Pundit
Ari Paparo
50 receipts • Tracked since Apr 2026
Adtech veteran (Beeswax founder, ex-DoubleClick, Bloomberg). Marcellus newsletter author covering programmatic and the open web.
Affiliations
No firm affiliations on record.
— Filed —
50 receipts
2026
6 predictions- Aug 10, 2026Too early to tell.“Not to put too fine a point on it, but businesses associated or dependent on the open web are pretty much dead in the water.”
This prediction was made only ten days before the current date of August 20, 2026, making it too early to observe meaningful outcomes or structural business failures. Detailed traffic and financial data for open-web-dependent publishers after this date would be required.
- Mar 16, 2026Too early to tell.“If you put enough trust in a single platform to get your media-buying job done with fewer employees, less waste, less error, and faster turnaround times, you might just find that you’ve accidentally ended up in Stage 4, with the wholesale elimination of certain types of partners.”
The 'wholesale elimination of certain types of partners' due to increased trust in single platforms powered by AI in media buying is a long-term trend that is still in its early stages.
- Feb 2, 2026Called it.“It’s actually I think it’s inevitable that some portion of the content people read and watch will be AI generated, and obviously it’s going to radically change the way that content is presented.”
By mid-2026, many major publishers and marketing teams acknowledge using generative AI for content creation, indicating that a significant portion of consumed content is at least partially AI-generated or AI-assisted.
- Feb 2, 2026Story still evolving.“The first one is the ad server ad exchange combo that Google has has built is going to break down in some way.”
Antitrust actions continue, but a complete structural breakup of Google's ad server and ad exchange has not yet occurred as of mid-2026.
- Feb 2, 2026Called it.““Consumers are moving to kind of what I call a multipolar world where they’re seeing their content from all [these different places]…””
Throughout 2026, media consumption data showed continued growth of niche platforms and individual creators, leading to increased fragmentation of consumer attention across many sources, aligning with Paparo's "multipolar world" prediction.
- Feb 2, 2026Called it.“[The open web] is obviously just absolutely getting destroyed right now. ... AI is radically changing the way consumers surf the web, find information.”
Publishers and analytics providers reported noticeable declines in click-through rates from search to open-web pages in 2026 due to expanded AI summary features, supporting Paparo's prediction of reduced traffic.
2025
19 predictions- Nov 26, 2025Too early to tell.“I think Google’s going to get their way and there won’t be a spin out.”
The outcome of the Google-DOJ adtech case regarding a potential AdX spin-out versus behavioral remedies is not provided in the supplied information as of August 20, 2026. The prediction's accuracy cannot be verified yet.
- Jul 22, 2025Too early to tell.“The digital market is getting more fragmented every day, making the centralized role of the DSP harder and harder to maintain.”
This describes an ongoing trend of fragmentation that makes the DSP's role harder. While it's a current assessment with future implications, a definitive 'right' or 'wrong' verdict requires a clear time horizon and measurable outcome not yet available in the provided material.
- Jul 22, 2025Too early to tell.“AI agents will make media-mix models more important.”
The prediction is about a future trend, and the evaluation period is still ongoing or has not yet begun definitively relative to the publication date of the reporting article. The Marketecture post itself is not provided, so its original date is inferred from the MediaPost article.
- Jan 2, 2025Partially right.“Ari predicted a Silicon Valley company (like Netflix) would buy TV Scientific.”
TV Scientific was acquired by Pinterest, which is a Silicon Valley company, but not Netflix specifically.
- Jan 23, 2025Story still evolving.“the era of like easy cheap targeting is sort of coming to an end”
Later commentary states 'targeting is becoming harder as third-party cookies and browser/privacy changes reduce signal availability,' which aligns with the prediction.
- Jul 21, 2025Partially right.“It’s very hard to believe that AI won’t be better than humans at campaign setup and targeting very soon.”
AI has become central to campaign setup and targeting, but full displacement of human campaign managers has not yet occurred, and DSPs remain in market.
- Jul 1, 2025Called it.“Agentic commerce is (still) a collective hallucination.”
Two years later, large-scale, mainstream deployments of fully autonomous agentic commerce systems are not yet present, supporting Paparo's contrarian view against hyperbole.
- Jul 21, 2025Partially right.“It's very hard to believe that AI won't be better than humans at campaign setup and targeting very soon.”
Major platforms have aggressively rolled out AI-driven campaign automation, consistent with Paparo’s thesis. However, a universal benchmark proving AI is *universally* better than humans in all contexts is not yet available, making it partially validated.
- Dec 23, 2025Story still evolving.“2026 will see a significant rise in M&A activity.”
As of mid-2026, partial year data shows ongoing M&A in ad-tech. While some verticals (CTV, retail media) are seeing heightened deal activity, it is too early to definitively confirm a 'significant rise' across the entire sector compared to prior years. The prediction is directionally plausible but not yet conclusively validated or refuted.
- Sep 9, 2025Story still evolving.“I think it's very likely that Google will be forced to spin out some of these crown jewels of its advertising business.”
Antitrust proceedings against Google’s ad tech stack are ongoing in the U.S. and EU. Google continues to command massive closed-loop performance ad share.
- Aug 18, 2025Partially right.“We’re in the midst of the biggest disruption since the iPhone and that’s AI. …we already [are] seeing market share declines from Google for the first time ever. SEO is dead.”
Google’s search share has faced its first meaningful competitive pressure from AI-native interfaces. Google introduced AI overviews/SGE, altering SEO tactics. SEO remains a large industry but has significantly evolved.
- Nov 26, 2025Called it.““I think Google’s going to get their way and there won’t be a spin out.””
Judge Brinkema did not order a structural breakup or spin-out of AdX in April 2026, aligning with Paparo's prediction of behavioral remedies.
- Sep 3, 2025Called it.““seems extremely unlikely to blunt Google’s market share.””
As of mid-2026, Google continues to hold a very large majority of general search queries, and there is no evidence of a sharp or remedy-driven decline in its overall search market share attributable to this ruling. Competitors have not materially closed the gap due to this specific remedy.
- Jan 1, 2025Partially right.““If that’s the case we would look for either a) some retailers giving up; b) generally a lot of noise about how it ‘might not be worth the hassle’; and c) aggregation platers across retailers.””
The prediction of aggregation players across retailers has largely materialized with the rise of platforms integrating multiple retail media networks. There are also signs of smaller retailers struggling or partnering, aligning with 'some retailers giving up' and 'a lot of noise about how it 'might not be worth the hassle'' due to complexity and ROI concerns. While not every retailer has definitively 'given up', the overall direction and trends described are visible.
- Jan 1, 2025Too early to tell.““Eric and I have talked on the pod before about a prediction that there would be an AI-first agency with large bookings and very low headcount.””
While there are AI-focused agencies, clear public evidence with verifiable "large bookings" and "very low headcount" to match the specific conditions of the prediction is not yet available, making it too early to definitively judge. The industry is moving towards AI integration in agencies, but a pure 'AI-first' agency with those specific metrics is not clearly established.
- Jan 1, 2025Called it.““Tik Tok Ban: It’s gone to the Supreme Court, and President-elect Trump has come out on the app’s side.””
The TikTok ban remained a majoy legal and political issue through 2024-2025, so the broad prediction that this conversation would continue was accurate. However, Donald Trump did not become President-elect in late 2024, altering the specific scenario of the prediction.
- Sep 3, 2025Called it.“The data-sharing remedy "seems extremely unlikely to blunt Google's market share."”
By mid-2026, publicly reported search-engine market-share data show Google maintaining very high market share in general search, with only modest changes despite the remedies and the rise of generative AI-based alternatives. There is no public evidence that the specific remedy described (limited data sharing with “qualified competitors,” excluding ad data) has significantly reduced Google’s share of general search or search text ads.
- Sep 3, 2025Called it.“One frequent criticism of antitrust is that it fights the last battles; in this case the judge was cognizant of the coming battle [regarding AI] and chose to take a very light hand on the remedies within that context, Paparo says.”
The remedies imposed by the judge have indeed been characterized as light, and despite the advancements in AI, Google's core power in search has remained largely undiminished through mid-2026, aligning with Paparo's interpretation and forecast.
- Sep 3, 2025Called it.“Ari Paparo, author of the book Yield that describes Google’s ad tech business, says that the data-sharing remedy “seems extremely unlikely to blunt Google’s market share.””
As of mid-2026, Google retains overwhelming dominance in the search market. The data-sharing remedy has not observably blunted Google's market share, confirming Paparo's prediction.
2024
25 predictions- Oct 6, 2024Too early to tell.“If you think about the total cost of a DFP plus an SSP [supply side platform] and also, like I said, that it’s a bigger market, I think that the very possible outcome is that they’re forced to spin out DFP. I think that’s very possible, and it’s also a clean outcome. It’s not the search trial. It’s hard to figure out what to do with search in this market. It’s quite easy actually to just spin out DFP, the problem is solved. I think it’s very possible that will be the outcome. I would say 75% is my current estimate.”
As of August 31, 2026, there is no public record of Google having announced or implemented a spin-out of DFP / Google Ad Manager or a court/regulator formally ordering such a divestiture as a remedy in the U.S. DOJ ad-tech antitrust case. The predicted outcome has not occurred yet.
- Oct 6, 2024Missed it.“It’s quite easy actually to just spin out DFP, the problem is solved. I think it’s very possible that will be the outcome. I would say 75% is my current estimate.”
Paparo put a 75% probability on Google spinning out DFP (Google Ad Manager). It has not happened: the DOJ ad-tech remedies phase ran through 2025 with closing arguments in November 2025 and the judge signaling hesitancy to order divestiture, and as of mid-2026 GAM remains inside Google.
- Dec 16, 2024Called it.“Google's dominance is too strong. The only possible way to break this is legal.”
The outcome states 'Google was declared a monopolist,' confirming that legal action was the route to addressing its dominance.
- Dec 16, 2024Story still evolving.“Well, yeah, I think AI is the biggest change.”
The prediction is about AI being the 'biggest change' in ad tech/marketing. The provided materials suggest AI is indeed a central force, indicating this is an ongoing development consistent with the prediction.
- Oct 6, 2024Missed it.“Paraphrase: Paparo estimated a 75% chance that Google would be forced to spin out DFP as an outcome of its ad tech antitrust case.”
Paparo gave a 75% estimate that Google would be forced to spin out DFP as a remedy in the ad tech antitrust case. Judge Brinkema's remedies opinion, unsealed September 16, 2026, refused the government's structural remedies, including a contingent DFP divestiture, and ordered behavioral remedies instead. DFP and AdX remain together. Appeals remain possible, but the court's remedy did not match his forecast.
- Oct 6, 2024Missed it.“It’s very possible and it’s also a clean outcome ... it’s quite easy actually just spin out DFP the problem is solved ... I would say 75% is my current estimate”
Paparo put 75% odds on Google being forced to spin out DFP. In the ad-tech remedies opinion unsealed on Sep 16 2026, Judge Brinkema declined to order an AdX divestiture and ordered behavioral remedies instead, including Prebid integrations for AdX and DFP. DFP was not spun out.
- May 7, 2024Called it.“Engagement is the new uniques. The traffic era of publishing has ended. Nobody brags about their ComScore uniques anymore; engagement is the new North Star.”
Industry commentary, publisher strategy, and conference messaging consistently support the shift toward engagement/attention as the key KPI.
- Oct 6, 2024Missed it.“it’s very possible that will be the outcome … I would say 75% is my current estimate”
Paparo put a 75% probability on Google spinning out DFP (Google Ad Manager). It has not happened: the DOJ ad-tech remedies phase ran through 2025 with closing arguments in November 2025 and the judge signaling hesitancy to order divestiture, and as of mid-2026 GAM remains inside Google.
- Aug 30, 2024Missed it.““…we’re entering in this new period where targeting is going to be harder… what’s going on in the browser with, you know, Google is going to eventually deprecate this third‑party cookie, it’s going away at some point…””
Paparo said Google would eventually deprecate Chrome's third-party cookie. Google abandoned that plan, kept its existing user-choice approach for third-party cookies, and later retired major Privacy Sandbox replacement technologies. The announced deprecation did not happen.
- Oct 6, 2024Missed it.““It’s very possible and it’s also a clean outcome… it’s quite easy actually just spin out DFP, the problem is solved. Um and I think it’s very possible that will be the outcome, uh I would say 75% is my current estimate.””
Paparo put a 75% probability on Google being forced to spin out DFP. In September 2026, the court rejected structural divestiture of DFP or AdX and imposed behavioral remedies instead, so the forecast missed the remedy selected in the case.
- Oct 6, 2024Partially right.“If Ari Paparo had one piece of advice for everyone looking to buy ads today it would be to stop putting so much faith in measurements and look for other (read: cheaper) opportunities. I think it's very possible that will be the outcome. I would say 75% is my current estimate.”
While outcome-oriented, cheaper attention channels have grown, and AI-driven modeling is replacing traditional attribution, most large buyers still invest heavily in measurement tooling and analytics. The core behavioral shift of broadly reducing 'faith in measurements' has not materialized at scale.
- Dec 9, 2024Story still evolving.“While we haven’t seen major shifts in the agency world due to AI, I think everyone knows it’s coming,”
His backward-looking assessment for late 2024, that major shifts hadn't yet occurred, is consistent with evidence. His forward-looking statement that major shifts are 'coming' is directionally supported by increased AI usage and discussions in the industry by mid-2026, though the shift is evolutionary rather than catastrophic. No specific timeframe was given.
- Oct 6, 2024Partially right.“I think that the very possible outcome is that they’re forced to spin out DFP. … It’s quite easy actually to just spin out DFP, the problem is solved. I think it’s very possible that will be the outcome. I would say 75% is my current estimate.”
As of mid-2026, Google has not spun out DFP/Google Ad Manager. Antitrust litigation is ongoing, but no final, enforced remedy has required a DFP spin-out. While structural remedies were contemplated, no binding order to divest DFP has taken effect. The 75% probability did not materialize.
- Dec 1, 2024Called it.“TikTok didn't get banned, but it will be controlled by a US entity.”
Both halves landed. TikTok was never banned in the US, and in 2026 the divest-or-ban saga closed with the creation of the TikTok USDS Joint Venture - US-controlled, with Oracle, MGX and other American investors holding the majority and ByteDance reduced to a minority stake. Not banned, and now controlled by a US entity.
- Jan 1, 2024Called it.“Second thing to watch is AI adoption among agencies and holdcos and that’s happening and is important and the question is how big a change is it going to be to their business models. ... We’re seeing or I’ve heard that agency pitches are coming in like 20% lower in headcount already for the same business. And AI is basically doing that.”
Major holding companies have reported large AI investments and internal AI tool launches. Industry commentary and earnings calls confirm efficiency gains and some headcount rationalization driven by AI, supporting the prediction of reduced labor intensity and changing agency models.
- Jan 1, 2024Called it.“There have been two counterforces one is CTV and the other is retail media and the retail media one has proven quite significant…”
Retail media networks expanded rapidly in revenue and product sophistication. CTV spending grew significantly, with YouTube and Netflix often described as a forming duopoly in streaming.
- Jun 25, 2024Called it.“The Open Internet is trying to come up with a way to keep whatever value their media once had. Overflowing these sites with ad units and making them primarily for advertising and AI created content is one way. Another way is impression generating bots. They make up a significant portion of what 'The Open Internet' is selling — as anyone who has seen traffic logs from these sites knows.”
Many open-web publishers have resorted to increased ad load and AI-generated content, and bot traffic remains an issue. The shift to logged-in environments further validates the open internet's relative decline.
- Jun 25, 2024Called it.“The biggest trends in advertising right now are the move to advanced audiences and the use of deterministic first-party data.”
First-party data and advanced audience solutions have indeed become central to advertising strategies, validating this as a dominant trend.
- May 19, 2024Partially right.“The SaaS apocalypse is real, with caveats: Ari argues AI and vibe-coding will eliminate a significant portion of SaaS businesses, but concedes that deeply operational infrastructure; pacing algorithms, supply access, click tracking across complex stacks isn't going away via a prompt.”
While AI has pressured lighter SaaS tools, a full 'apocalypse' has not materialized. Complex infrastructure, as Paparo predicted, remains essential.
- Jan 1, 2024Called it.““Agentic Commerce Is (Still) a Collective Hallucination””
As of mid-2026, there is no evidence that fully autonomous “agents” have become the primary modality for consumer shopping at scale. Industry coverage continues to treat agentic commerce as an emerging or experimental area, not a mainstream reality.
- Dec 1, 2024Called it.“If that’s the case we would look for either a) some retailers giving up; b) generally a lot of noise about how it ‘might not be worth the hassle’; and c) aggregation platers across retailers.”
Several smaller or less-successful retail media efforts have slowed or wound down. The trade press has questioned if every retailer needs its own media network. There has been growing investment in multi-retailer retail media platforms and aggregators. This prediction is substantially accurate at the trend level.
- Feb 2, 2024Called it.“We’re moving into an era where ‘outcomes’ – not impressions or even clicks – will become the defining metric in digital advertising over the next few years, driven by AI optimization and the consolidation of media platforms. By 2025 and beyond, the platforms that can prove incremental sales or other hard outcomes will gain share at the expense of those that just deliver reach.”
From 2024–2026, major platforms and advertisers have increasingly emphasized incrementality, sales lift, and business outcomes in their narratives and product updates, especially in retail media networks, CTV/commerce integrations, and performance-oriented video/social formats. AI-driven campaign optimization and closed-loop measurement have become core selling points, consistent with Paparo's forecast.
- Dec 30, 2024Partially right.“Walmart buying Vizio was arguably the most important advertising-related M&A of 2024. It will force Amazon, Kroger, and the big streamers to accelerate their own CTV and retail-media integrations in 2025, with at least one major, defensive acquisition of a TV OEM or CTV OS.”
Large platforms and retailers have intensified investment and partnerships in CTV and retail media. However, as of mid-2026, there has not been a clearly 'defensive' acquisition of a major TV OEM or full CTV operating system by Amazon, Kroger, or a top streaming platform.
- Dec 1, 2024Called it.“Tik Tok Ban: It’s gone to the Supreme Court, and President-elect Trump has come out on the app’s side.”
The ban law did reach the Supreme Court, which upheld it unanimously in January 2025 (TikTok v. Garland) - but enforcement was then delayed repeatedly through 2025 while a US ownership deal was negotiated, and TikTok never went dark for good, consistent with Paparo reading Trump as on the app side. (Prior summary was corrected: there WAS a definitive Supreme Court ruling, and Trump was president-elect in late 2024.)
- Dec 30, 2024Called it.““Tik Tok Ban: It’s gone to the Supreme Court, and President‑elect Trump has come out on the app’s side. So the ban that seemed super‑likely in 2024 is now less likely in 2025, although the Court could still surprise us.””
Despite high political pressure, TikTok was not actually banned in the U.S. in 2025, aligning with Paparo's prediction that a ban was "less likely."
