Somebody at a major network told me how they buy measurement. Not as a leak. Not with any sense that it was interesting. The way you'd describe your commute.

They partner with a few companies for always-on outcomes, so they don't have to commission custom work every time they want to know whether a campaign did anything.

Survey vendors for the upper funnel, awareness and recall.

Behavioral vendors for the lower funnel, the part where somebody theoretically buys something.

Full scope of metrics across a variety of industry categories.

It was a description of competence. That's what made me start pulling.

Because the thing about a system that has become completely normal to the people inside it is that nobody in there can see its shape anymore. You have to be standing somewhere else, holding the paperwork, being annoying.

On July 29, Fox Advertising announced it was extending its partnership with iSpot. The framing was outcomes. Kym Frank, Fox's SVP of Research, put the pitch in the only words that survive contact with a chief marketing officer: did what I buy work?

It is the right question. It is arguably the only question. It is also the one nobody independently checks.

Here is what is actually verified in that arrangement, precisely and narrowly. iSpot holds accreditation from the Media Rating Council for national TV ad occurrence reporting.

Translated out of trade-body English: did the spot air, on which of 178 networks, at what time. MRC assessed the accuracy at 98.6 percent or higher. Google's Ads Data Hub holds accreditation for SIVT-filtered viewable video impressions and TrueView views, which Google described at the time as a first for a clean-room-based measurement product.

Both accreditations are real. Both were earned. Both required an independent firm to spend months inside the company verifying the claim, at the company's own considerable expense.

And both answer delivery.

Not effect. Not outcome. Not whether the money did anything. Delivery. Verifying that across 178 networks is genuinely hard, which is why it takes months and an accounting firm and why 98.6 percent counts as an achievement rather than a rounding error.

It is also, definitionally, not an answer to Frank's question.

Neither accreditation covers outcome attribution.

Neither covers cross-platform deduplicated reach. Which means the audited portion of a nine-figure arrangement is the portion nobody was arguing about, and the portion that decides where the money goes next year is carrying a badge earned for something else entirely.

Nobody is lying. Nobody has to be. That is the whole reason this is worth six thousand words.

The Media Rating Council has a name for what happens next, which tells you it happens often enough to need one.

Ron Pinelli Jr. is MRC's SVP of Digital Research and Standards. He is the person who actually runs this, as opposed to the person who gives the keynote about it. He calls it the halo effect, and defines it in a parenthetical so precise it reads like something he has had to explain roughly nine hundred times: "I am accredited full stop without distinction."

A note on how this reporting happened, because it should shape how you read the rest. MRC answered a long set of questions in writing, in detail, with no conditions attached.

They produced internal policy documents I didn't know existed.

They corrected a premise I had gotten badly wrong, which cost me a week of reporting and improved the piece considerably.

And they volunteered a line about their own flagship standard that no communications professional on earth would have cleared.

Almost nobody in this industry does any of that. Hold onto it for later, when we get to the companies with real money and see how they handle the same questions.

So. Full stop without distinction. Accredited, generally, at large, vibes-wise. Accredited the way a restaurant is clean or a lawyer is licensed.

Not accredited for a specific metric, in a specific environment, in a specific geography, which is the only way accreditation has ever once been granted in the entire history of the organization.

Pinelli says the reports come in constantly. Members flag claims. Outsiders flag claims. MRC corrects the record directly with the company, and when a claim gets amplified far enough, publicly, through the trades, which is the industry equivalent of being grounded in front of the neighbors.

And here is where he declines to take the easy shot, which is the part I keep returning to. Asked why it happens, he doesn't reach for malice. "Usually this winds up being a sales team not fully educated or getting a little aggressive. Rarely is it ill intent."

Rarely ill intent. A seller who doesn't know what the certificate says, telling a buyer who doesn't know either, and eleven figures of television advertising sliding through the gap between them.

MRC is not passive about this, and the lazy version of this story, in which a toothless regulator waves at an industry doing whatever it likes, is simply false. I like to call out how the “industry press” often tries the easiest way to frame things, because the other option, is actually doing journalism.

Auditors at accredited services inspect internal and external disclosures as part of the audit itself. Not just the methodology. The marketing. Somebody reads the deck. The recourse is citing noncompliance and threatening accreditation decisions, and Pinelli says the vast majority of vendors comply quickly once the phone rings, because losing the badge is considerably worse than overselling it.

There is a written rule, too: an Accreditation Representation Policy, dated February 15, 2024, furnished to every audited service, governing exactly how accreditation may be described to customers. I asked for a copy expecting a fight. They sent it in an hour.

So: a published standard. An inspection regime. An enforcement mechanism. A policy document with a date on it and everything.

And the halo effect persists like mold.

Two reasons, both structural, neither anyone's fault.

The first is jurisdiction. Where MRC has no audit relationship with whoever made the claim, Pinelli says the council will still reach out and try to fix it, but has limited recourse. You cannot threaten to revoke an accreditation you never granted. The bouncer only works one door.

The second is headcount, and this is the number that reframes everything.

MRC has seven full-time employees.

One hundred sixty-five member organizations. More than a hundred audits a year, conducted by outside CPA firms because seven people cannot conduct a hundred audits. Set that against a measurement industry that has spawned dozens of new vendors since 2021, each with a sales team, each with a deck, each with a slide where the logo goes.

The scope problem is not carelessness. It is arithmetic.

Seven people, and Pinelli's own summary, delivered without a trace of self-pity: it is hard to have eyes and ears everywhere. He isn't complaining. He is describing a physics problem.

Which brings us to the sentence every media buyer in America should tape to the bottom of their monitor.

I asked Pinelli what a buyer should actually do when a vendor waves "MRC accredited" at them in a pitch. His answer is short and carries the particular flatness of a man who has said it many times to people who then went ahead and did not do it:

"Ask to see or go to our site and see the MRC accreditation letters. They are the record of truth."

Here is the thing. The letters are public. They have been public. MRC publishes them next to each accredited service, updates a status chart quarterly, maintains a parallel services-under-review list, and organizes the whole thing by medium like a library nobody visits. It's free. It's indexed. It is, at absolute most, an afternoon of reading.

Almost nobody reads it.

Not the buyers, who are spending the money. Not a great many of the sellers, who are quoting it.

And certainly not the trade press, which has spent three years covering the currency wars while treating "accredited" as a binary property a company either has or doesn't, like a driver's license or a soul. I include myself in that. I have written about this market for years and I had never once opened an accreditation letter until three weeks ago. It took an afternoon. It changed the story.

That is how a badge for ad occurrence becomes, four slides into a deck, a badge for whether advertising works. Not through fraud. Through a scope statement nobody opens, a sales rep who was never taught the distinction and would be genuinely startled to learn it, and a buyer with no particular reason to know that accreditation is granted per format, per inventory type, per environment, per geography, and covers precisely what the letter says and not one metric more.

Pinelli is unapologetic that the narrowness is the design.

MRC has to audit every format, inventory type, environment and geography before it can accredit any of them. He describes conveying that as increasingly complex, which is a regulator's way of telling you the market has outrun the language available to describe it, politely, while looking directly at you.

There is an accreditation covering whether a commercial aired on 178 networks, verified to 98.6 percent, by an independent accounting firm, over a period of months.

There is no accreditation, anywhere, held by anyone, covering whether any of it worked.

That is not MRC's failure. They wrote the standard for exactly that, four years ago. What happened next is the actual story, and it is stranger than negligence.

Kym Frank is SVP, Research and Data, Ad Sales at Fox Corporation. Her quote opens this series, and not because she said something incriminating. She said something correct.

Did what I buy work? Most of this business has built an elaborate vocabulary for gliding past that question. She built an announcement around it.

Here's why she knew what she was doing. Before Fox, Frank spent six years as President of Geopath, the out-of-home industry's audience measurement body. She ran a joint industry committee. She took the old Traffic Audit Bureau, rebranded it, relaunched the currency with mobile location data, doubled membership past 380 companies, tripled revenue past $12 million, and managed a 25-person board, which anyone who has sat on a 25-person board will recognize as the hardest item on that list. Her ARF bio also notes volunteer service with the Market Research Council, where she is President Emeritus, and the Media Ratings Council.

A former JIC president who has worked inside the standards machinery, now sitting on the sell side. When she describes how measurement gets bought, she is describing a system she helped build, from three different chairs.

Here's the part that matters. This series started because she answered a simple question about her job. Survey vendors upstairs, behavioral vendors downstairs, always-on outcomes so nobody commissions custom work every time. A description of competence, from someone with no reason to think it was interesting.

Most of what I write starts with somebody angry. A grievance, a competitor with a knife, an ex-employee with a folder. Useful material, and always pre-shaped: somebody decided what the story was before I arrived.

This started with somebody who wasn't trying to tell me anything. Which is exactly why I went and pulled the accreditation letters. When a person with her résumé calls an arrangement routine, it genuinely is routine. Which means whatever is wrong with it isn't buried in one company. It's the water everyone is swimming in.

When I asked how a network handles vendors disagreeing about the same campaign, she corrected my premise rather than managing me. Different vendors, different verticals, different funnel stages. Two of them are rarely answering the same question. I checked it against NBCUniversal's published framework, WBD's statements, Disney's clean room announcements and the JIC's own documents. She was right about all of it.

And then the line I haven't been able to put down: "It makes it incredibly challenging for non-research practitioners to talk about research."

That's the thesis of this entire series, from somebody who watches it happen daily. She's describing her own working conditions. She's also describing exactly why this was possible to write.

Disclosures. ADOTAT has no financial relationship with any company named in this series, and no company named has paid for placement, sponsorship, or advertising in connection with this reporting. Quoted material was submitted to speakers or their representatives for accuracy review before publication; the Media Rating Council reviewed and cleared its quotes, as did EDO, which asked that its responses be attributed to an EDO spokesperson. No source was given advance review of this series' framing, analysis, or conclusions. Readers should know that in January 2026 iSpot won an $18.3 million judgment against EDO in a data misuse case, that both companies compete in outcomes measurement, and that EDO is the only vendor in this category that agreed to answer detailed methodology questions on the record, while iSpot, VideoAmp, Comscore, MNTN, Fox and Nielsen either declined, did not respond before deadline, or responded as noted in the text. Kevin Krim of EDO, representatives of the Media Rating Council, and Kym Frank of Fox have all appeared in prior ADOTAT coverage, and Frank's description of how Fox buys measurement is what prompted this reporting.

ADOTAT published "The Nielsen Bonfire: Who's Holding the Match?" in April 2026, which took a skeptical view of the alternative-currency narrative and noted that the Video Advertising Bureau is funded in part by Nielsen's competitors; that position informs this series and readers should weigh it accordingly, as they should the fact that Nielsen's communications team has separately encouraged ADOTAT to publish critically about its competitors.

ADOTAT accepts no paid placement as editorial, is funded by subscriptions and reader support, and publishes uncomfortable findings regardless of the PR fallout, including about people we like.

This series took six weeks. Pulling accreditation letters nobody reads. Reading an ANA tax filing line by line. Comparing an SEC registration statement against a help-center article. Getting the Media Rating Council to answer, in writing, questions nobody had put to them before.

None of that has a sponsor. There is no advertiser in ad tech who wants a piece about whether ad tech's numbers are real.

Personal subscription: $50/month, or $600 a year. One named human, non-transferable. Every part of this series in full, the source documents I worked from so you can check my work instead of trusting me, updates as companies respond, and the reporting that can't run in public, which is most of the interesting reporting.

Corporate and team licenses are available. If you need an invoice, a W-9, net-30, or seats for a floor that's already passing this around on one login, email me directly and I'll sort it. Corporate single seat runs $1,200 a year. Team licenses are quoted on request.

What ADOTAT doesn't do: paid placement in editorial, sponsored coverage dressed as analysis, or the polite silence that keeps a trade publication's ad sales pipeline healthy. That's the entire business model. It works only because enough of you pay for it directly.

If you buy measurement, sell measurement, or answer to someone who does, this is the one you want to be ahead of.

[SUBSCRIBE — $50/month] · [CORPORATE — email [email protected]]

This series took six weeks. It meant pulling accreditation letters nobody reads, reading an ANA tax filing line by line, comparing an SEC registration statement against a help-center article, and getting the Media Rating Council to answer questions in writing that nobody had put to them before.

None of that has a sponsor. There is no advertiser in ad tech who wants a piece about whether ad tech's numbers are real.

At $50 a month, this is not a casual subscription, and it isn't meant to be. It's priced for people who make decisions where being wrong is expensive: buyers moving eight figures, vendors defending a methodology, investors underwriting a category, executives who need to know what their own team hasn't told them yet.

What you get: every part of this series in full. The source documents I worked from, including the accreditation letters, the representation policy and the S-1 language, so you can check my work rather than trust me. Updates as companies respond, and they will respond. And the reporting that can't run in public, which is most of the interesting reporting.

What ADOTAT doesn't do: paid placement in editorial, sponsored coverage dressed as analysis, or the polite silence that keeps a trade publication's ad sales pipeline healthy. That's the entire business model. It works only because enough of you pay for it directly.

If you buy measurement, sell measurement, or answer to someone who does, this is the one you want to be ahead of.

[SUBSCRIBE — $50/month]