Newsletter · · Ashutosh Agarwal

Google Escapes an Ad Tech Breakup as Judge Rejects Forced Exchange Sale - Weekly Online Ads Podcast Recap - Week of September 13, 2026

Weekly Online Ads Podcast Recap for the week of September 6 to September 13, 2026. Podcast synthesis on Judge Brinkema declining to break up Google's ad-tech business in the DOJ antitrust case, retail media's aggressive push into connected TV led by Amazon and Walmart, the bull and bear debates over The Trade Desk, AppLovin, Meta, Netflix and Reddit, and AI emerging as both the ad industry's growth engine and its new brand-safety threat.

Weekly Online Ads Podcast Recap

Week of September 13, 2026: Google Escapes an Ad Tech Breakup as Judge Rejects Forced Exchange Sale

Coverage window: podcasts published 2026-09-06 through 2026-09-13 (the past 7 days). This is primary-window coverage, 44 episodes surfaced across programmatic ad tech, retail media, connected TV, Meta and Google's ad-and-AI story, AppLovin, Reddit, the DOJ's Google ad-tech antitrust case, agency holding companies, and measurement/identity. No fallback to the older 30-day window was needed.

Executive Summary

Google dodged a breakup for the second time in a year, and the ad-tech world mostly shrugged. On September 2, Judge Leonie Brinkema declined to force Google (GOOGL/GOOG) to sell its ad exchange, opting for "behavioral" fixes (changes to how it runs its business, watched by a monitor) instead of a structural breakup. The full ruling is sealed until after September 16. Legal experts called this more disappointing than the search case because control of the exchange was central to the wrongdoing, but the government showed up with no buyer for the exchange, and the judge concluded the cure could be worse than the disease. Google is expected to appeal regardless. (Bloomberg Law, Votes & Verdicts, Total Leo/This Week in Tech)

"Nothing changes" was the week's blunt verdict on the FTC's separate case against Amazon's ad auction. Because big advertiser budgets are locked in months or even a year ahead, and because Amazon (AMZN) wins on shopper traffic rather than auction trust, commentators expect the money to stay put. (Retail Media Breakfast Club)

The retail media land-grab into TV accelerated. Amazon merged its whole ad system (DSP, sponsored ads, and its Amazon Marketing Cloud data tool) into one back end on July 30, mirroring Google's setup, and Walmart (WMT) completed its buy of streaming-TV ad platform Vibe.co and struck a Roku (ROKU) partnership, using its store-level purchase data to chase big brand TV budgets. (Ecommerce Braintrust)

The Trade Desk (TTD) is the fallen darling of the moment. A dedicated episode titled "WTF Happened to The Trade Desk?" argued its roughly 20% take rate (its cut, including data fees) was never sustainable, that it leaned too hard on the open web while Amazon's DSP gained, and that its stock could keep sliding, with fringe talk of going private or being delisted. (Next in Media)

AppLovin (APP) is the mirror image: a "broken stock, not a broken business" debate. Its shares have halved this year even as revenue grew more than 50%, investors are split on whether AI is a real threat or whether the market is mispricing a cash machine that reaches 1.4 billion daily users. (The Intrinsic Value Podcast)

Connected-TV economics got real numbers. Netflix's (NFLX) ad-free plan earned about $21.50 per US user in 2025 versus just $9.80 for its ad-supported plan, the core tension in every streamer's flirtation with a free tier. Disney (DIS) is reportedly "sold out" of ad inventory, and YouTube now commands about 13% of all TV viewing time. (Inside The Stream)

Reddit (RDDT) is having a moment as the web's trust layer for both shoppers and AI models. Its data-licensing deal with Google (GOOGL), about $60 million a year, is up for renewal, and nearly everyone expects Reddit to demand far more. It posted an eighth straight quarter of revenue growth and is still only ~2% of US social ad revenue, leaving lots of runway. (Behind the Numbers/EMARKETER)

Two big structural themes ran underneath everything: "everything is becoming an ad network," and AI is the new brand-safety battleground. Server-to-server data pipes (called CAPI) are letting streaming services, banks, airlines and even ChatGPT build ad businesses, while a new worry emerged that a tiny amount of planted text (one study found just 13 words) can poison what an AI says about your brand. (Mobile Dev Memo, The Programmatic Digest)

Coverage gaps this week: no episodes with specific ad-spend forecasts from GroupM/Magna/Zenith, no dedicated Apple Search Ads episode, nothing substantive on Chinese ad platforms' advertising businesses (Baidu/Alibaba/PDD/Tencent), and no dedicated political-ad-cycle episode. The one agency-holding-company nugget: PepsiCo moved its global media account from Omnicom (OMC) to Publicis. (Sleeping Barber)

(a) Dominant Themes

1. Google escaped a breakup, and the shared reaction was weary resignation

The single most-discussed story of the week was Judge Leonie Brinkema's September 2 decision in the Department of Justice's ad-tech case against Google. Back in April 2025 she had found Google illegally monopolized two advertising-technology markets. The remedy she chose last week: not a forced sale of Google's ad exchange (the marketplace that sits between websites selling ad space and advertisers bidding for it), but "behavioral remedies", court-ordered changes to how Google runs the business, overseen by a monitor. The full opinion is sealed for redactions and is expected to become public after September 16.

Across legal and tech shows the consensus was that Google got off relatively lightly. On Bloomberg Law, NYU antitrust professor Harry First made the striking point that this outcome is arguably more disappointing than the earlier search case, because here the thing the government wanted sold off was the heart of the wrongdoing: "The divestiture that the plaintiffs sought in the ad tech case is actually central to the theory of liability... their control of the exchange... was critical for Google's control of this whole market and for Google managing to squeeze more money out of the parties than they otherwise would have. The publishers in particular. And the judge found that." But he also noted the practical hole in the government's case: "They don't have a buyer. So that cuts the other way."

Why do judges keep declining to break up Big Tech? First's answer was that the government spends most of its energy proving wrongdoing and too little building the case for a specific cure. On Votes & Verdicts, Bloomberg Intelligence's Justin Teresi framed the market impact bluntly: "Google dodges a breakup again. That is the big news here." He stressed how much is still unknown until the sealed remedies drop, and flagged the real swing factor for rivals, whether the behavioral rules get frozen during an appeal that "could go on for years." The hosts of This Week in Tech reached the folksiest version of the same point: the judge decided "the cure would be worse than the disease," and Google's businesses have become so intertwined since its DoubleClick acquisition that pulling one piece out is "heart surgery... and they want to take out the heart." The through-line: everyone agrees Google is a monopoly in ad tech; almost no one thinks the legal system moves fast enough to do much about it.

2. Retail media's expansion into TV, the "duopoly to triopoly to quadopoly"

The second dominant narrative was retail media networks (the ad businesses retailers build on top of their shopper data) pushing aggressively into connected TV and consolidating their tech. On Ecommerce Braintrust, the hosts detailed two moves. First, Amazon's July 30 consolidation of its DSP (its programmatic buying tool), its sponsored ads, and Amazon Marketing Cloud into "a single global hub", a change one host said is "clearly matching what Google has done with GAM, with their Google ad manager." Second, Walmart's completed acquisition of Vibe.co, "a leading self-service streaming TV advertising platform," plus a new Roku partnership.

The strategic logic was spelled out clearly: retailers can only tap so much money from e-commerce teams, whose budgets are performance-driven and smaller. To reach the giant global brand budgets, "you need to have a really high functioning, powerful streaming TV option." The host argued Walmart's edge is the same as Amazon's, verified purchase data down to what you actually bought, which beats the "behavioral intent signals" and "lookalike modeling" that independent platforms like The Trade Desk rely on: "What Amazon brings to the table and what Walmart is bringing to the table here is they have the data all the way down to... what thing did you buy." The framing of the week: the old Google-Meta-Amazon "triopoly is now becoming a quadropoly," with Walmart the serious fourth player.

3. "Everything is becoming an ad network", powered by a plumbing change most people missed

Eric Seufert used his Mobile Dev Memo episode to revisit and extend his well-known "Everything is an Ad Network" thesis. The short version: after Apple's 2021 App Tracking Transparency (ATT) privacy change kneecapped the cross-app tracking that made Meta so powerful, companies with their own first-party purchase data (retailers, banks, airlines) suddenly had a reason to sell ads. The newer insight is that the durable enabler isn't ATT at all; it's CAPI (the Conversions API), a server-to-server pipe that sends purchase data directly from an advertiser's own systems to a platform, "entirely outside of the purview of a browser or mobile operating system." That makes the data flow immune to Apple's cookie crackdowns.

Seufert's list of the newest "ad networks" is the tell for how broad this has gotten: Chase Media Solutions (a bank using card-transaction data), United's Connective Media (an airline using traveler data and seatback screens), Ace Hardware, Academy Sports, and OpenTable Media. Crucially, he noted that Netflix added a CAPI in March 2026, and ChatGPT "launched a CAPI, a Pixel, and SelfServe" ad tooling in May 2026, evidence the model now stretches from streaming to AI chatbots. The catch he flagged: advertisers will only wire up this deep data connection for a handful of their most important partners, "placing a practical ceiling on the number of CAPIs they are willing to support", which paradoxically makes the biggest platforms (led by Meta) even more entrenched.

4. AI as both the growth engine and the new brand-safety threat

AI ran through nearly every episode, but two angles dominated. The first: AI is the story platforms are selling to justify enormous spending. On The Programmatic Digest, guest Aisha put it sharply, Meta makes "over 90 plus percent of their revenue from digital advertising," pushes its Advantage+ AI ad tools and its Llama model, and there's an "illusion of choice" because "the particular digital platforms themselves need you to use AI in order to justify the investment they've already made." Dr. Cecilia added the memorable image: the same companies selling the ad space are selling the AI tools to buy it, "We'll sell you the picks. We'll sell you the land. And good luck getting that gold."

The second, newer angle: brand safety is no longer just about what content sits next to your ad, it's about what AI systems say about your brand. The Programmatic Digest hosts cited a Cornell study finding it took only 13 words of planted text on Reddit to shift an AI search answer, opening the door to "SEO poisoning" and "corporate espionage or sabotage" that is "hard to track and trace... not auditable." They predicted a new "cottage industry" of brand-defense and verification firms, and noted TIME "recently started experimenting with ads designed to influence AI systems," with one person behind it saying "maybe it's more important to influence the agent than even the human."

5. The great ad-tech "rebundling", the stack is collapsing

A quieter but recurring theme: the layered ad-tech supply chain is being squeezed down. On Next in Media, the analysts argued "the next wave of the whole ad tech industry... is the tech stack has to collapse," with SSPs (the sell-side platforms that help publishers) pulling buying logic onto their own systems, more direct deals, and agencies claiming they're "becoming platforms, not agencies." The same convergence showed up in CTV transparency: on AdTechGod Pod, Gracenote's Bill Condon described "content intelligence" signals (knowing the exact show, episode, actors and rating your ad runs against, not just a vague genre) becoming "table stakes," now plugged into Index Exchange, PubMatic and The Trade Desk, with Gracenote's ad product having launched in December 2025.

(b) Active Debates

Debate 1: The Trade Desk (TTD), permanently impaired, or oversold?

The bear case got a full airing on Next in Media's "WTF Happened to The Trade Desk?" The analysts argued TTD's economics were built on a take rate that "was consistently around 20% to include the data fees... and that was not going to be sustainable." They said Jeff Green's company bet too heavily on the open web and misjudged CTV as an open-web-style buying opportunity when it turned out to be "much more direct." They faulted the company's messaging, "writing op-eds saying Google has broken the consumer's trust... just shows that they're not focused where marketers want to spend money", and warned sentiment could get worse: "if you imagine a few more quarters of weak results... their valuation falls in a pretty meaningful way from here? Sure."

The bull/nuance case, from the same episode, is that this is still "a sizable company... cash flow generative in a pretty meaningful way," and the opportunity remains if it refocuses products on "what it seems like marketers actually want." The counter to the "agencies will replace them" fear: building a real DSP "is not that easy for an agency, AI coding or not, to step in" and replicate. Investors on The Intrinsic Value Podcast landed firmly bearish, noting they passed on TTD and still won't buy it "despite it falling 75% from the date we published" their episode.

Debate 2: AppLovin (APP), broken stock or looming AI victim?

The bull case: on The Intrinsic Value Podcast, the setup is a classic value opportunity, "AppLovin stock has halved this year and over that same stretch revenue has grown by over 50% while free cash flow continues to compound at a very high rate... this isn't a broken business, it's a broken stock." The scale is enormous: its MAX platform "reaches up to 1.4 billion daily active users across more than 140,000 apps," and "growth advertising spend on AppLovin is more than Pinterest, Snapchats and Reddit's combined revenue." The flywheel, better ad-to-user matching lets AppLovin widen its spread (estimated at 30-40% of ad spend) without squeezing either side, is genuinely powerful.

The bear/hesitation case, from the same hosts, is what gives them "pause": both AppLovin and The Trade Desk "have been absolutely crushed due to AI disruption concerns this year," and it's hard to be sure the model is safe when "these large tech behemoths like Google and Meta... have such a strong hold on the space." The open question they kept circling: what stops Google (with AdMob) or Meta from replicating AppLovin's matching engine? At a peak "$250 billion market cap," AppLovin is now big enough that its niche "would be material to the economics of some of these big tech giants", a reason for them to attack it.

Debate 3: Meta (META), best-in-class compounder, or a one-legged stool?

This was the sharpest investor disagreement, laid out on Animal Spirits by Dan, a portfolio manager at Alger.

The bear worry: Meta "has kind of a bigger... existential risk than, say, Google does," because "other than the advertising business, they've not been successful in any subscription-based businesses, really." The concern is five years out: "will they have completely saturated us with... ads and AI?" He also flagged the deteriorating user experience, "about 85% of it is not from my friends anymore. They're ads."

The bull case (why Alger still owns it): "you are talking about the global leader in social media, and you're talking about a leader in advertising. We will tolerate some volatility and uncertainty around Mark Zuckerberg... because... that team there is executed. No doubt." He noted Meta trades around "15 times forward earnings", cheap, and that "it has proven in ad tech extremely hard to do two things. Turn around once your tech is behind... And two, beat Meta." The contrast with Google was explicit: Google's diversification (YouTube, Cloud growing ~100% last quarter, Waymo, its own chips) makes its AI story easier to believe than Meta's ad-only model.

Debate 4: The FTC v. Amazon (AMZN) case, will advertisers actually move money?

On Retail Media Breakfast Club, host Kiri Masters framed the only question that matters, "Does any of this change where the money goes?", and answered: "sadly, I don't think so." Either ad costs fall and performance marketers simply "buy more of them," or costs don't fall and "budgets are going to stay exactly where they are... allocated months, quarters, even a year in advance." Her key line: "moral disappointment is a much weaker motive to change than poor performance."

The other side came from her guests. Retail commentator Brian Gildenberg argued this case is legally easier than the Google one, "a little bit of a fake bid to make you pay more is a little more Tony Soprano than astrophysics", and predicted Amazon will "probably settle and not admit wrongdoing." Consultant Katie McKee raised a more alarming possibility: that years of inflated Amazon clearing prices may have "contaminated" the benchmarks, bid recommendations and third-party tools the whole industry plans against. Tom Limongello of The Middlemen Podcast held out hope advertisers "might consider moving some budget based on the perceived savings," but conceded that if budgets are locked, "the outrage will have settled long before there's an opportunity to make the change."

Debate 5: Should streamers launch fully free tiers? (NFLX, DIS, and the ad-economics tradeoff)

On Inside The Stream, Ampere Analysis's Rahul laid out both sides of the rumored free-tier moves at Netflix, Disney+ and Paramount+.

For free tiers: roughly 37% of US consumers already use a free ad-supported service, and huge shares of them don't subscribe to the big players (72% don't take HBO Max, 61% don't take Paramount+). A free tier expands the audience, increases engagement, creates more sellable ad inventory, and works as "a more in-depth shop window" to upsell people to paid. Disney is a prime candidate because it "mentioned within its most recent financial presentation... that it was fairly well sold in terms of its advertising and that more ad inventory would actually help the company accelerate their ad revenue growth."

Against: the "churn down" risk. Because ad-free users are worth so much more than ad-supported ones, Netflix's US ad-free ARPU was about $21.50 in 2025 versus $9.80 for ad-supported, a free tier risks pulling paying subscribers down to "a less valuable access point." Rahul's bet: a full free-tier launch by a major player before the end of 2026 is "unlikely," with markets like India and Japan the most logical first targets.

Debate 6: Reddit (RDDT), durable AI-era winner, or hostage to Google traffic?

On Behind the Numbers/EMARKETER, analysts Max Willens and Minda Smiley debated Reddit's future.

Bullish: Reddit is uniquely trusted, "the top cited domain across ChatGPT search, Google AI mode and perplexity, according to SEMrush." Its Google data-licensing deal (~$60 million a year) is expiring and "pretty much everyone I talked to expects that Reddit is not going to do this again for $60 million a year. They are going to demand much, much, much more" (for context, OpenAI was reportedly paying ~$70 million). Reddit posted "its eighth consecutive quarter of revenue growth," ad growth "well outpaces social competitors," and it's still "only... 2% of social network ad revenue in the US," implying lots of room to grow.

Bearish: Reddit's fate is tied to Google. If the relationship sours and referral traffic declines, both engagement and ad value could suffer. And Reddit's text-first nature limits how much lucrative video advertising it can add, Willens is "skeptical that they're going to be able to get much more in" without breaking what makes Reddit, Reddit.

Debate 7: Is AI-powered advertising actually good for anyone?

Underneath the AI hype, The Programmatic Digest hosts pressed a consumer-value question the platforms rarely address: "I have yet to find a consumer that said, oh, goodness, this wonderful AI generated ad has solved all of my life's problems." A related trust gap surfaced on Sleeping Barber, which cited eMarketer data showing that as of April 2026, AI assistants were the least trusted source for shopping recommendations at just 23%, behind family and friends (two-thirds), professionals (59%), and review sites (58%). The skeptics' takeaway: AI is being sold as transformative for advertising while consumers remain unconvinced.

(c) Tickers Mentioned

Bull/bear angles below reflect how each name was discussed on the podcasts this week, not investment advice. Every entry lists the source episode, speaker/affiliation where identifiable, episode date, and a verbatim quote.

GOOGL / GOOG (Alphabet / Google)

  • Bull angle: Diversified and winning the AI transition; avoided a breakup again. On Animal Spirits, Dan (portfolio manager, Alger) argued the market got the AI-kills-search fear wrong: the easiest way to introduce ordinary people to AI "is you have the leading search product and then you power it with AI," while "Google Cloud is growing 100% last quarter."

  • Bear angle: A found monopolist in ad tech whose remedy may still bite once unsealed; facing an appeal and lingering private litigation.

  • Source: "Google Survives Second Breakup Attempt", Harry First, NYU Law professor, on Bloomberg Law, 2026-09-09.

  • Quote: "Their control of the exchange, which is what the plaintiffs wanted divested, was critical for Google's control of this whole market and for Google managing to squeeze more money out of the parties than they otherwise would have. The publishers in particular. And the judge found that." (Harry First, Bloomberg Law)

  • Additional quote (bull): "The easiest way for introduce an ordinary person to AI is you have the leading search product and then you power it with AI and you give them even better and better search results. So that's exactly what's happened with Google." (Dan, Alger, Animal Spirits, 2026-09-07)

META (Meta Platforms)

  • Bull angle: Global leader in social and advertising, cheap at ~15x forward earnings, a management team that "has executed," with Reels "on fire" and AI-powered ad targeting working.

  • Bear angle: Over-reliant on a single business (ads = 90%+ of revenue); risk of over-saturating users with ads and AI over five years; huge AI spend on unproven future businesses.

  • Source: "Talk Your Book: AI Winners & Losers", Dan, portfolio manager at Alger, on Animal Spirits, 2026-09-07.

  • Quote (bull): "It has proven in ad tech extremely hard to do two things. Turn around once your tech is behind and off the edge. And two, beat Meta. So one of the reasons that we still own Meta is really simply, you are talking about the global leader in social media, and you're talking about a leader in advertising." (Dan, Alger)

  • Quote (bear): "We're just a little nervous about if you look five years out, will they have completely saturated us with, you know, ads and AI?... they have only one business. Unlike Google, they've only got basically an advertising business." (Dan, Alger, Animal Spirits)

  • Also on Meta's model: "They make over 90 plus percent of their revenue from digital advertising itself. But what is Meta also pushing? Advantage Plus." (Aisha, guest, The Programmatic Digest, 2026-09-09)

AMZN (Amazon)

  • Bull angle: Consolidated ad stack (DSP + sponsored ads + AMC) into one Google-like hub as of July 30, improving usability; dominance rooted in unmatched purchase data and shopper traffic; FTC case unlikely to move advertiser budgets.

  • Bear angle: FTC antitrust case over its ad auction alleges manipulated pricing; a "conceptually simpler" case than Google's that Amazon may prefer to settle.

  • Sources: Ecommerce Braintrust (2026-09-08); Retail Media Breakfast Club (2026-09-08).

  • Quote (bull/consolidation): "As of the July 30th upgrade, existing Amazon advertiser accounts automatically combined... programmatic, which is Amazon DSP, search ads, which is sponsored ads... Amazon marketing cloud into a single global hub... it's clearly matching what Google has done with GAM, with their Google ad manager." (Pat, Ecommerce Braintrust)

  • Quote (bear/why nothing changes): "An ad platform can mislead advertisers and still be the best place to spend a performance dollar. The money isn't being spent on Amazon simply because its auction is trusted. It's there because that's where the shoppers are." (Kiri Masters, Retail Media Breakfast Club)

WMT (Walmart)

  • Bull angle: Serious fourth force in digital ads, completed Vibe.co acquisition and a Roku partnership to chase big-brand TV budgets with store-level, verified purchase data.

  • Bear angle: Still chasing Amazon's head start in the DSP/CTV game (none voiced strongly this week).

  • Source: "The Retail Roundup", Pat and hosts, Ecommerce Braintrust, 2026-09-08.

  • Quote: "Walmart announced the completion of the company's acquisition of Vibe.co. So that's a leading self-service streaming TV advertising platform... this gives advertisers access to Walmart's first party customer data... I think Walmart is a serious, serious player in this game." (Pat, Ecommerce Braintrust)

TTD (The Trade Desk)

  • Bull angle: Still a large, cash-generative player with a real opportunity if it refocuses on what marketers want; hard for agencies to replicate a true DSP.

  • Bear angle: ~20% take rate seen as unsustainable; over-exposed to the open web; losing share to Amazon's DSP; stack collapsing around it; negative sentiment with fringe talk of going private/delisting; passed over by value investors even after a 75% fall.

  • Sources: "WTF Happened to The Trade Desk?" Next in Media (2026-09-09); The Intrinsic Value Podcast (2026-09-06).

  • Quote (bear): "The trade desk was, their take rate was consistently around 20% to include the data fees, right?... And that was not going to be sustainable. You could get away with that for a certain period of time, but not long term." (Next in Media)

  • Quote (value-investor bear): "This is why we decided to pass on the trade desk and why still haven't opened a position in that stock, despite it falling 75% from the date we published it." (Kyle Grieve / Shawn O'Malley, The Intrinsic Value Podcast)

APP (AppLovin)

  • Bull angle: "Broken stock, not a broken business", shares halved while revenue grew 50%+; massive scale (1.4bn daily users, 140,000+ apps); a powerful matching flywheel; growth spend on the platform exceeds the combined revenue of Pinterest, Snap and Reddit; so far shielded from Google-style regulatory overhang.

  • Bear angle: Crushed on AI-disruption fears; competitive threat that Google/Meta/Unity could replicate the model; hard to build a durable moat in online ads against the giants.

  • Source: "AppLovin (APP): The 30-Bagger Down More Than Half", Kyle Grieve & Shawn O'Malley, The Investor's Podcast Network, on The Intrinsic Value Podcast, 2026-09-06.

  • Quote (bull): "AppLovin stock has halved this year and over that same stretch revenue has grown by over 50% while free cash flow continues to compound at a very high rate. In other words, it appears that this isn't a broken business, it's a broken stock." (The Intrinsic Value Podcast)

  • Quote (bear): "If AppLovin can do this, well, what's stopping Google or Meta or Unity from just replicating the strategy? And why hasn't AppLovin taken even more of a market share than it already has?" (The Intrinsic Value Podcast), Alger also noted on Animal Spirits that it "still own[s] AppLovin, which we like."

NFLX (Netflix)

  • Bull angle: Very high ad-free ARPU ($21.50 in 2025 US); huge original library (~4,800 pre-2024 titles) that could seed a free tier; building ad measurement (added a CAPI in March 2026); "YouTube-ifying" with video podcasts and vertical video.

  • Bear angle: Ad-supported ARPU ($9.80) is less than half the ad-free tier, so a free tier risks "churn down" of valuable paying subscribers; lacks sports rights and local content in key growth markets like India.

  • Source: "Ampere Analyst Interview on SVODs' Opportunity to Launch Free Tiers", Rahul, analyst at Ampere Analysis, on Inside The Stream, 2026-09-11.

  • Quote: "In 2025 in the US, we found that the ad-free subscription to Netflix generated an average revenue per user of about $21.50. When we look at the same metric for the ad-supported tiers, we find that that weighs in at less than half, coming in at $9.80." (Rahul, Ampere Analysis, Inside The Stream)

DIS (Disney)

  • Bull angle: Ad inventory reportedly "sold out," with the highest fill rates in the market (Hulu and Disney+), expanding inventory could directly accelerate ad revenue.

  • Bear angle: Same free-tier "churn down" and brand-image risks as peers.

  • Source: Inside The Stream, Rahul (Ampere Analysis), 2026-09-11.

  • Quote: "Disney mentioned within its most recent financial presentation, where it said that it was fairly well sold in terms of its advertising and that more ad inventory would actually help the company accelerate their ad revenue growth." (Rahul, Ampere Analysis, Inside The Stream)

ROKU (Roku)

  • Angle discussed: Increasingly a distribution/partnership hub for retail media, new partnerships with both Amazon and Walmart; the Roku Channel cited as a major free ad-supported platform.

  • Source: Ecommerce Braintrust, 2026-09-08.

  • Quote: "We talked about on our last month's episode... the sort of aggressive moves by Walmart in the connected TV space acquisition of Vibe.co. We also are looking at their integration with Roku... now seeing a partnership on Walmart's end with Roku." (Ecommerce Braintrust)

RDDT (Reddit)

  • Bull angle: The web's trust layer, most-cited domain across AI search engines; Google data deal (~$60M/yr) up for renewal at a likely much higher price; eighth straight quarter of revenue growth; only ~2% of US social ad revenue, so long runway; owned in the Intrinsic Value portfolio.

  • Bear angle: Heavy dependence on Google referral traffic; limited ability to add lucrative video without breaking its text-first culture; a scare over a drop in ChatGPT citation rate hit the stock.

  • Source: "Reddit's Growing Importance", Max Willens & Minda Smiley, EMARKETER analysts, on Behind the Numbers/EMARKETER, 2026-09-11.

  • Quote (bull): "Even though the negotiations between Reddit and Google are ongoing, because that deal is now... about to expire pretty soon, pretty much everyone I talked to expects that Reddit is not going to do this again for $60 million a year. They are going to demand much, much, much more." (Max Willens, Behind the Numbers/EMARKETER)

  • Quote (bull/runway): "It still only represents like... 2% of social network ad revenue in the US... So I think it has a lot of room to grow." (Minda Smiley, Behind the Numbers/EMARKETER)

PINS (Pinterest) & SNAP (Snap)

  • Angle discussed: Referenced mainly as scale benchmarks against AppLovin, underscoring how large AppLovin's ad spend has become relative to these platforms. No standalone bull/bear thesis on either this week.

  • Source: The Intrinsic Value Podcast, 2026-09-06.

  • Quote: "Growth advertising spend on AppLovin is more than Pinterest, Snapchats and Reddit's combined revenue." (The Intrinsic Value Podcast)

U (Unity)

  • Angle discussed: Named as a potential competitor that could try to replicate AppLovin's advertiser/publisher matching model.

  • Source: The Intrinsic Value Podcast, 2026-09-06.

  • Quote: "If AppLovin can do this, well, what's stopping Google or Meta or Unity from just replicating the strategy?" (The Intrinsic Value Podcast)

OMC (Omnicom) & Publicis (PUB FP)

  • Angle discussed: A concrete account shift, PepsiCo moved its global media business from Omnicom to Publicis, cited as a marker of competitive movement among agency holding companies (though not analyzed in depth).

  • Source: "The Barber's Brief", hosts of Sleeping Barber, 2026-09-08.

  • Quote: "PepsiCo moved its global media account from Omnicom to Publicis." (as summarized on Sleeping Barber)

Non-ticker platforms and newer names mentioned

  • Gracenote (Nielsen's content-data business): launching CTV "content intelligence" ad products, now integrated with Index Exchange, PubMatic and The Trade Desk (TTD). Bill Condon, head of global advertising sales, Gracenote, AdTechGod Pod, 2026-09-09. Quote: "The roster of partners that we're working with from a programmatic platform standpoint, index exchange, Pubmatic, and now trade desk, extremely exciting... these signals are really becoming table stakes." He cited a Heineken/Dentsu live-sports campaign that achieved "0% waste" and roughly "7% efficiency in CPMs."

  • Vibe.co: self-service streaming-TV ad platform acquired by Walmart. (Ecommerce Braintrust)

  • ChatGPT/OpenAI (private): launched a CAPI, a pixel and self-serve ad tooling in May 2026; OpenAI reportedly pays Reddit ~$70M/yr for data. (Mobile Dev Memo, Behind the Numbers/EMARKETER)

  • New "everything is an ad network" entrants (all private): Chase Media Solutions, United's Connective Media, Ace Hardware's Red Best Media, Academy Sports' Academy Retail Media, OpenTable Media. (Mobile Dev Memo)

  • TIME (private): experimenting with ads designed to influence AI systems/agents rather than human readers. (The Programmatic Digest)

Coverage Notes & Gaps

  • Chinese ad platforms (BIDU, BABA, PDD, TCEHY, NTES, JD): no substantive advertising-business discussion this week. An Alibaba.com interview on Schwab Network (2026-09-10) covered its Accio AI e-commerce tool, not its ad platform.

  • Ad-spend macro forecasts (GroupM, Magna, Zenith, eMarketer): no episode with specific market-size or growth numbers surfaced.

  • Apple Search Ads / App Store ads (AAPL): no dedicated episode; Apple appeared only via its ATT/ITP privacy changes on Mobile Dev Memo.

  • Political ad cycle: no dedicated episode this week.

  • Specialist verification/identity names (DV, IAS, LiveRamp, CRTO, MGNI, PUBM, Zeta): aside from PubMatic being named as a Gracenote partner, none were the subject of a dedicated segment in the primary window.

  • Note on The Trade Desk share-price figures: one investor podcast referenced TTD price levels in passing, but the transcript numbers were garbled, so specific price points were omitted here; the 75%-decline figure quoted above is stated clearly on the record.