In an earlie piece, we misspelled Tania Yuki's name and incorrectly described her as having been hired as Comscore's CMO. In reality, Tania joined Comscore following its acquisition of Shareablee, the company she founded, and her responsibilities extended well beyond marketing to include oversight of Comscore's digital portfolio and related initiatives.

Tania also pushed back on my characterization that she was trying to position Comscore as "the future of media measurement." According to her, that messaging existed long before she arrived and was something she considered "pompous, tone deaf and unhelpful." Her focus was on rebuilding Comscore's reputation through original research, thought leadership, and cross-platform measurement initiatives rooted in the company's actual data assets.

Frankly, her broader point is worth noting. A data company without interesting things to say about its data is a very expensive filing cabinet. Tania's argument is that Comscore once produced genuinely useful industry analysis, and that part of the company's identity had faded by the time she returned. Whether those efforts ultimately moved the needle is a separate question. But it's fair to say the challenges Comscore faced were larger than any marketing strategy, and on that point we largely agree.

The theme this week is simple: the ad industry has decided it no longer trusts anything. Not its data. Not its inventory. Not its AI. Not its users. Not its platforms. Not even its own supply chain. After spending the better part of two decades building an infinitely scalable digital advertising machine, everyone is suddenly looking around the casino asking who checked the cards. The result is a week full of stories about control, verification, curation, measurement, and oversight. Or, put another way, an industry desperately trying to clean up a mess it spent years creating.

Nielsen announced new integrations with Mediaocean, Polk and MRI-Simmons, which sounds boring until you realize it's actually a classic power move. While every ad-tech company is busy stapling "AI-powered" onto its pitch deck, Nielsen is pursuing something much more valuable: dependency. The company wants to be sitting inside your planning tools, your audience data, your automotive targeting, your research stack and eventually every decision you make. This isn't a measurement strategy. It's a toll-booth strategy. Nielsen understands a brutal truth about advertising: the company that owns the pipes often makes more money than the company selling the water.

The ANA's latest transparency report should cause widespread embarrassment across media departments everywhere. The biggest performance gap in programmatic isn't coming from hidden fees, shady intermediaries or mysterious ad-tech taxes. It's coming from advertisers buying crappy inventory. The top-performing marketers converted 54% of spend into quality impressions while the bottom half managed only 32%. Even funnier, the smart advertisers paid lower CPMs than the bad ones. That's right: the people buying the best inventory spent less money than the people buying the worst inventory. The industry's favorite excuse machine just suffered a catastrophic malfunction. Turns out you can't optimize your way out of bad judgment.

Locality hired longtime ad-tech executive Tom Wolfe to expand its programmatic local television marketplace, making him the latest industry veteran tasked with delivering advertising's favorite fantasy: simplification. The pitch is familiar—more transparency, cleaner supply paths, better automation—but the timing matters. Advertisers are increasingly fed up with bloated supply chains and paying five companies to accomplish what one company should have done. The entire local TV market is being dragged toward the same destination as digital advertising: fewer middlemen, more direct relationships and endless PowerPoint presentations explaining why this time the middlemen are really necessary.

VIOOH's latest argument for curated programmatic DOOH boils down to one incredible industry realization: context is dead, data is king. Premium advertising used to mean prestigious environments. Today it means finding the right audience regardless of whether they're walking through an airport, standing in a gym or pumping unleaded into a leased SUV. The outdoor industry has quietly stopped selling locations and started selling probabilities. The billboard isn't the product anymore. The screen isn't the product anymore. You are the product. The gas station just happens to be where the algorithm found you.

The most refreshingly honest thing said about AI all week came from IAB UK's James Chandler, who suggested Chief AI Officers will eventually disappear. Of course they will. Nobody has a Chief Electricity Officer. Nobody has a Chief Internet Officer. The role exists because companies are currently terrified of missing the AI wave and need someone to organize the panic. What's more revealing is that marketers continue talking about autonomous agents, AI-generated creative and machine-led media buying while simultaneously refusing to trust any of it without human oversight. Everyone wants the robot to drive. Nobody wants to take their hands off the wheel.

Substack launched Reply Rules, allowing creators to establish comment guidelines while AI helps identify responses that should probably be hidden. In other words, Substack has discovered the same thing every platform eventually discovers: the comments section is undefeated. For years the company built its identity around creator freedom and minimal interference. Then reality arrived carrying spam, trolls, bad actors and people who type entire paragraphs in capital letters. The internet runs this experiment every few years. It always ends the same way. Absolute freedom sounds fantastic right up until somebody actually starts posting.

Louisiana's age-verification law is facing another legal challenge as digital rights groups argue that requiring users to verify their age threatens privacy, anonymity and free expression online. The state's position is straightforward: protect kids. The opposition's position is equally straightforward: don't build a surveillance system disguised as child safety. Both arguments are reasonable, which is why these battles never end. Lawmakers keep trying to regulate the internet like it's a shopping mall, while the internet stubbornly refuses to behave like one. The result is another courtroom showdown where everyone claims to be protecting the public and nobody can agree what the public actually needs protecting from.

Stay Bold. Stay Curious. Know More Than You Did Yesterday.

For about a decade, programmatic has run on a comforting bedtime story. It goes like this. You, the marketer, are a good and decent person buying good and decent media, and somewhere out in the plumbing a cabal of shadowy intermediaries is skimming your budget before it ever reaches a human eyeball. The supply chain is a crime scene. The SSPs are the suspects. You are the victim. Sleep tight.

The ANA's Q1 2026 Transparency Benchmark just flipped on the overhead fluorescents, and the answer to the headline is the one nobody in the room wants written on the whiteboard. It's option B. And this time there's a chart.

Here is the stat that should settle this forever, and won't, because settling it would require people to look in a mirror they spent a decade avoiding.

The Benchmark split participating advertisers into two cohorts, top half and bottom half, and measured where the performance gap actually comes from. Transaction costs between the two cohorts differed by just 2.4 percentage points. The gap in media productivity, which is the polite term for "the quality of the inventory you chose to buy," differed by 19.4 percentage points.

Read that twice. The "ad-tech tax" everyone screams about at conferences, the hidden fees, the mysterious middleman skim: rounding error. The real money gets set on fire by buyers who keep buying garbage and calling it a media plan.

So no. The intermediaries didn't rob you. You walked into the digital dollar store, filled your cart, and tipped on the way out.

The full spread between the best and worst advertisers is now the widest the Benchmark has ever recorded. The top cohort converted 54.0% of spend into qualified impressions. The bottom cohort managed 32.1%. That is a 21.9-point chasm, and you could lose a CMO down it. ANA chief Bob Liodice said the quiet part in trade-association dialect: lower-performing advertisers are "falling further behind." Translation from the original Diplomatic: the gap is widening, and pretending otherwise is no longer a defensible career strategy.

It gets funnier, in the way a pie to the face is funny if it isn't your face.

The good advertisers paid less. Not less per quality impression. Less, period. The higher-performing cohort paid an average CPM of $5.45 against $7.40 for the bottom half, while generating substantially more quality impressions. The people buying the best inventory spent the least money. The people buying the worst inventory paid a premium for the privilege.

Somewhere a media buyer is in a QBR right now explaining that their higher CPMs reflect "premium placements." Sir. The data says you overpaid for slop.

Adjust for what actually got delivered and the gap turns biblical. On a quality-adjusted TrueCPM basis the top cohort paid $7.46 per thousand qualified impressions while the bottom paid $19.04. A $1.95 difference in raw CPM balloons into an $11.58 difference once you account for the waste. That is not a fee. That is a self-inflicted wound with a line item.

This demolishes the last economic alibi standing. The entire argument for chasing cheap programmatic was efficiency: lower CPMs, more reach, same budget. The math is now public and it runs the other way. The mechanism is not subtle. An advertiser buying a $0.50 CPM into an open exchange running at 30% quality is effectively paying $1.67 per impression that counts. The buyer paying a $1.00 CPM into a curated marketplace at 60% quality pays the same $1.67, and also walks away with better measurement, better brand safety, better viewability, and attribution that survives contact with a finance team. Cheap inventory is the most expensive thing in the building.

Controlled tests across four global brands made the point in the bluntest possible terms: prioritizing quality delivered a 32% reduction in CPMs, a 33% lower cost per action, and a 5% lift in ROAS. Reckitt's Sameer Amin, who buys media for Lysol, Mucinex, and Enfamil, explained why everyone keeps missing it. Econometric models and marketing mix analysis reward cheap inventory because they never look at whether the impression was any good. Your measurement was built to flatter your worst instincts, and it has done so loyally for years.

For a decade the industry's reflex response to every waste study was to point at the plumbing. The Benchmark series has now walked through the morgue and tagged each excuse individually.

"The supply chain is too complex to control." Both cohorts used the same supply chain. The 19.4-point gap came entirely from buying decisions. Demolished.

"Better inventory costs more." Quality buyers had lower CPMs, lower cost per action, and higher ROAS. Demolished.

"We don't have the data to optimize." Log-level data exists. Top performers use it. Bottom performers don't open it. Demolished.

"Transaction costs are eating us alive." Transaction costs differed by 2.4 points between the best and worst. Demolished.

"This requires expensive technology nobody has." Same DSPs, same SSPs, same tools, opposite results. As the Alkimi Exchange put it, this is not a technology gap, it is an operations gap, and most trading desks already know it. Demolished.

"Fraud and MFA are the real problem." This one earns an asterisk, and it is the most instructive of the bunch. Made-for-advertising spend was 15% of programmatic in 2023. The industry mobilized, and exposure fell below 1%. Genuine progress. And yet total waste over the same window grew 34%, to $26.8 billion. Killing the MFA weeds did nothing for the garden, because the root cause, undisciplined buying and CPM-chasing, was never touched. And the weed is already growing back: MFA ticked up to 1.1% this quarter, with the Benchmark flagging AI slop as an emerging subtype. We built machines that generate infinite garbage, and we are now paying CPMs to advertise next to it. Progress. Complicated.

Here is the part that should sting the most, because it is so unglamorous.

The top cohort did not buy a smarter algorithm. They did not unlock a secret exchange. They were disciplined, and they were boring, and they did the thing your intern suggested in 2019 before someone overruled her.

They ran inclusion lists instead of exclusion lists, defining the universe of acceptable publishers rather than playing reactive whack-a-mole with bad ones. The 2023 study found the average campaign sprawled across 44,000 websites. The ANA's own finding is that you can reach 95% of a valuable audience on fewer than 5,000, and that optimal performance is achievable on as few as 75 to 100 trusted sellers. This is not a new or contested idea. JPMorgan Chase cut its list from 400,000 domains to 5,000 back in 2017 and measured no performance drop. That finding is nearly a decade old. The bottom half of the industry has spent ten years failing to operationalize a one-line memo.

They put their money where it could be watched. Top performers directed north of 92% of spend into private marketplaces rather than the open exchange, the highest-waste channel in programmatic. They reconciled log-level data against what platforms reported instead of taking the dashboard's word for it. They enforced supply-path controls at the insertion-order level, not just the account level, which is the single most common place quality governance quietly dies. And they retired raw CPM as a buying metric in favor of TrueCPM, so they were optimizing toward media that exists rather than media that is cheap.

None of this is a moonshot. All of it is unglamorous, repeatable, and free. It is flossing. The bottom half does not floss.

So if the fix is known, proven, and cheaper than the disease, why are roughly half of all programmatic advertisers still hemorrhaging value every quarter?

The measurement alibi. When your marketing mix model rewards cheap impressions without grading their quality, optimization dutifully flows toward more cheap impressions. The metrics don't lie. They just measure the wrong thing, beautifully, forever.

The incentive structure. Trading desks were built in an era that rewarded volume and scale over quality, and the KPI frameworks that encode those habits are still bolted to the wall. Nobody got fired for buying a billion impressions. People get fired for asking why.

The talent gap wearing a complexity costume. A 21.9-point quality gap between cohorts using identical infrastructure is not an algorithmic mystery beyond human control. It is the cumulative result of inclusion lists never built, MFA classifications never refreshed, supply paths never enforced below the account level, and log-level data never reconciled. Each is a choice. Each was made by a person. Each is on a resume somewhere.

The scale fallacy. The comfort of running across tens of thousands of domains was diagnosed as a structural failure years ago, and the bottom half maintains it anyway, mistaking reach for results and breadth for diligence.

So, corrupt or stupid?

The corruption story was always the more comfortable one, because being robbed is something that happens to you, while being bad at your job is something you do. For a decade the industry preferred the role of victim to the role of suspect. The Benchmark has now quietly reassigned the parts.

The sin in programmatic was never theft. Theft would at least be interesting. The sin is sloth. It is the buyer who never opened the log-level data, never pruned the list, never asked why a campaign was running across fifty thousand sites nobody has heard of, and then blamed "the ecosystem" when the numbers came back ugly. The ecosystem did not betray you. You just never read it.

The transparency report now arrives every quarter like a Yom Kippur you cannot skip, and every quarter the message is the same. The tools are free. The methodology is public. The case studies show 40% lower cost per conversion for the advertisers who simply decided to act. The waste is not a tax levied on you by villains. It is a choice you keep making, and signing off on, and expensing.

You cannot optimize your way out of bad judgment. You can only stop exercising it.

For years, the easiest story in advertising has been that somebody else stole the money.

The SSPs took it. The exchanges took it. The middlemen took it. The pipes took it. Somewhere inside the labyrinth of programmatic acronyms, a shadowy figure was allegedly running off with the budget while innocent media buyers stood around looking confused.

Then Andrew Casale came on The ADOTAT Show and ruined the fairy tale.

"The buyer is responsible for determining value but also separating good from bad."

That's not exactly the sort of quote that gets standing ovations at media conferences. It requires accountability. Accountability is unpopular. It has terrible branding.

Casale spent much of the interview dismantling one of ad tech's favorite defense mechanisms, the idea that bad inventory somehow sneaks into media plans like a raccoon breaking into a garbage can.

Speaking about MFA sites, Casale didn't sound particularly sympathetic.

"It was very clear that it's traffic arbitrage. It's very clear that the content is of dubious value."

That matters because the ANA's latest benchmark found something the industry is going to spend the next six months pretending not to notice.

The gap between winning advertisers and losing advertisers wasn't primarily transaction fees.

It was inventory quality.

A lot of it.

The report found only a 2.4-point difference in transaction costs between the best and worst performers. The gap in media productivity was 19.4 points.

In other words, the industry spent years chasing pennies while lighting dollars on fire.

Which brings us to perhaps the most uncomfortable thing Casale said all day.

"Historically those APIs have never really been used."

Translation: the receipts were sitting on the table the whole time.

For a decade the industry demanded transparency while treating log-level data like the treadmill in a hotel gym. Everybody liked knowing it existed. Very few people actually touched it.

Casale's broader point was that the market is entering an efficiency era. Growth is harder. Supply is exploding. Budgets are under pressure. The days when buyers could spray ads across 50,000 domains and call it a strategy are ending.

What makes this particularly awkward is that none of his solutions sound revolutionary.

Use the data.

Audit the supply.

Ask harder questions.

Stop rewarding garbage.

That's it.

No magical AI.

No blockchain.

No mystical optimization layer powered by buzzwords and venture capital.

Just operational discipline.

The ad tech industry spent years hunting for criminal masterminds hiding in the supply chain. Entire conference panels were built around the premise. Consultants made careers from it. Vendors sold software to fix it.

What Casale suggests is far more uncomfortable.

The villain may not be a shadowy intermediary.

The villain may be a media plan that nobody bothered to read twice.

And judging by the ANA benchmark, half the industry is still trying to optimize bad decisions instead of stopping them.

The full ADOTAT Scorecards. Vendor comparisons that name winners and losers, with the analysis your planning team needs before Q3 budget meetings.

The takes nobody else will publish. Independent means we don't answer to vendors, agencies, or the trade press establishment.

The archive. Every Scorecard, every analysis, searchable. The reference layer for how this industry actually works.

Read by senior leaders at every major holding company agency, GroupM, Publicis, Omnicom, IPG, Havas, Dentsu, and inside the buying and strategy teams at Netflix, Roku, Pinterest, Disney, NBCUniversal, Spotify, Amazon, Meta, and most of the platforms shaping where ad dollars actually go.

If your competitors are reading ADOTAT+ and you aren't, you're planning with less information than they are.

$99/month. $999/year (save $189). Expense it.

Founding member rate. Locks in for life.

Also, no ads. Zero ads. No Ads.