Having published over 100,000 sweepstakes, I can say that humans are still superior verses AI. We hand place 100 percent of the promotions. Further, there is no AI program that can show over 700 sweepstakes in the AI presentation format. There is a place for AI but not in advertising other than a expanded bean counting program. 

Don't worry Wayne ... the under 30s won't have time to read your comment (hehehe).

Adam, the Nielsen stats refer to overall "viewing" time by persons aged 2+ and they include a lot of TV set activity--over 10%---- which can not be identified as to exactly what the content is or where it comes from--most likley videogames, digital audio and DVR plays of old movies, etc. Also, a lot of the "viewing" takes place when the "viewer" is not even in the room or, if so, is paying little or no attention.

Most importantly, this particular Nielsen report includes viewing to ad-free services like linear TV's "pay cable" channels, PBS stations, etc. And a substantial amount of streaming activity is to ad free content. Add to that, the fact that almost all ad supported streaming services offer many fewer commercials per hour than their linear TV counterparts and consequently, streaming  share of TV's ad impressions is barely 20% ( some have the percentage lower )--which is way below its share of viewing.

The people in charge of Streaming's ad-supported operations will have to decide if they are willing to risk the ire of their current users by greatly increasing their commercial clutter in order to match linear TV's profit margins, or whether they can find some way to convince marketers that paying very high CPMs for their  viewers will pay out --because of "better targeting"---in terms of ROI. That's not an easy sell, it seems. 

As for You Tube---which I like and use many times per day, myself--there are serious questions about whether it's "TV" from an advertiser's point of view. This goes to how the commercials are placed and ad attentiveness, the kinds of content--professionally made or otherwise--- issues about "audience" measurement and reporting and ad blocking--I pay Google $14 per month not to get ads when I go to You Tube. So let's not get carried away with the viewing stats--there's more than that to be considered.

Great piece, Adam, and much appreciated.

@Ed Papazian: You were ahead of your times. I think the difference is it's not a model, but synthetic respondents responding as if they were actual people. I'm not sure it's necessarily better, but from everything I've ever heard of -- and reported on -- vis a vis human respondent bias, I'm not sure it's any worse either.

Agreed, Joe.

Long ago, when I was at BBDO, I created a "model" that used fake--or "synthetic"--- groups of viewers and their demo characteristics plus TV viewing habits and used it to estimate the ratings for network TV shows for an upcoming season--including, of course, new programs with no rating track record. It was much simpler than what Dentsu is doing and much more focused on one aspect, but we got interesting results. Typically, we  predicted about 65-70% of the new show ratings within a few share points of what actually happened--which was a little bit better than the human experts were doing. Then we refined the "model" and improved its performance slightly so that its predictions for TV's second season--that time in the first quarter when an additional slew of new programs replaced those that bombed out in the fall----and our "accuracy" level rose about 5 points on the newbies and, of course over 90% for those shows that weren't cancelled. It was an interesting experiment--but as upfront buying moved away from show-specific buys to multiple show bundles there was no need for our model as the rating ups and downs for a mass of TV show ad placements more or less cancelled out any errors. 

I use Google Gemeni and in certain cases it is quite helpful--especially when it hunts for specific data. But when I asked it , "Who is Ed Papazian?", just for kicks, it began with a correct description but used the past tense--now that worried me. Then it closed by saying that I had died way back in 2003--because it found a guy with the same name who sadly passed on then. Sigh!

@Ed Papazian: That is actually how it works. I'm no expert on how you do that, but I asked one -- ChatGPT -- and here's what it said:
Creating a synthetic respondent for a consumer survey involves generating a simulated persona or dataset that mirrors the characteristics, preferences, and behaviors of a real consumer—without actually surveying a real individual. This can be useful for modeling, testing, hypothesis development, or scaling insights when real data is limited. Here's a step-by-step overview of how to do it:


1. Define the Purpose
Clarify why you're using synthetic respondents. Common use cases include:

Testing survey design
Modeling market segments
Simulating future scenarios
Running pre-research experiments

2. Build or Acquire a Foundational Dataset
You need a basis for realism. Sources can include:

Actual survey data
Public datasets (e.g., census, Pew Research, Nielsen)
CRM or digital behavior data
Market research reports

3. Identify Key Variables
Decide which attributes your synthetic respondent needs to emulate. These typically include:

Demographics: age, gender, income, education, location
Psychographics: values, attitudes, interests
Behavioral: purchase habits, media usage, brand preferences
Situational: employment, family status, device ownership

4. Generate the Respondents (Methods)
There are multiple approaches depending on the fidelity you want:

A. Rule-Based Generation
Use if you want control or simplicity.

Create profiles manually or with logic rules
E.g., "A 35-year-old urban mom likely shops online weekly and values convenience."
B. Statistical Modeling / Simulation
Use probabilistic sampling from real distributions.

Techniques: Monte Carlo simulation, Bayesian modeling, bootstrapping
Helps generate statistically valid synthetic populations
C. Machine Learning-Based Generation
Use when you have enough real data.

Clustering (e.g., K-means): to generate personas from real data clusters
Generative models (e.g., GANs, VAEs): to synthesize data matching real distributions
Can be refined with techniques like SMOTE for balancing attributes
D. Language Models (LLMs)
Use when you're simulating open-ended responses or behavior narratives.

Prompt an LLM (like GPT-4) with a persona and a question:

"You are a 42-year-old suburban dad who drives a hybrid and cares about sustainability. How do you decide which cereal to buy?"
This is useful for qualitative survey pre-testing or ethnographic simulation

But Joe, when you create a systhetic respondent don't you have to give that synthetic person characteristics, mindset and behavior patterns which, I assume,  would be based on some sort of complex profiling. For example, say you ask one of your synthetic respondents ----who is a man,  36 years  old, married with two children, working  at a markting company who is a light viewer of TV, etc. etc.---- questions about what media he uses or how he might respond to excess ad frequency, etc. To get that far, don't you have  to define each synthetic respondent's activities,  interests, attitudes, etc. so "he" and others like  him can supply useful answers? Or is everything purely random?  

@Ed Papazian: It's not a model. Synthetic audiences are AI-generated respondents that participate in panel-based research just like human respondents. Basically, they're asking the synthetic respodents questions to generate responses as if they were humans responding to the questions.

Sounds like a multiple correlation simulation model  designed to predict actual audience attainment and, maybe, ad reach/frequency, etc--all supposedly producing outcomes.  What I don't get is that media plans for a future year are usually not  specific as to what TV or other platforms---what TV network or social media website, etc--- are to be used nor what vehicles--TV shows, magazines, etc---wil carry a brand's ad messages. So what data is being used? Or is the model creating synthetic data for its machinations?

Joe,

you are suibright on this. Keep at it. 

phil

Hi Wynn, Thanks for reading! Research (from a multitude of sources) shows that most EV owners end up putting in a level 2 home charger. But even with simply a 120 volt household outlet, you could easily add 75 miles of range overnight.  Folks can do that even out of apartment or condo windows with extension cords.

Tony, re your comment about JWT, you may be thinking about the early 1980s fiasco where Mari lousi, the head of the agency's TV syndication unit obtainted a time bank of some $18-20 million worth of spot TV commercial time that could be cashed in --but wasn't as yet sold off to any advertiser or broker. Some idiot in the shop's financial group  claimed a third of that amount in a quarterly income earnings report to Wall Street. This caused a huge flap--misleading the analysts, being the big complaint ----and at least one head, Mari's, did get loped off--as I recall.

But there's an important point implied in your post that I'd like to comment on. It's about determining the "standard" currency that national TV needs--without which no buyer--principal or otherwise--could function. There can be only one standard--most likely its "audience" ---and only one source for such data---barring a monumental screw up that causes another supplier to be recruited  to do the same job.Without that single standard currency no buyer or seller or advertiser could  know whether he had made a good or a bad one as there would be no basis for comparisons with other possible deals. This does not preclude the use of  non-standard metrics nin addition to the standard one--again, "audience".

 We have got to stop confusing our terms. Nielsen is not the standard, it's just a research company that has been supplying the indusrty's standard currency for a long time. Many years ago--in the early 1940s---Nielsen replaced radio's standard audience counting  methodology--telephone coincidental calls made by C.E. Hooper----because it had a better method--meters. But it was still trying to measure exactly the same thing--radio's audience.   

BMW Group may be hearing their market research the same way that Ford heard the good news on the Edsel. The touted 2025 Mini Countryman SE ALL4 EV has a range barely exceeding 200 miles. That doesn't square with their 75% of consumers wanting a 75-mile daily range, which would require a charge every other day. Or call AAA for a tow every third day. 

Josh: As Maarten eloquently and I believe appropriaiately stated it's really about, "transparency, trust and integrity", ever decreasing characteristics in the media buy/sell arenas, notably in the US as exemplified by "Principal Media Buying".  Years ago when JWT went well beyond aggregating brand buys and spending for just a single advertiser/client in the upfronts via a JWT 'principlal media buy', i recall there were threats of law suits for unethical conduct and I believe some at the agency lost their jobs. 
Sadly this decline in transparency, trust and integrity, is also reflected in the media currency and research arenas.  As an industry research guru and champion you are on the front lines to addresss the current damaging practices and failures to follow the best practices and established media research principles.
As a colleague opined, Advertisers ultimately determine the status and behaviours of our industry via their ad dollars. So perhaps "we" are merely getting what "they" deserve? 
As an early member of https://www.advertisingwhocares.org/ , "we" are trying to underline and remind advertsiers of best practices across all the cornerstones of the ad industry to help them understand and embrace the most effective advertising practices and investments possible to dirve brand equity and brand strength for both for the long and short term. 

All the Turner guys mentioned in the article got "retired," but Nielsen rolls on.

Trump's total DEI opposition is shameful. This together with his characterization of immigrants as violent black and brown criminals and other actions seem to be elements of a campaign to appeal to racists such as White Nationalists, replacement theorists, Proud Boys, and others who blame their grievances on people who are different from themselves. 

Can you share the research that includes those stats? Is it B2B CFOs or B2C or all?

No Josh. The agency assigned the UTV upfront buying assignment for an advertiser handles only the aggregated brand requirements and spending for that advertiser. There may be isolated cases where principal buying is involved--but usually for very small sellers. It's not a common practice.

Please forgive a naive question, but I'm genuinely curious. Isn't this basically how the upfronts work? Agencies buy blocks of network inventory in advance, and then parcel these out to their clients? 

This is a solid take on how social media advertising has changed. The environment is definitely more unpredictable, and brand safety is a real concern. But at the same time, it’s worth remembering that an ad on a social platform doesn’t automatically mean a brand supports whatever happens to be in someone’s feed at that moment. Most people, especially younger users who’ve grown up with this stuff, understand how ad targeting works. Everyone’s joked at some point about just mentioning a product and suddenly getting bombarded with ads for it.

So yes, brands should be smart and think long-term, but that doesn’t mean they have to run from social media entirely. Used the right way, it can still be effective. The key is staying intentional and not letting short-term wins come at the cost of long-term trust.

Nice piece Jon and I agree.

As my firm's focus is on understanding and marketing to older adults (55+), I often see this large and growing demographic either ignored or stereotyped in marketing efforts.  Given their spending power, it is a mistake not to put some brand resources and budget against reaching and targeting this group.

As for the technology gap, our research shows that the vast majority of older adults are quite savvy although they need to see the practicality of the tech to explore and adopt it.  As we have seen in other categories and with other brands, my reco would be to meet the needs of these older fans as that will also be beneficial to younger ones.