Commentary

DASH-Enhanced Nielsen Drives Reset On TV Viewing Picture

by Dave Morgan , Featured Contributor, Yesterday

The media industry has wanted better and more accurate viewing measurement data from Nielsen for decades -- and streaming, audience fragmentation and the “alt measurement movement” of the past few years only amplified those wishes.

Yesterday, Nielsen announced it has now integrated data from DASH, a syndicated baseline study conducted by the ARF (Advertising Research Foundation) in partnership with a polling firm at University of Chicago and accredited by the Media Rating Council for TV universe estimation that measures how U.S. homes connect to televisions, streaming platforms, and mobile video.

Finally, Nielsen can put to bed questions about whether its TV viewing panel and Big Data sets conform to the U.S. population.

However, what I found most interesting in digging into the announcement is some of the non-intuitive insights about U.S. TV viewing that the changes bring to light. They include:

Nielsen’s panel ratings numbers for broadcast and cable viewing were low. Nothing comes through clearer when looking at the newly weighted Gauge from Nielsen than the fact that the panel-only ratings that the ad industry has been living on were low. In fact they were too low by almost 10% each, requiring about 2+% (of all TV households) increases each.

Only 30% of U.S. households are “streaming TV only.” Yes, for all who are certain that streaming TV has killed and destroyed linear TV, never to be heard from again, the message is: Linear TV is still quite alive.

The June 2026 DASH report notes that streaming viewing adoption by households has not caused the dropping of linear viewing: “This pattern suggests that, for most households, streaming adoption is additive rather than substitutive.”

Antenna-based TV viewing is increasing. Yes, the use of antennas to get free over-the-air broadcast TV is going up, not down. It grew from 16% of U.S. households in 2025 to more than 17% this year.

Heavy dual use of streaming and linear TV. What the DASH report also screams is that if advertisers want to reach Americans on TV, they need to be on both linear and streaming. Neither on their own can deliver everybody -- not even close, and certainly not at an efficient scale.

Please check out the new DASH-enhanced Nielsen and let us know what you think.

10 comments about "DASH-Enhanced Nielsen Drives Reset On TV Viewing Picture".

Check to receive email when comments are posted.

Artie White from Zoom Media Corp, August 20, 2026 at 5:11 p.m.
Hi DAve, I'm not sure the conclusions here follow from the research. DASH is credible, but it estimates household access to TV services and devices; it does not measure what people actually watch. A household with an antenna or cable connection may still do most of its viewing via streaming. Likewise, the increase in antenna access from 16% to 17.1% is small and could easily reflect sampling variation without context (sample size, statistical significance, etc.) More broadly, Nielsen is introducing several methodological changes at once and explicitly says they are not guaranteed to increase ratings. Recalibrating the estimated universe is not, by itself, proof that broadcast and cable viewing had been undercounted by 10%.

I agree that it behooves advertisers to embrace linear and streaming in their strategies but to frame these numbers as a comeback story for linear seems premature at best.
Reply

Dave Morgan from Simulmedia replied, August 20, 2026 at 5:30 p.m.

Artie, very good points. Thanks very much for raising them.
As I called out in the headline of the piece, it is the DASH enhancement of Nielsen's core viewing data from its panel and from its Big Data set of second-by-second viewing from set-top boxes, smart TV's and apps that makes the difference. DASH's value is to help remove macro sample biases in those data sets, which capture viewing from many tens of thousands in the first case and more than a hundred million in the second. The almost 10% undercount number is what is relfected in the new Nielsen Gauge numbers from a viewing reach number. And, as all who use Nielsen data know, the recalibratoins have had signifcant impacts in increasing total TV rated impressions. Thus, I felt comfortable describing it that way.
For sure, all samples and porojections carry error rates. On the antenna issue, the fact that it is not down says a lot. Of course, growth in antennas is also validated by consumer sales data of digital antennas on Amazon and smart TV sales with installed digital antennas.

Ed Papazian from Media Dynamics Inc, August 20, 2026 at 6:27 p.m.

Dave, while I agree with most of your comments about how advertisers should look at TV and I have no beef agaist the ARF's DASH sstudy, all it is, using afairly small sample of 10,000 per year, is a survey. Various sets of respondents are interviewed and asked questions about how they get their TV and the results are melded together to estimate the way the whole country gets its TV--linear only, over-the-air or cable, streaming-only, etc. Calling it a "measurement" implies a level of precision in obtaining the answers that may be unwarranted.

As for what is happening, Nielsen has tradionally been forced to estimate the size of the TV-owning universe as reliable and current tat on this subject is rarely. I recall long ago--1960, to be exact--when I was very, very young---tht Nielsen pegged U.S. TV ownership at 90% but when the census came out the "correct" figure was 89%. So Nielsen had been putting out slightly "wrong" audience numbers for a while--but there was no big fuss about it. Nielsen merely revised its projection base downward slightly to conform to the "truth".

Nielsen is doing more or less the same thing again, only now the various segments --linear-only, streaming-only, both, etc.--- are being reweighted according to the DASH estimates.

As a result, there will be changes in individual show and seller ratings and in highly generalized general reports like The Gauge, but, as the folks at Nielsen will point out, the basic trends will no doubt continue--slow but, for a time, steady growth in Streaming's share of total viewing.

Ed Papazian from Media Dynamics Inc, August 20, 2026 at 6:54 p.m.

Again, I apologize for my "typos". My new computer has a hair trigger and deletes parts of words without any warning. Sigh!

Jack Wakshlag from Media Strategy, Research & Analytics, August 20, 2026 at 8:33 p.m.

Certainly we know the basic trend will continue Ed, but the point here is the old way exaggerated non linear viewing and the temporary discontinuity will immediately fix that. Of course, smart advertisers and sources will know the change is only in the metrics, not the behavior. Still there will be bragging rights gains for some and lost for others.

Dave Morgan from Simulmedia, August 20, 2026 at 8:34 p.m.

Spot on Jack.

Dave Morgan from Simulmedia replied, August 20, 2026 at 8:37 p.m.

Ed,, there is no question that Flks have complained -- for good reason -- about Nielsen numbers. My point here is that a big part fo the complaint was around sample bias and making sure panel and the big data were "fit" to population realities. DASH helps there. Yes. It is small. But the panel is much bigger than it used to be and its big data is enormous. Plus, the ARF created DASH and the MRC accredited it for the purposes that Nielsen uses it. We will never have perfrect TV ad measurement. But no question that it just got a good bit better.

Ed Papazian from Media Dynamics Inc, August 21, 2026 at 7:47 a.m.

Dave and Jack, of course, nobody can say that Nielsen is perfect --after all it, too, is only a survey--or now, a collection of surveys all mashed together. I think that our industry has become far too reliant on the ratings--carried down to a decimal point--as if they were reporting reality. Truth is that we will never know from any survey what reality is. Even when a seller like Amazon can pinpoint exactly how many sets got its content, we don't know who watched.

The Gauge was a nice effort by Nielsen to promote itself and at the same time to inform everybody, in a general way, what's happening. Unfortunately it became a promotional toy for the pro -streaming, anti- "pay TV" folks.

Dave Morgan from Simulmedia replied, August 21, 2026 at 8:23 a.m.

Good points Ed. However, when adveriteers spend the better part of $80 billion a year on linear and streaming ads on TV, a tenth of a ratings points is many tens of millions of dollars a month, so it matters. Wehter it is perfect or not is less of an issue than it is the best that an be done with the available data and technology and that it is stable and comparable to other TV programming, other media alternatives and the past and, critically, trusted (as Jack taught me a long time ago).

Ed Papazian from Media Dynamics Inc, August 21, 2026 at 10:54 a.m.

Dave, I understand the point about stability--or consistency. Obvously if you become overly reliant on the ratings that's what you want, not abrupt changes in methodology--like the PPMs for radio or the people meters in TV --in both cases replacing diaries.

What's striking, to me, at least, is that while previous dramatic alterations in radio and TV "audienve currencies" didn't make everyone happy, they were accepted and the buyers and sellers made the required adjustments to cope with the changes in the data. But now, we seem to be going overboard on our reliance---and expectations--from Nielsen. I guess that's largely a function of the wave of "digital" thinking that's sweeping the industry. We are demanding levels of micro-precision in our TV ratings that can not be attained. Sad.