Ad tech industry analyst | blockchain enthusiast | author of Amazing Heights | Opinions and views are my own
Joined May 2010
- Tweets7.7K
- Following898
- Followers2.1K
- Likes22.1K
The term "Power Elite" comes to mind. They don't care...because they don't have to (for one reason or another🤑)
Seth Ulinski retweeted
Cheers to @Gartner_inc for disclosing the investment changes and payment status behind otherwise random movements in the Magic Quadrant. Love to see it.
SSPs gaining strategic value as programmatic control points evolve.
The 2026 OpenX Supply-Side Intelligence Trend Report is here.
Backed by 4 years of proprietary data and industry insights, explore the trends shaping media buying and the growing role of supply-side decisioning.
Get the report: openx.com/research/supply-si…
Just a matter of time before they release a "kid-friendly" version.
Meta’s cuddly spy called Muse has exposed private conversations by making links to shared chats publicly searchable. Predictable.
When someone shared a conversation, Meta created a web link with the title of that conversation. Those links ended up in publicly accessible records where strangers could find them.
Security firm Claroc discovered more than 6,800 shared conversations with a single search. Some contained sensitive personal and business information, including a loan application, a bank fraud alert and confidential company documents.
Anyone who found these links could open the conversations without a password or permission from the person who shared them.
Meta says it has suspended the sharing feature, but records of the exposed links remain publicly accessible.
💡 Meta should shut down this service. It’s riddled with security vulnerabilities and poses serious risks to everyone’s privacy, including people who don’t use it.
It doesn’t matter if you don’t use any of Meta’s products. If someone gives Muse access to their emails, messages or other private accounts, highly sensitive information about you could end up in Meta’s hands without your knowledge or consent.
If someone shares conversations you’ve had with them, or talks about you in messages Muse can access, Meta will learn who you know, what you’ve said, what you’re thinking, your emotional state and your personal vulnerabilities.
That kind of information is being used to build detailed profiles and social graphs about people who’ve never used Muse or agreed to share anything with Meta. You’d have no way of knowing what information had been collected about you or how it was being used.
And because Meta is a US company, information it holds is subject to demands from US authorities.
This is an extraordinary invasion of privacy, including the privacy of people who’ve never used the service.
Seth Ulinski retweeted
Don't miss tomorrow's @AdExchanger webinar presenting the results of an @OpenX research project looking into what is going on with ad targeting on the supply-side, conducted by me. Lots of cool insights - you will not be disappointed.
Thursday, October 14, 8am PT/11am ET.
bigmarker.com/AdExchanger/th…
Facebook Marketplace borrowed a page from eBay's sales playbook but forgot to update Ad Manager.
$META $EBAY
Facebook Marketplace's little-known Cost per Sale ad option promises sellers they'll only pay if an item sells, but Meta’s own ad tools show those campaigns using impression-based billing behind the scenes. 💸
$META
valueaddedresource.net/faceb…
Seth Ulinski retweeted
HUMAN Security Launches Ad Integrity Suite, Putting DV and IAS on Notice
Today, HUMAN Security fired a direct shot at ad verification's cozy duopoly, rolling out its AI-Powered Ad Integrity Suite for Platforms with inaugural partner OpenX.
HUMAN, long recognized as the elite cybersecurity powerhouse taking down sophisticated botnets like Methbot and 3ve, spent recent years quietly expanding into AdTech. Now, they are aggressively pushing beyond pure-play invalid traffic (IVT) mitigation, bundling bot detection, viewability, and brand safety into an integrated, explainable engine designed specifically for DSPs, SSPs, and ad exchanges.
Why Platforms Should Care
For years, programmatic pipes have suffered from verification acting as a black box. Buy-side verification slapped platforms with bare-bones pass/fail verdicts without explaining the reasoning.
HUMAN is mending that measurement weak spot. By packaging explainable signals, granular reason codes, and custom AI safety agents directly into platform architecture, infrastructure players can bake pre-bid inventory decisioning directly into deal curation and supply-path optimization. No longer, platforms have to cobble together fragmented vendors or build detection infrastructure from scratch. For launch partner OpenX, integrating these signals sharpens supply-side intelligence that gives media buyers tangible proof of clean inventory.
What’s Next?
Still, HUMAN has work to do. The platform offering is phased, with some capabilities coming out in the first half of 2027. The company is also relatively new to the game and must gain clients’ trust: HUMAN’s viewability accreditation is recent, brand safety is younger still, and this edition launches with one only one named partner. And yet, as soon as HUMAN can prove the efficacy of the Ad Integrity Suite with but one or two clients, none of this should hamper the new product’s success.
Shaking Up the Duopoly
For HUMAN, this marks their arrival as a full-spectrum verification player. For the broader industry, it lands a disruptive jab against DoubleVerify and Integral Ad Science.
DoubleVerify and Integral Ad Science dominate roughly three-quarters of verification budgets, but both carry historical baggage. They grew up as buy-side, post-bid cops, heavily tethered to blunt keyword blocklists and rigid taxonomies that frustrate publishers and platforms alike. HUMAN enters the fray unencumbered by legacy measurement stacks, leveraging deep cybersecurity DNA, real-time platform APIs, and behavioral machine learning.
They won't dethrone the incumbents overnight. Yet, by empowering sell-side curation and giving programmatic exchanges teeth, they are shifting verification from defensive compliance into proactive marketplace intelligence.
@SecureWithHUMAN @OpenX @doubleverify @integralads
Default search engine relationship (billions paid annually by Google) may include sweetheart provisions for Big G
Seth Ulinski retweeted
Good news, bad news in Google-accountability tonight.
First, good: antitrust adtech case survived nearly all summary judgment motions. 1st peek at damages shows two pubs alone at $1.5B (this gets tripled), AdX class at $1.7B, and still unpacking auction rigging effects. 1/2
Magnite moves into protocol territory rather than software (SSP)
Power up MCP + AdCP
$MGNI
Excited to announce the expansion of Magnite Orchestration.
Orchestration now extends across linear and addressable TV, CTV high-impact formats, digital audio, and display. New supply partners include DIRECTV, Samsung Ads, iHeartMedia, local linear via ITN, and many more
What this means for publishers: We're working with you to turn your inventory into machine-readable products at scale. Richer context, audience insights, and rates.
What this means for buyers: We're making it easier than ever for buyer agents to discover premium supply, launch omnichannel campaigns, and optimize spend.
We've integrated agents from agency holding companies, independent agencies, and tech vendors like Newton Research and Databricks with more in the works.
By connecting agents across every screen, we're partnering with the ecosystem to drive a major shift in how advertising dollars transact.
Full press release in comments 👇
Seth Ulinski retweeted
Meet MADJobs.ai.
Whether planned or unplanned, the job search process can be a drag.
MADJobs is here to help you get hired, fast with the world’s largest job board built specifically for advertising and marketing.
Powered by AI, MADJobs connects job seekers with more than 12,000 opportunities worldwide, from the biggest tech giants to the most innovative ad startups. MADJobs is brought to you by Marketecture Media, and is totally free to both job seekers and hiring managers.
Get started at madjobs.ai.
Read the full press release: marketecturemedia.com/p/mark…
Seth Ulinski retweeted
Scooplet: Filling in for @KendraEBarnett on this week's AI, Tech & Money newsletter —digging inside The Trade Desk's layoffs, its market cap collapse, and why some in adtech (including @sam_khoury1 @bokelley) think its days as an independent could be numbered.
Subscribe!
DOJ attempts to seize stolen assets, preventing victim restitution.
The U.S. seized $225 million in crypto for scam victims in Operation Big Tuna. Returning the funds has proved harder. on.wsj.com/4d6Oxdi
I am one of the 200+ victims of the Turkish money-laundering ring identified by Nivie Kaul and Digital Defenders Group.
I thought the immutability of the blockchain would let us trace the stolen assets. It did.
I thought that evidence would help stop the criminal operation. In Turkey it has produced more than 130 arrests.
As a law-abiding American citizen I thought the Department of Justice would be on the victims’ side. It is not.
We are now fighting our own government for funds that were already under a foreign court seizure order.
Yield vibes 😎
@aripap
BREAKING: We just sued Amazon over what we believe was a vast, ad-rigging scheme.
We believe Amazon found a way to charge a few extra pennies to an enormous number of transactions, adding up to over $20 billion in illegal charges.
This is about how something becomes a "Sponsored product."
Amazon lets businesses bid to be "Sponsored" but it told businesses that the highest bid would only pay slightly more than the second-highest bid.
For example: If you bid $1 to have your product “Sponsored” and the next-highest bid is $0.50, you should pay about $0.51.
But we believe that Amazon would run the auction, see those bids, and then secretly raise the second-place price so it could charge the winner more.
It was a gradual strategy based on concealment.
Internal Amazon documents show they tested how much they could raise those hidden charges without advertisers catching on, then raised them more when businesses didn’t react, then gave false answers when businesses started asking questions.
This is a case about affordability, fair competition, and basic honesty in the marketplace. My thanks to the big bipartisan team on this, from AGs across the country to the FTC.
Seth Ulinski retweeted
Just got home from Judge Mehta’s courtroom for the motion to dismiss hearing in Penske Media v Google. The core issue: can Google use its adjudicated search monopoly to force publishers to provide content for AI uses at zero cost? Some notable exchanges today. 1/5
GenX+Boomers = 70% of disposable income in the US 🤑
@bendingspoons let's make semantic sexy again
😍📨#youvegotmail
The World's leading B2B platform needs a thorough house-cleaning if it wants a slice of the +$500B sports marketing opportunity
@LinkedIn @kingjames @NBA
I accidentally found a fascinating personal-brand experiment on LinkedIn.
Search Stephen Curry.
It’s a reasonably common name, so you get plenty of results. But it takes about two seconds to figure out which one is the @StephenCurry30. His presence is obvious, credible, and connected to who he is beyond basketball: entrepreneur, investor, founder, businessperson.
Now search LeBron James.
It’s chaos.
Hundreds, maybe thousands, of results. Fan accounts. Impersonators. Questionable profiles. Real people who happen to share his name. And somewhere in all that noise, the most commercially significant LeBron James in the world is effectively absent.
That leaves me with two questions.
First: What are @KingJames people doing?
This is someone who has built an enormous business ecosystem around his name: entertainment, investing, media, philanthropy, consumer brands and more. LinkedIn would seem like an obvious place to establish and protect that professional identity, even if he personally never posts a thing.
Second: What is @LinkedIn doing?
At some point, identity integrity becomes a platform issue. When searching for one of the most recognizable people on Earth produces a wall of obvious impersonators and fan accounts without an authoritative identity to anchor the results, that feels like a problem.
The Curry/LeBron comparison makes it especially visible.
One search says: this is the person, and this is what he’s building.
The other says: good luck figuring it out.
For marketers and brand leaders, there’s a broader lesson here too:
If your brand has enough value that other people want to borrow its identity, absence from a platform doesn’t mean you don’t have a presence there.
It may just mean you’ve surrendered control of it.
Seth Ulinski retweeted
Sell-side folks - how many bidders tend to show up when a deterministic ID is present vs not?
I’m looking for a rough cut of distribution of bidders per auction and how it tracks with ID presence.
Does the ID matter, or is any auth ID roughly the same once you have coverage?
