# AI Search, Digital Ads & Retail Media - Week of June 14, 2026: Meta Passes Google in Ad Revenue, and the Open Web Feels It > Digital ads and retail media newsletter for the week of June 14, 2026. An eMarketer forecast has Meta overtaking Google in 2026 ad revenue for the first time, retail media concentrates ~90% into Amazon and Walmart, and measurement distrust hits 91% of marketers as the open web rations trust. ## AI Search, Digital Ads & Retail Media ### Week of June 14, 2026: Meta Passes Google in Ad Revenue, and the Open Web Feels It --- ## TL;DR - An eMarketer forecast making the rounds has **Meta overtaking Google in 2026 ad revenue for the first time** ($243.46B vs $239.54B), the talking point of the week, and a clean proxy for the AI-tooling-vs-keyword-bidding shift. - **Retail media's reckoning is here:** 90% of new US retail-media dollars go to Amazon and Walmart, leaving ~270 networks to fight over the scraps. The IAB just told most of them to pick a lane or get sorted. - **The open internet's whole problem in one number:** $2 CPMs vs $20 CPMs, and the difference is whether you know who your users are. Measurement distrust (91% of marketers) and signal loss were the week's connective tissue. It was a practitioner-heavy tape, lots of ecommerce operators, not much sell-side. But the operator chatter rhymed in a way that matters for the book: the walled gardens keep widening the moat, and everyone downstream is rationing trust. --- ## What's new **1. Meta is now the biggest ad seller, full stop (per the forecast everyone's quoting).** On *The Art of the Brand*'s "[Google Didn't See It Coming](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOgNy36WSrSS4VIJGbgXNSJkLcP9ePloBkN0mzcka1zzlXfVDJLylyJaY-2BvG0TZmJBNb8UQ3B5g434s8sXGaOcNfdUPyqDPTde6mCfl0dzXkIA-3D-3DWR2E_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI-2FLetGS6Ws-2BwFOVJ-2BCOqAyzmif7OwJPB88vpiiTv91PQ5QnGf9u6QVjWOjK61UzAGLbYnG1ZkHxFyRdPBZi69L5nnHn21ZgPKIUml0efv9MIXL-2F8OMzvrleYEsD-2FXLFDTw-3D-3D)" (Jun 11), the hosts walked through an eMarketer forecast pegging **Meta at $243.46B vs Google at $239.54B for 2026**, Meta's first time on top. Treat the number as a forecast quoted by brand marketers, not gospel, but the *why* is the thesis: they called Google's keyword-bid interface "prehistoric" next to Meta's "tell us what you want and our AI builds the ads." That's the Advantage+ flywheel showing up in a marketer's mouth, unprompted. **2. The IAB put a name to the retail-media shakeout.** On the *Retail Media Breakfast Club*'s "[The Retail Media Reckoning](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOidOoWEFy9y60T6PJUwiX3aGNO1ZwDp8n0Zw1lNzhJZ9Tm2OHiRiQFQz0LX3p69mkoiwuvo2I94v2fLvH9eVm7aGOHyOmViBTe2Z1f2lWckEw-3D-3DgBI-_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI3-2BKGgObifl-2B8-2FkQsJoK-2FLb-2FuxC9Op6QcsYlsmOkyNNEXIeJuyGkZq232E8FO-2BOJiQVJ3QNaLrldf-2B3vmVIul-2FL4wylO9GPsk8vGXrJWBAXcLdR4iPMrL2wcxDsBtIP7aw-3D-3D)" (Jun 11), columnist **Kiri Masters** did the napkin math: of roughly $10.5B in new US retail-media spend, **$9.4B, 90%, goes to Amazon and Walmart**, while a database tracks ~270 retail media networks worldwide chasing the other tenth. She cited the IAB's April white paper warning that "over the next 24 to 36 months, networks that don't make explicit strategic choices will be implicitly sorted by the market." Read-through: the long tail of retailer ad businesses is a value trap; the duopoly inside the channel is the trade. **3. The open web's escape hatch costs a 10x CPM.** The most quotable insider of the week was **Scott McKinley**, Truthset founder and former Nielsen EVP, on *AdTechGod Pod*'s "[The Data Quality Crisis in Digital Advertising](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOhxSV83bfk5KuAarmiWHB6nPRYtSDE4GAN7Z6Mr8hpW1g7LurH6oEO-2BfRHBG9IRekAGub1SEZv1Ax6hUEoVBWfYFlN-2FL-2FWqqv9i-2Fgk0BHGQ-2FA-3D-3D7Zk0_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI0wKFKbvHOciUEFmEpb94Qzcz9Y95LqjgolBg2dJ555OVkFLwQr80O2J0dd9xN01qaZ92cx2ozHV0ARas276jmAk-2FhtnIm9GtG7OPyVheupKkvpvriP-2FyCQsCeYIU7AApQ-3D-3D)" (Jun 9). His line, "the way you get a $20 CPM instead of a $2 CPM is you authenticate your users," is the whole walled-garden story in a sentence. He called digital ads "a $700 billion industry that runs on faith," warned that leaning on IP addresses as cookies crumble is a trap (a SIM study put IP-to-email accuracy at **13%**, IP-to-postal at **16%**), and noted data degrades from ~90% to **22–23% accuracy** across programmatic supply-chain hops. This is the structural tailwind for identity and first-party-data owners and the structural headwind for open-exchange CPMs. **4. Nobody trusts the platforms' own scorecards.** On *Marketecture* (Jun 8), **DoubleVerify** Chief Strategy Officer **Doug Campbell**, alongside Affinity's Damian Garbaccio, unveiled a survey finding [**91% of marketers believe platform-reported results are overstated**](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOhsjzhK-2BpdCbFiaoqly5QlMSxGPwsP1ayvmUQrWTRPOxh68bXnJ414M51Udvmb-2F8W0h-2Fvtz0hpGKbM2tXop8gifVcwcJwFq-2BRtNkrN5qJ-2Bnzg-3D-3DsXi8_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI6QDnWdtqmcCl5MkxbVC8IA1r2oFrQy19LdxxEuoGVdE33-2BVwF4RxI9wgI7w1Nqu4O2XWr1REBBsb58HdiZpqb1QFDwbG1B9HR2OqQFL0JaQNWmbXX593QgV-2Br64idRZVQ-3D-3D) in some way. Campbell's framing, it's "still pretty difficult to connect ad exposure to actual purchase data" in real time, plus the double-counting gag (one platform claims 100 sales, the register says 90, a second platform also claims 90 = 180) is the entire bull case for independent, outcomes-based measurement. Good for the measurement vendors; awkward for the self-grading gardens. **5. The upfront is being re-plumbed as enterprise software.** Two angles converged. On *The Digiday Podcast*'s "[How automation and AI are rewriting the upfront marketplace](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOjJ-2B9dQ1LzV4YpCViaL-2FDOShuOlfcd9pZF2-2FkSSpnILMUiVPY7jnE12h6X-2FY8Sjhj-2BvStUWsOXUvHXbA8txYRdAAf8X4fXcwXdUN-2FxEnfBRDw-3D-3D031T_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPIwepVqAUUwh7IxKZCU0xnQw1NPe2N0zBmtevMCp-2FHTDnfr1hH4LHalL3B-2F2-2FcQfSapnI3YooQquB4mok9wQ-2BcCt5EDPvNWuJJSQ6-2FyTdYQpfazLqyfsMPIv2AoitNuFBqQ-3D-3D)" (Jun 9), editors noted Amazon and YouTube increasingly treat the upfront "as enterprise tech deals more so than traditional media deals," you reach YouTube programmatically through DV360, and Amazon keeps funneling buyers into its DSP. Meanwhile NBCU's **Allison Levin** on *Next in Media* (Jun 9) gave the [seller's side of the same story](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOhF1KlWdzOOzcbxrk1pWuUwSWb7PL8wV8XVOkfVdXkfzFIyT3HK5jzjWIPzEPuLAjTXsjZo9nylyzLBIoCwtgEhJw4aQffxIwRRshbdx2gO3Q-3D-3DMvCb_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI-2FGpYaI9GD704k-2F82ihhwznpTozBy7GnhCMTJq4495T15-2B3ODxek0GfFcz5ZMRUaA30G30UujpRcS3Gp0KTvnvy7olW2AZGnwWpiqWbJKJvaChORrcZWUmFTZbYhSgWGww-3D-3D): scarce, IP-driven inventory (live sports, Bravo, Love Island) is commanding the demand, NBA viewership ran **+100% YoY** on comparable games, and, the number worth tattooing on a CTV bull, **"still 70% of all impressions are happening in linear."** --- ## The debate **Bull, the pie is growing and the gardens own it.** AI ad tooling (Advantage+, Performance Max) is lowering the skill floor and pulling SMB and long-tail budgets onto Meta and Google, which is exactly how Meta laps Google on revenue. Retail media is concentrating into a high-margin Amazon/Walmart duopoly. And as signal loss bites, the platforms with logged-in users and first-party data capture the $20 CPMs while the open web bleeds. Owning the gardens *and* the measurement/identity layer that polices them is the cleanest expression of the trade. **Bear, saturation, distrust, and the AI-search overhang.** Retail media's growth is 90% spoken for, so ~270 also-rans face a margin reckoning that could pressure ad-tech vendors levered to the long tail. The 91%-overstated-results stat says budgets are being defended on faith, not proof, fragile if a downturn forces marketers to actually audit ROAS. CTV's land grab is splintering inventory and pricing. And the quiet overhang on Google: if discovery keeps migrating to ChatGPT/Gemini/Perplexity, the most profitable real estate in advertising, the search results page, gets disintermediated faster than AI-search ad formats can backfill it. --- ## Stocks in play **Meta (META)**, *Bull:* Now the #1 ad seller on the forecast, with Advantage+ doing the heavy lifting; operators describe it as the path of least resistance for budgets. *Bear:* That same automation "optimizes for Meta's goals, not the advertiser's" (more below), and reported results sit inside the 91%-distrust bucket. *Watch:* Advantage+ adoption/revenue-mix commentary and any impression-vs-price disclosure into Q2. **Alphabet / Google (GOOGL)**, *Bull:* Google Shopping is still described as highly incremental and outperforming Performance Max for ecommerce, and DV360 is the toll booth for programmatic YouTube. *Bear:* The week's narrative was Google ceding the ad crown and search relevance to AI chat. *Watch:* Search ad resilience vs. AI-search erosion; YouTube's push to monetize upper-funnel video (a new branded-search beta metric is rolling out). **Amazon (AMZN)**, *Bull:* Half of the retail-media duopoly capturing ~90% of growth, now turning Sponsored Brands into multi-product "discovery" surfaces and pushing buyers into its DSP for premium CTV. *Bear:* Coverage this week was practitioner-tactical, not investor-grade, no fresh read on segment economics. *Watch:* Ad-segment growth and DSP attach rate; the Brand Gallery/AI-visual-search rollout as a CPM lever. **The Trade Desk (TTD)**, No dedicated coverage by name this week. The relevant tape is indirect and mixed: the open-web CPM/identity problem (McKinley) is exactly the gap TTD's UID2/Kokai pitch addresses, but it's also the bear case if authentication consolidates toward the gardens. --- ## Read-throughs - **Emerging-platform ad players (RDDT, PINS, SNAP, APP):** Quiet week, **zero dedicated episodes** on Reddit, Pinterest, Snap, or AppLovin/AXON as stocks. One indirect Reddit positive: operators note AI-search engines increasingly lean on Reddit threads to understand and recommend products ([Selling on Giants](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOhqRNIMNYgyCRnNUUJqV79VJkpuMTCGXc4eBxU1gca4PrnrcAc-2Fc0-2BGRsiinOdss2OajNBHgdTOUW33rl08cb0tgmG0MQ1cJs-2BB2XbWAILvqA-3D-3DsEcU_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI-2BrEECFH3wytvEu-2F5unlikI6NwqS72CT07hrJhp5QQ1jmTAEI6JgKcICvYNdgHzDhLy8DVzUEwNBjBfzN9YjQ34MNAq-2FRt7AVAYofTuZgE1D0hswY8igs0yKJti3xi02Qg-3D-3D), Jun 9), a data-licensing/visibility tailwind, not an ad-revenue one. - **Retail-media operators (WMT, CART):** Walmart shows up only as the co-monopolist in Masters' 90% stat; **no Walmart Connect or Instacart episode by name.** The actionable layer is structural, own the duopoly, fade the long tail. - **CTV / streaming:** NBCU's Levin is the operator anchor, linear still 70% of impressions, scatter at record-low availability, and a shift to "parallel marketplaces" blending reserved upfront with always-on programmatic. Bullish for premium, IP-heavy inventory; a reminder that "linear is dead" is years early. - **Ad-measurement vendors (DV, IAS, RAMP, CRTO, MGNI, PUBM):** No stock-level discussion, but the *theme* was loud, DoubleVerify's exec on the 91%-distrust survey, and the broader signal-loss/identity case. Note the consolidation tell Masters flagged: **Publicis acquiring LiveRamp**, folding a neutral-identity backbone into a holding company. - **AI ad-tooling, with a caveat:** A Meta-ads strategist on [Powerful Women Rising](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOhHe6kaI83xpobYXX-2BHJ42sh0CtNZpjSiHH0sH1HxMrSSkG7jO5Gq64-2FUNl59ZTWTyC7aSplRJA8PLjxaMDknXJqUaSbZaE-2Fk4EsHRbYmOa-2Bw-3D-3DrVsq_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI7FhVeXBtlbsNO02VGtefJjr0SgW1wQFEUdRhseWndUuuRgXyUgt9vHc7nqxR3vC4kD9pTpapy8RDOIZORgN7RNmMRmBEDVjIYkkjAG7yMZgLzB5Pa1vLBmaPzIFSIzrBw-3D-3D) (Jun 8) warned Advantage+ pushed a daily budget from "$40 to $439" with no sense of the advertiser's constraints, useful color that the tooling optimizes for platform spend. And on [DTC Podcast](http://url7324.matterfact.com/ls/click?upn=u001.idHmPrr2Geh7KYLAsTy7NkrIVb-2FgA4pmf2rMXQwGcOhWJAxeetJHgui8EnfULu7Sr-2F02XriUUQyl9-2BMniCPWFS6obSrsDfZ2sCduabiKbVTjE1ESF8nusxzDLkTsrsVJmjsV5GPwM9rkIZfredBiow-3D-3DWeEK_7mLGwmUci-2BLaXswv9WX1yTgqn3Wad-2FotHhzHgSNAZbU0ojgoOrEP8jMx6y8-2FhHnRxCUDl5rhh5QY6u5DfQWPI9tFi-2FjDemQAJPrLfWYUY-2BAyD6KG-2BhcQ8g-2FI3YGB7v2LPGRgtj7BAsi2yZmWW8OgpqzsxQxxNGFXc-2B9vcL5noGRhGFGmDYvtBkgrnTHMsep8c4Iynx-2BHvkaUYImysfu5EA-3D-3D) (Jun 12), Pilothouse called ROAS a lagging indicator that breeds a discount "death spiral," the practitioner version of the measurement-trust problem. --- ## What changed vs last week Honest read: a **thin, operator-skewed tape**, heavy on ecommerce and DTC practitioners, light on sell-side or company-specific investor commentary. No earnings, no guidance, no analyst day. What *did* sharpen is the through-line: the same forces, AI tooling pulling budgets into the gardens, retail-media concentration, and signal-loss-driven measurement distrust, were independently restated by a brand-marketer panel, a retail-media columnist, a former Nielsen EVP, a DoubleVerify exec, and an NBCU ad chief. When five unrelated voices in one week describe the same moat from five angles, that's the signal worth carrying into the book. No major narrative reversals; the walled-garden thesis got reinforced, not challenged. *Coverage note: zero dedicated episodes this week on RDDT, PINS, SNAP, CART, AppLovin/AXON (as a stock), TTD by name, or the measurement names (DV/IAS/RAMP/CRTO/MGNI/PUBM) as named tickers. Where those appear above, it's via theme or indirect read-through, flagged as such.*