The Trade Desk’s AI gamble has moved from experiment to enterprise: Kokai is live, adoption has accelerated, and the company is consciously steering its product and go-to-market strategy around the very forces — generative search and AI-driven ecosystems — that threaten the open web it sells into. (thetradedesk.com, fool.com)
The Trade Desk (TTD) long pitched itself as the independent demand-side platform (DSP) for advertisers who prefer transparency and control outside Big Tech's walled gardens. That positioning became more than rhetoric when the company launched Kokai in June 2023 — a distributed deep‑learning architecture that the company says ingests the equivalent of more than 13 million ad impressions per second and applies thousands of signals for prediction, scoring, and optimization. Kokai is explicitly designed to act as a copilot for marketers: humans set objectives and guardrails, AI drives allocation, bidding, and measurement. (thetradedesk.com, fool.com)
Through 2024 and into 2025, Trade Desk executives reported rapid migration to Kokai: management told investors that about two‑thirds to three‑quarters of client spend had moved to Kokai by early to mid‑2025, and the company has publicly targeted full client adoption by the end of the year. Those numbers, combined with early performance case studies Trade Desk cites, are core to the company’s argument that AI will deepen client stickiness rather than hollow it out. (fool.com, investing.com)
At the same time, the advertising market itself is being reshaped by two powerful currents. First, AI-powered search and summarization features from large platforms are reducing referral traffic to publishers and compressing the pool of open‑web impressions that programmatic buyers have historically relied on. Multiple independent studies and industry reports show dramatic reductions in link clicks when AI summaries appear in search results. Second, major platform owners are beginning to monetize those AI experiences — inserting ads directly into chatbot answers and AI overviews — which concentrates advertiser demand inside a smaller number of ecosystems. (searchengineland.com, blog.google)
These forces create an ambiguous picture for Trade Desk: Kokai makes the platform technically stronger and more efficient, but the upstream shift in supply and attention could reduce the raw inventory pool and shift budgets toward the very walled gardens Trade Desk wants to compete with.
The strategic pitch is clear: automate the mechanical and micro‑optimizations so human strategists can focus on creative strategy, brand objectives and higher‑level planning. In practice, Trade Desk emphasizes that Kokai is a copilot, not a replacement. Advertisers set goals and constraints; Kokai optimizes towards those targets in real time. This model mirrors an industrywide shift toward automation and outcome‑based buying, but Trade Desk frames Kokai as uniquely tuned to the open internet and to privacy‑first identity solutions. (thetradedesk.com, fool.com)
To be precise: a drop in referral clicks is not a one‑to‑one decline in all ad impressions. Many publishers still host large audiences through direct visits, apps, social distribution and subscription models. But for ad inventory that relies heavily on search discovery and organic referrals, the impact is acute. Publishers whose revenue model is tightly coupled to referral traffic are most exposed. (forbes.com, wsj.com)
That’s the threat at scale: smaller supply and more concentrated demand can raise CPMs for available open‑web inventory, shrink reach for some audiences, and shift allocation opportunities toward those who control AI search inventory and the associated first‑party data.
Yet the external environment is hostile in ways product upgrades can’t fully control. AI search's capacity to keep users inside summary experiences and the growing practice of inserting ads into those interfaces represent a structural shift that favors companies controlling both discovery and monetization. If those ecosystems keep expanding and advertisers migrate significant budgets to them, Trade Desk may find its role more limited or more commoditized than it hopes.
This is a classic technology strategic fork: invest early and deeply in AI to improve intrinsic value and customer lock‑in, while aggressively expanding into channels that are less likely to be cannibalized. Trade Desk is doing both — which is the right playbook — but execution risk is high, competition is intense, and the regulatory environment adds further uncertainty.
For readers focused on either client strategy or capital allocation: monitor Kokai adoption and performance metrics closely, watch how ad inventory dynamics evolve as AI search scales, and judge Trade Desk’s ability to turn product differentiation into sustainable margin expansion. If Kokai becomes the industry’s de facto copilot for open‑internet buying and Trade Desk captures creative and measurement value as planned, the company stands to benefit. If AI search concentrates attention and dollars inside the major platform owners faster than open‑web alternatives can adapt, the company will face meaningful headwinds.
Trade Desk’s path is neither binary nor predetermined. The company has built a strong technical response and a coherent channel strategy — but its fate will be decided at the intersection of execution, advertiser behavior, and how rapidly the big AI players monetize their discovery layers. The coming 12–24 months will tell if Kokai is a growth engine or an insurance policy against a shrinking open web. (thetradedesk.com, fool.com, blog.google, searchengineland.com)
Source: The Globe and Mail AI at The Trade Desk: Risk or Opportunity?
Background
The Trade Desk (TTD) long pitched itself as the independent demand-side platform (DSP) for advertisers who prefer transparency and control outside Big Tech's walled gardens. That positioning became more than rhetoric when the company launched Kokai in June 2023 — a distributed deep‑learning architecture that the company says ingests the equivalent of more than 13 million ad impressions per second and applies thousands of signals for prediction, scoring, and optimization. Kokai is explicitly designed to act as a copilot for marketers: humans set objectives and guardrails, AI drives allocation, bidding, and measurement. (thetradedesk.com, fool.com)Through 2024 and into 2025, Trade Desk executives reported rapid migration to Kokai: management told investors that about two‑thirds to three‑quarters of client spend had moved to Kokai by early to mid‑2025, and the company has publicly targeted full client adoption by the end of the year. Those numbers, combined with early performance case studies Trade Desk cites, are core to the company’s argument that AI will deepen client stickiness rather than hollow it out. (fool.com, investing.com)
At the same time, the advertising market itself is being reshaped by two powerful currents. First, AI-powered search and summarization features from large platforms are reducing referral traffic to publishers and compressing the pool of open‑web impressions that programmatic buyers have historically relied on. Multiple independent studies and industry reports show dramatic reductions in link clicks when AI summaries appear in search results. Second, major platform owners are beginning to monetize those AI experiences — inserting ads directly into chatbot answers and AI overviews — which concentrates advertiser demand inside a smaller number of ecosystems. (searchengineland.com, blog.google)
These forces create an ambiguous picture for Trade Desk: Kokai makes the platform technically stronger and more efficient, but the upstream shift in supply and attention could reduce the raw inventory pool and shift budgets toward the very walled gardens Trade Desk wants to compete with.
How Kokai works — and why Trade Desk thinks it matters
Distributed AI as a trading co‑pilot
Kokai builds on Trade Desk’s earlier AI work (Koa) but expands deep learning across the full buying stack: impression scoring, predictive clearing (estimating auction dynamics), dynamic audience segmentation, budget pacing and more. The company’s launch materials describe Kokai analyzing “more than 13 million advertising impressions every second,” each carrying thousands of signals that can be evaluated in milliseconds to decide whether to bid and how much. That capability is the technical backbone for features such as Koa Audiences, TV Quality Index, Retail Sales Index and other measurement and optimization tiles inside the Kokai UI. (thetradedesk.com, fool.com)The strategic pitch is clear: automate the mechanical and micro‑optimizations so human strategists can focus on creative strategy, brand objectives and higher‑level planning. In practice, Trade Desk emphasizes that Kokai is a copilot, not a replacement. Advertisers set goals and constraints; Kokai optimizes towards those targets in real time. This model mirrors an industrywide shift toward automation and outcome‑based buying, but Trade Desk frames Kokai as uniquely tuned to the open internet and to privacy‑first identity solutions. (thetradedesk.com, fool.com)
Early performance and customer migration
Trade Desk’s investor commentary and earnings transcripts highlight client case studies — Samsung and Cashrewards among them — showing multitudes of percentage improvements in reach and acquisition metrics. Management also reported that the “bulk” of platform spend had migrated to Kokai and that spend from Kokai adopters was accelerating faster than from legacy users. Those claims are consistent across multiple earnings transcripts and company presentations in 2024–2025. While company case studies should be read with healthy skepticism, the consistent message from different quarters is that Kokai materially changed campaign performance for many early adopters. (fool.com, investing.com)The downside: AI search and the shrinking open web
Measurable drops in referral traffic and click‑throughs
Independent research shows that AI search features significantly reduce clicks to publisher sites. The Pew Research Center’s analysis of U.S. users found that searches triggering AI “overviews” produced link clicks at roughly 8% of sessions, compared with about 15% when no overview appeared — a near‑halving of click propensity. Other analytics firms reporting on publisher traffic trends note year‑over‑year declines for major outlets that range broadly but can be severe for individual titles. These shifts translate directly into fewer ad‑supported impressions on the open web. (searchengineland.com, euronews.com)To be precise: a drop in referral clicks is not a one‑to‑one decline in all ad impressions. Many publishers still host large audiences through direct visits, apps, social distribution and subscription models. But for ad inventory that relies heavily on search discovery and organic referrals, the impact is acute. Publishers whose revenue model is tightly coupled to referral traffic are most exposed. (forbes.com, wsj.com)
AI platforms become ad destinations
Beyond reducing referral traffic, AI search products are becoming ad surfaces themselves. Google confirmed expansion of ads inside AI Overviews and its new “AI Mode,” making existing Search, Shopping and Performance Max campaigns eligible for placements inside AI responses. Microsoft has similarly integrated ads into Copilot and associated generative search interfaces and is developing ad formats tailored to conversational contexts. When AI answers begin to display sponsored recommendations and shopping placements, advertiser budgets naturally follow those attention pockets. That process risks concentrating demand inside a few platform owners and carving away share from independent DSPs’ open‑internet inventory. (blog.google, about.ads.microsoft.com)Why this matters to Trade Desk’s core product
The Trade Desk’s value proposition historically rests on three pillars: access to scale inventory across the open internet, transparent bidding and measurement, and an independent stack that lets advertisers avoid Big Tech’s vertically integrated ad stacks. If AI search reduces scale (fewer impressions to buy) and shifts advertisers to platforms that can place ads inside AI answers, then the addressable inventory pool and the bargaining leverage of an independent DSP could erode.That’s the threat at scale: smaller supply and more concentrated demand can raise CPMs for available open‑web inventory, shrink reach for some audiences, and shift allocation opportunities toward those who control AI search inventory and the associated first‑party data.
Trade Desk’s defensive and offensive moves
Double down on premium, curated supply
Trade Desk has been explicit about pushing curated, premium inventory with initiatives such as OpenPath and its Sellers and Publishers 500+ ranking — attempts to gather brand‑safe, high‑quality placements that advertisers will still pay to reach, regardless of search referral flows. This is both a defensive tactic (protecting margins and brand suitability) and an offensive one (creating unique, measurable inventory the company can promote to clients). (investors.thetradedesk.com, thetradedesk.com)Prioritize CTV and retail media
Trade Desk has emphasized connected TV (CTV) and retail media as growth channels that are less dependent on search referrals and thus relatively insulated from AI‑search cannibalization. On the numbers: industry forecasts and surveys show continued double‑digit growth in CTV ad spending and robust year‑over‑year increases in retail media budgets, with retail networks forecast to add tens of billions in global spend in the coming years. For Trade Desk, which already has a strong CTV play, this remains a strategic hedge. (content-na1.emarketer.com)Extend into creative with generative AI partners
Trade Desk’s Kokai ecosystem now includes partnerships with generative‑AI creative vendors such as Rembrand, Spaceback, Bunny Studio and others. The Rembrand integration is notable because it enables virtual product placements and spatially aware brand insertions in video content — effectively pushing Trade Desk to capture a slice of creative value rather than just media buying. If executed at scale, that could expand the company’s TAM and make Kokai an attractive one‑stop solution for both creative and buying workflows. (thetradedesk.com)Strengths: what Trade Desk is doing well
- Platform modernization at scale. Kokai represents a substantial engineering effort to deploy distributed deep learning across programmatic workflows, and management has demonstrated traction with adoption and case studies. (thetradedesk.com, fool.com)
- Curated supply and transparency plays. OpenPath and the Sellers & Publishers 500+ are concrete products that help advertisers find premium, measurable inventory — important counterweights if open‑web scale softens. (investors.thetradedesk.com, thetradedesk.com)
- Channel diversification into CTV and retail media. Both channels are outpacing many others in advertiser interest, and Trade Desk’s strong CTV footprint positions it to capture an outsized share of programmatic video dollars. (content-na1.emarketer.com, emarketer.com)
- An early move into creative tooling. By aggregating generative‑AI creative partners inside Kokai, Trade Desk is trying to capture more of the “ad creation to delivery” funnel — a strategic move that reduces dependency on third‑party creative vendors and increases value per customer. (thetradedesk.com)
Risks and unanswered questions
1) Inventory compression and pricing pressure
If AI search permanently reduces referral traffic and certain open‑web impressions disappear or become harder to target, advertisers may either (a) bid more aggressively for the remaining high‑quality open impressions — raising CPMs — or (b) allocate budgets to AI ecosystems that own attention. Both outcomes could reduce the efficiency and reach of campaigns that rely on programmatic open‑web buys, tightening margins for agencies and advertisers. This structural shift would be a net headwind for independent DSPs if not offset elsewhere. Evidence that AI Overviews materially cut click‑throughs is already documented by multiple independent analysts. (searchengineland.com, euronews.com)2) Platform monetization of AI experiences
When Google, Microsoft and other AI providers put ads inside their generative responses, advertisers will have a low friction way to place budgets there. Google’s own documentation confirms that Performance Max, Shopping and Search campaigns can be eligible to appear inside AI Overviews and AI Mode; Microsoft is actively positioning Copilot as an ad surface. Those mechanics favor companies that control both the AI discovery layer and the ad inventory, a structural advantage Trade Desk cannot replicate. (blog.google, about.ads.microsoft.com)3) Client and agency friction with Kokai
Not all customers embrace forced automation. There have been industry reports and anecdotal accounts of agencies pushing back on Kokai migrations, citing workflow differences and trader‑efficiency concerns. If a meaningful subset of high‑spend advertisers resists full migration due to functionality or control issues, that could slow the monetization curve and create churn risk. The company’s public numbers on adoption are strong, but independent industry commentary suggests operational frictions exist. These dissenting signals deserve attention. (ppc.land, investing.com)4) Measurement reliability in a fragmented landscape
AI‑driven discovery, zero‑click search results and platform‑native ads complicate measurement. Trade Desk’s TV Quality Index and Retail Sales Index are responses to measurement opacity, but as the landscape fragments — with closed AI experiences and proprietary retail media networks — building a cross‑platform truth set will be harder. Advertisers will demand consistent, auditable measurement; failing to deliver could tilt budgets toward vendors who control the measurement and inventory stack. (thetradedesk.com, content-na1.emarketer.com)5) Regulatory and legal uncertainty
Major platform behavior around AI search has attracted scrutiny from regulators and publishers. Antitrust litigation, content‑licensing disputes and evolving rules around LLM training data could alter the economics of AI search over time — potentially creating either a headwind or a tailwind for independent ad tech players depending on outcomes. Investors and clients should treat regulatory risk as a persistent, non‑linear factor. (investors.com, theguardian.com)What this means for advertisers and investors
For advertisers
- Reassess channel mix: prioritize channels where measurement, control and inventory quality align with campaign goals (CTV, retail media, curated open web).
- Test and measure Kokai’s automation on controlled budgets before shifting core spend; use holdout experiments to verify vendor claims.
- Build a creative + media workflow that can take advantage of generative tools while preserving brand safety and quality controls.
For investors
The Trade Desk’s strategy attempts to straddle two simultaneous trends: use AI to widen platform defensibility, and shift buyer focus toward channels less exposed to AI search disruption. That dual approach is strategically coherent, but not risk‑free.- Positive scenario: Kokai drives measurable performance gains, client stickiness increases, and Trade Desk captures market share in CTV and retail media, offsetting any open‑web compression.
- Negative scenario: a sustained migration of discovery and purchasing to integrated AI search experiences concentrates ad budgets in platform owners and reduces the long‑term addressable inventory for independent DSPs.
- The percentage of client spend running through Kokai and whether client retention and budget expansion metrics improve for Kokai adopters. (fool.com)
- Trends in open‑web supply and CPMs for premium inventory — is supply compression materializing, or is premium inventory holding pricing and scale? (thetradedesk.com)
- The pace at which AI search features monetize with ads and whether those placements materially shift advertiser demand away from open‑web buys. (searchengineland.com, about.ads.microsoft.com)
Final assessment — risk, opportunity, and the line between them
Trade Desk faces a pivotal moment that could define its trajectory for years. Kokai is both a defensive necessity and an offensive bet: it modernizes the product, embeds AI deep into the stack, and creates new levers for measurement and creative orchestration. Those are meaningful strengths that can translate into higher client retention and share gains in programmatic CTV and retail media.Yet the external environment is hostile in ways product upgrades can’t fully control. AI search's capacity to keep users inside summary experiences and the growing practice of inserting ads into those interfaces represent a structural shift that favors companies controlling both discovery and monetization. If those ecosystems keep expanding and advertisers migrate significant budgets to them, Trade Desk may find its role more limited or more commoditized than it hopes.
This is a classic technology strategic fork: invest early and deeply in AI to improve intrinsic value and customer lock‑in, while aggressively expanding into channels that are less likely to be cannibalized. Trade Desk is doing both — which is the right playbook — but execution risk is high, competition is intense, and the regulatory environment adds further uncertainty.
For readers focused on either client strategy or capital allocation: monitor Kokai adoption and performance metrics closely, watch how ad inventory dynamics evolve as AI search scales, and judge Trade Desk’s ability to turn product differentiation into sustainable margin expansion. If Kokai becomes the industry’s de facto copilot for open‑internet buying and Trade Desk captures creative and measurement value as planned, the company stands to benefit. If AI search concentrates attention and dollars inside the major platform owners faster than open‑web alternatives can adapt, the company will face meaningful headwinds.
Trade Desk’s path is neither binary nor predetermined. The company has built a strong technical response and a coherent channel strategy — but its fate will be decided at the intersection of execution, advertiser behavior, and how rapidly the big AI players monetize their discovery layers. The coming 12–24 months will tell if Kokai is a growth engine or an insurance policy against a shrinking open web. (thetradedesk.com, fool.com, blog.google, searchengineland.com)
Source: The Globe and Mail AI at The Trade Desk: Risk or Opportunity?