[scuttleslops] AMZN, CME, IQV, WIX, SNOW, VEEV, GWRE
[scuttleslops] AMZN, CME, IQV, WIX, SNOW, VEEV, GWRE
Mobile Dev Memo Podcast — Adam Epstein on Amazon’s advertising advantages, June 4, 2026: Amazon’s identity-spine moat and the structural decline of The Trade Desk
Adam Epstein (CEO of GG; ex-Perpetua) laid out why Amazon’s advertising business is becoming much more than retail media: it is evolving into a scaled programmatic, CTV, and identity infrastructure business. Epstein’s core claim is that Amazon’s path from roughly $50 billion in ad revenue toward $100 billion will come “primarily from Amazon’s DSP,” (Demand-Side Platform – software that lets advertisers buy ad inventory automatically across many publishers through real-time bidding, rather than negotiating directly with each one) because sponsored ads on Amazon.com are constrained by inventory and seller profitability, while CTV (Connected TV – TV delivered over the internet) remains a large and growing market.
The big strategic asset is Amazon’s authenticated identity graph, which now deterministically reaches “90 percent of US households,” giving Amazon an “unparalleled advantage” in building a deduplicated reach-and-frequency layer for TV advertisers. The Roku partnership is central to that advantage because Roku fills gaps in Amazon’s household identity coverage through its massive CTV hardware footprint, creating what Epstein calls a “unified identity spine.”
The strategic implication is that Amazon can tell large brands and agencies to centralize TV buying in Amazon DSP because it has a better signal layer than other DSPs for measuring and controlling reach and frequency across CTV. Epstein argues that Amazon has already done this well with endemic advertisers, especially e-commerce and CPG brands, but the larger growth opportunity is non-endemic categories like auto, pharma, and financial services.
Epstein views Amazon’s strategy as the opposite of the traditional walled-garden model: instead of keeping all data inside Amazon-owned inventory, Amazon is projecting its identity and measurement layer into premium third-party inventory. Prime Video is the anchor because it is exclusive to Amazon DSP and gives advertisers a reason to enter the platform, but Amazon’s broader pitch is that once a buyer is there, they should also centralize Disney, Netflix, NBC, Paramount, and other premium CTV buys in Amazon. He describes Amazon’s positioning as becoming “the premium CTV DSP in the world,” in contrast to Google’s YouTube-driven CTV pitch. This creates a serious threat to The Trade Desk, which Epstein says lacks both proprietary inventory and proprietary data.
Epstein is especially skeptical of The Trade Desk’s competitive position, saying Amazon and Google have created “existential doubt” about Trade Desk because both can use owned inventory plus proprietary data to pull CTV budgets into their DSPs. He criticizes Trade Desk CEO Jeff Green’s public comments about Amazon, particularly the idea that Amazon DSP is not strategically important to Amazon, calling that “categorically untrue.” Still, Epstein does not argue Trade Desk disappears; he notes that it remains entrenched with large agencies and brands and has a real advantage in log-level data, which matters in categories like pharma where advertisers need to match impression-level exposure to downstream outcomes.
Amazon is still the dominant retail media network, likely representing “70 to 80 percent of all retail media,” while also owning the fastest-growing scaled CTV DSP and the largest online store in the world. But he also identifies blind spots: Amazon does not have a meaningful position in social commerce, and unlike Meta or YouTube, it cannot endlessly generate organic content inventory.
Epstein argues that if OpenAI wants to build a large-scale advertising and e-commerce business, it likely needs Amazon because a serious commerce product requires reliable catalog, inventory, pricing, delivery, and purchase data. He says a Shopify-enabled ChatGPT commerce product would be “simply niche,” because Amazon still represents about half of U.S. e-commerce. At the same time, Amazon faces a disruption dilemma because if shopping shifts from Amazon search pages into Rufus or third-party chatbots, Amazon could lose keyword and product-detail-page ad revenue. Epstein’s view is that the natural solution is an Amazon-OpenAI advertising or commerce partnership, where Amazon’s product listing ads, DSP, identity spine, and commerce infrastructure extend into ChatGPT-like surfaces.
Source: Podcast: Understanding Amazon’s advertising advantages (with Adam Epstein)
CME Group (CME) — Piper Sandler Conference, June 4, 2026: Institutional risk-management moat, perpetuals as systemic risk
Terry Duffy framed CME’s core competitive position as an institutional risk-management utility, not a retail speculation venue, and that distinction drove almost every answer in the interview. The most forceful discussion was around the CFTC’s approval of Kalshi’s Bitcoin perpetual futures contract, which Duffy argued was procedurally and substantively wrong because the CFTC treated a “novel and complex” product with less scrutiny than even a normal self-certification process.
His central legal argument was that the Commodity Exchange Act defines a futures contract as one with “a delivery date at a later date or an expiration date at a later date,” while a perpetual contract “never ends,” so in his view “that is not a futures contract” and is closer to a swap. Duffy’s risk argument was even more direct: perpetuals in Europe can trade with “20x to 250x leverage,” whereas CME’s institutional crypto leverage is closer to “5:1,” and he questioned how such leverage can be sustainable under a retail auto-liquidation model.