Tech Explained: Here’s a simplified explanation of the latest technology update around Tech Explained: CEO Says AI Agents Will Enhance Ad Tech, Not Replace It in Simple Termsand what it means for users..
Digital Turbine logo
AI agents will enhance ad tech, not replace it: Executives at Digital Turbine, Verve and LiveRamp said agents are being used to boost efficiency across QA, campaign setup, targeting and optimization (e.g., DT iQ, Verve’s Atom, LiveRamp’s agentic orchestration) rather than disintermediate the ad-stack.
Data and privacy are decisive: Firms with superior on-device signals and distribution are best positioned as AI “fuel,” and secure environments like clean rooms are becoming essential for training models and enabling multi-party agent workflows while protecting privacy.
Measurement and channel dynamics are evolving: deterministic matching remains preferred when available while probabilistic models fill gaps, and ad spend is migrating toward in-app, CTV, retail/commerce media and emerging LLM-driven surfaces as search monetization shifts.
Executives from Digital Turbine (NASDAQ:APPS), Verve, and LiveRamp discussed how artificial intelligence is reshaping advertising workflows, measurement, and data strategy during a Roth Capital Partners event moderated by internet analyst Rohit Kulkarni. The panel repeatedly framed AI—including the fast-emerging topic of “agents”—as an efficiency and performance enhancer rather than a wholesale replacement for the existing ad tech stack.
Digital Turbine CEO Bill Stone said AI is being used “across our entire business,” citing applications in QA, coding, back office automation (including responses such as RFPs), and automating parts of campaign setup and creative. He said it can be difficult to isolate ROI from any single AI initiative, but pointed to operating leverage as evidence of broad-based efficiency gains, referencing reported 25% gross profit growth in the company’s most recent quarter alongside declining operating expenses.
On the customer-facing side, Stone said the company’s AI and machine learning platform, branded DT iQ, is used to improve targeting and deliver better return on ad spend, which in turn supports increased customer spending.
Verve CEO Remco Westermann also emphasized AI-driven efficiency, saying the technology helps reduce costs and accelerate workflows such as sales preparation and programming. He described AI as central to revenue initiatives as well, highlighting Verve’s “Atom” SDK on device that collects data in a privacy-compliant way and uses on-device AI to optimize outcomes. Westermann stressed that AI’s effectiveness depends on data quality and availability.
LiveRamp CFO Lauren Dillard outlined recent AI-related product enhancements, positioning LiveRamp as software that helps companies use data to deliver and personalize advertising by connecting data across an ecosystem of media buying platforms and partners. She highlighted three areas:
AI segmentation that allows marketers to build audience segments using natural-language prompts across first-, second-, and third-party data.
Marketplace expansion in which LiveRamp’s third-party data marketplace is increasingly used not just for personalization, but also to make “models and agents” available to customers.
“Agentic orchestration” via public APIs that allow autonomous agents to operate across LiveRamp’s network and platform.
Asked whether AI agents could disintermediate ad tech companies, Westermann described agents as a way to automate manual workflows and improve optimization, but said they still require an operating environment and data. He cited media planning as an example where agents could reduce time spent from days to hours, while cautioning that he would not yet trust an agent to negotiate ad pricing.
Stone said investors often focus on whether agents “replace” parts of the workflow, while he views agents primarily as tools that “enhance” budgeting, targeting, and measurement. Dillard similarly said she does not expect the ecosystem to “collapse,” but rather to become orchestrated through “multi-agent communication,” describing specialized agents for planning, segmentation, targeting, and measurement in a more iterative loop.
Data ownership and access emerged as a central theme. Stone argued that “data is king,” and said companies with the best data and distribution scale are positioned to benefit most as AI lowers barriers to entry. He highlighted Digital Turbine’s Ignite technology, which he said has on-device permissions and access to usage attributes such as apps used, time spent, deletions, location, and other device/user signals. Stone said OEMs and operators want to leverage that data more aggressively, describing it as “fuel” for AI and agents.
Dillard said the evolution of clean rooms—secure environments for sharing data and running analytics without exposing personal information—has been foundational for retail and commerce media networks, and could become increasingly important for training models and enabling agents to operate on collaborative datasets. She said this is already happening today.
Westermann, speaking from a European regulatory perspective, said privacy is becoming a larger topic in the U.S. as well, driven in part by platform changes and consumer behavior. He noted that publishers now face mixed inventory—some with IDs/cookies and some without—often leading to lower pricing for non-addressable inventory unless improved targeting solutions are available. He said Verve has invested in approaches to target without IDs using AI, on-device intelligence, and multiple data sources. As an example of AI-enabled targeting, he described using AI to find “contextual pockets” where ads perform well.
The panel discussed a shift toward more model-based approaches in measurement. Stone said “last click” attribution remains common in mobile and is unlikely to change as the basis for payment, but AI can improve upstream modeling—such as testing different attribution windows and understanding the broader marketing mix that contributed to conversion.
Dillard said LiveRamp customers value deterministic matching for closing the loop on measurement and optimizing based on outcomes such as known purchases, including offline conversions. Westermann said deterministic measurement is preferred when available (for example, with retail point-of-sale integration), with probabilistic methods used when deterministic measurement is not possible.
On AI’s impact on search and the open web, Dillard said LiveRamp is less exposed to the open web than some investors assume, and she expects AI to create new ad-supported “surfaces” over time. She cited experiments with advertising by ChatGPT, Perplexity, and Gemini, and said LiveRamp would treat these as new channels to connect data into and measure against outcomes.
Stone said declining open-web traffic is a net positive for app-centric models, arguing that activity is shifting into native app experiences. Westermann said search advertising dollars that are disrupted will need to move elsewhere, including in-app and connected TV (CTV), and he expects LLM platforms to eventually introduce advertising given their need to monetize.
On CTV, Dillard said it has been LiveRamp’s fastest-growing channel for the past 2–3 years and now represents the majority of LiveRamp’s largest integrations by data volume. She also said CTV platforms are often more willing than traditional social walled gardens to share data back with advertisers for measurement and optimization, frequently in protected environments such as clean rooms. Westermann called CTV highly competitive with margin pressure and limited differentiation. Stone said CTV has become commoditized and cluttered, and positioned Digital Turbine’s in-app channel as a differentiated way to reach audiences at scale.
On retail media, Dillard said the concept is expanding into “commerce media networks,” citing examples such as airlines and on-demand delivery services, with banks also exploring the area. Stone said retail media is among Digital Turbine’s fastest-growing areas, describing a use case where a retailer’s own network may not fully reach an advertiser’s desired audience, creating an opportunity for Digital Turbine to extend reach through its device footprint. Westermann said the proliferation of retail media networks adds complexity for advertisers and increases the need for aggregation and curation through platforms.
Discussing cyclical ad drivers such as elections and major sports events, Stone said advertising spend follows “eyeballs,” creating a tailwind for media and ad tech. Westermann added that increased demand can raise CPMs across the market, benefiting companies even if they are not heavily concentrated in political advertising. Dillard noted LiveRamp’s subscription model based on data volume makes revenue less sensitive to cyclical ad spending swings, though she said its data marketplace—where LiveRamp takes a take rate—could see uplift during tentpole events.
In closing, the executives pointed to the pace of change as a major risk and challenge. Stone emphasized the importance of adaptability and timing adoption correctly. Westermann said companies must be agile or risk being left behind, and he also noted investor uncertainty has contributed to valuation pressure across the sector. Dillard said the industry’s growing complexity increases the need to help customers navigate change, adding that AI is likely to increase fragmentation more than consolidation.
Digital Turbine, Inc (NASDAQ: APPS) is a mobile technology company that streamlines content delivery and app advertising across connected devices. Its platform enables carriers, OEMs, app developers and advertisers to engage users through personalized app recommendations, in-app promotions and turnkey monetization solutions. By integrating software directly on smartphones and tablets, Digital Turbine simplifies the user journey from discovery to installation without requiring additional downloads or redirects through traditional app stores.
The company’s flagship Ignite Platform offers end-to-end campaign management, combining demand-side advertising, real-time analytics and automated content fulfillment.
