Media Solutions, is the named spokesperson. Innovid, Clarivoy and Mastercard supplied outcome measurement, and MediaScience conducted the viewer study.

What: Findings from the Multiscreen Performance Report, which succeeds the Multiscreen TV Advertising Report. Comcast Advertising states that multiscreen TV campaigns delivered 2X the incremental return on ad spend and 2.4X the website visits of campaigns using one tactic, that linear supplied 71% of unique multiscreen reach and streaming 21% with 8% overlap, that sports streaming recorded 52% unique reach, that addressable advertising produced a 38% incremental sales lift, and that TV exposure drew 36% more viewing time and 2.2X the brand recall of social media. The release does not disclose absolute returns, sample sizes, the definition of a tactic or which partner measured each result, and its attention figures match a streaming-only FreeWheel study.

When: The findings were published on October 1, 2026 at 8:00 AM Eastern Daylight Time. The underlying metrics cover the first half of 2026, January 1 to June 30.

Where: The release was issued from New York through Business Wire. Comcast Advertising's Media Solutions footprint covers nearly 125 million households across all 210 US designated market areas.

Why: Comcast Advertising positions the report as proof that pairing linear and streaming television drives measurable outcomes across the buyer journey, at a time when CTV budgets are rising and sellers compete on performance claims. For buyers, the significance lies both in the reach data and in the gaps: ratios without baselines, a seller grading its own campaigns, and sports-driven results drawn partly from NBCUniversal rights that are due to leave Comcast in a spin-off expected to take about 12 months.