Ogury's new agentic tool promises to eliminate a longstanding disconnect between how campaigns are planned and how they actually run - using persona-based data to drive continuity from brief through to activation.

Ogury on June 9, 2026, announced the launch of SONA, an agentic solution designed to address what the company describes as a persistent structural gap between audience planning and campaign execution. The New York-based announcement marks a notable expansion of Ogury's Persona Intelligence technology into autonomous workflow territory - one of the most actively contested areas of ad tech in 2026.

The timing is deliberate. The advertising industry has spent the past 18 months building infrastructure for agentic campaign management, with platforms ranging from supply-side operators to DSPs racing to embed autonomous capabilities into their pipelines. Ogury's entry differs from most of those moves in its starting point: rather than optimizing what happens during a campaign, SONA focuses on what happens before one starts.

What SONA does, step by step

According to Ogury, SONA begins by ingesting any campaign brief and structuring it into scalable, combinable personas. Each persona comes with full visibility into the behaviors, preferences, and geographic insights behind the recommendation - not just a label, but the underlying signals that generated it.

From there, SONA identifies the most relevant publisher environments and formats to reach those personas. It then generates media strategy recommendations that include projected reach and expected outcomes. Once the strategy is confirmed, campaigns can be activated directly through the same system, maintaining continuity between planning intent and execution parameters.

In practice, according to Ogury, this sequence can be completed in a few clicks. The company estimates that SONA saves 20 to 30 hours per campaign compared to manual planning and activation workflows. That is a significant claim in an industry where media planning remains heavily manual, spread across disconnected tools and spreadsheets.

The data model behind the personas

SONA's personas are not derived from a single signal. According to Ogury, the system draws on the company's audience-first data model, which brings together three categories of input: observed behaviors, declared data from surveys, and purchase signals based on real transactions.

The combination is designed to produce personas that reflect how people actually think and behave - not just demographic proxies or inferred interest categories assembled from browsing patterns. By structuring these signals into scalable personas, Ogury argues that SONA surfaces audience profiles that traditional approaches overlook, because those approaches tend to optimize for the audiences most easily found in existing platform segments rather than the full picture of relevant demand.

At every stage of the recommendation process, according to Ogury, the output remains fully transparent. Traders can inspect, adjust, and validate recommendations before any activation occurs. That emphasis on human oversight sits alongside the agentic framing rather than in tension with it - the system automates the assembly work while leaving strategic decisions legible and reversible.

Channels and integration pathways

SONA is available across CTV, desktop, and mobile environments, covering both web and in-app inventory. That cross-channel scope builds on Ogury's February 24, 2026 extension of Persona Intelligence to connected television, which marked the company's first formal deployment of its audience framework beyond mobile and desktop. SONA now provides a unified workflow capable of activating the same persona across all three environments within a single campaign.

Access to SONA runs through two pathways. The first is Ogury One, the company's own platform interface, where SONA is accessible via a chat-based interface. The second pathway uses Model Context Protocol (MCP), the connectivity standard developed by Anthropic that has become a common integration layer across ad tech platforms throughout 2025 and into 2026. Through MCP compatibility, SONA can be embedded into existing workflows via Gemini, Claude, and ChatGPT - meaning teams do not have to leave their preferred AI environment to use it.

That interoperability point matters in the current landscape. MCP has quickly become the connectivity fabric of agentic advertising, with Google, Amazon, PubMatic, AdRoll, and a growing list of platforms building MCP-based integrations throughout late 2025 and early 2026. Ogury's adoption of the protocol positions SONA as a participant in that ecosystem rather than a walled-off proprietary system.

Nicolas Bidon on the planning problem

"Many solutions today focus primarily on optimizing execution, but the real challenge is maintaining a consistent audience definition from planning through activation across platforms," said Nicolas Bidon, CEO of Ogury. "With SONA, we're bringing Persona Intelligence directly into the campaign workflow, enabling brands and agencies to turn any brief into a structured, persona-based strategy they can understand, refine, and activate consistently across channels to drive more measurable outcomes."

Bidon's framing identifies a specific failure mode in current campaign operations. Audience definitions often shift between the planning stage, where strategies are presented to clients and stakeholders, and the execution stage, where traders work within platform constraints that may not map cleanly onto the original intent. The result is a gap between what was sold in a media plan and what actually gets bought and delivered. SONA's architecture - where personas built during planning are the same ones used to drive activation - is designed to close that gap structurally rather than through manual reconciliation after the fact.

Why the planning-execution gap persists

The gap Ogury is targeting is not new. It reflects a fragmentation that runs through the architecture of the ad tech stack itself. Planning tools have historically operated in isolation from buying platforms. Data management platforms generate audience segments that then need to be exported, reformatted, and re-ingested by DSPs. Publishers accept targeting parameters that may approximate but rarely replicate the original audience intent. Each translation introduces drift.

Audience targeting has been converging toward real-time, AI-generated signals and away from static segments, a shift PPC Land has tracked across multiple platform launches in 2025 and 2026. The pressure comes from two directions simultaneously: the deprecation of third-party identifiers that once enabled cross-platform audience portability, and the rise of AI tools capable of generating and refreshing audience models far faster than human planners can.

SONA's approach - anchoring personas in a proprietary, multi-signal data model and carrying them through the entire workflow - represents one architectural response to that pressure. The alternative responses cluster around clean rooms and identity resolution, where platforms attempt to match first-party data at the point of activation rather than maintaining a consistent audience construct throughout.

Where Ogury sits in the agentic ad tech landscape

The broader agentic advertising ecosystem has moved quickly. PubMatic launched AgenticOS in January 2026, framing it as infrastructure for autonomous campaign execution across premium digital environments. Yahoo DSP embedded AI agents across five campaign lifecycle categories earlier that year. AdRoll and PubMatic demonstrated a working MCP-based deal diagnostics integration in April 2026. IAB Tech Lab has spent much of 2025 and 2026 building out its Agentic Advertising Management Protocols (AAMP) framework, which drew nearly 400 industry participants to its Tech Lab Summit in May 2026.

Ogury's participation in that wider ecosystem was already visible at the IAB Tech Lab Summit 2026, where the company appeared among a broad cross-section of sponsors that included Kerv, Meta, Optable, Permutive, The Trade Desk, LiveRamp, and others. The SONA launch translates that positioning into a specific product.

What distinguishes SONA from most of the agentic tools that have launched in this period is its entry point. Most agentic advertising products to date have focused on campaign management - bid adjustments, creative rotation, pacing corrections, deal diagnostics. SONA starts upstream, at the brief, and argues that the execution problems those other tools are solving are downstream symptoms of a planning problem that has not been addressed.

Company background

Ogury was founded in 2014 and currently operates in 18 countries with more than 500 employees. The company has built its commercial offering around Persona Intelligence - a proprietary technology that positions personas, rather than individuals or demographic cohorts, as the primary unit of audience strategy. That framework has been extended progressively across channels: mobile and desktop first, then CTV in February 2026, and now, with SONA, into the workflow layer that governs how audiences are defined and activated in the first place.

The Ogury One platform provides the primary interface for SONA access. Clients working in third-party AI environments can use the MCP integration pathway without migrating to a new system. According to Ogury, the setup does not require complex configuration or changes to existing campaign processes.

What this means for media buyers

For traders and media planners, SONA represents a workflow change more than a data change. The underlying Persona Intelligence data model is not new. What is new is the degree to which that model is now embedded in the end-to-end process - structuring the brief, selecting publisher environments, projecting outcomes, and executing the buy - rather than being consulted at one stage and then set aside.

The 20-to-30-hour per campaign time saving estimate, if it holds across diverse campaign types and scales, would be material for agencies running high volumes of programmatic activity. Planning and activation workflows for a single mid-size campaign can routinely consume a week of analyst and trader time when account management, data pulls, media plan construction, and trafficking are combined. Compressing that into a guided agentic flow has obvious operational implications.

The IAB Europe Virtual Programmatic Day in April 2026 surfaced a recurring theme that SONA is directly designed to address. Panelists at that event noted that audience strategies are not consistent across channels, and that campaign execution often diverges from planning intent. An Ogury representative at that event described persona intelligence as building aggregated audience models based on behaviours, motivations, and intent - arguing that those personas are more durable and more resilient across channels than individual identifier-based approaches.

Whether SONA delivers on that framing in practice will depend on how the persona model holds up across the publisher environments and inventory types it activates against - particularly in CTV, where the signal environment and inventory architecture differ substantially from mobile and desktop. The full transparency built into each recommendation stage, which allows traders to inspect and override before activation, is the design choice that provides most of the assurance for practitioners who need to maintain client-facing accountability for every campaign decision.

Timeline

Summary

Who: Ogury, a global adtech company founded in 2014 and operating in 18 countries with over 500 employees, with the announcement attributed to CEO Nicolas Bidon.

What: The launch of SONA, an agentic solution that takes a campaign brief through persona structuring, publisher environment identification, media strategy generation, and direct campaign activation - all within a single workflow. SONA integrates with Ogury One via a chat-based interface and with external AI environments including Gemini, Claude, and ChatGPT via MCP. The system is available across CTV, desktop, and mobile inventory.

When: Announced June 9, 2026, from New York.

Where: SONA is accessible globally through Ogury One or via MCP-compatible AI tools, across markets where Ogury operates. Campaign activation covers CTV, desktop, mobile web, and in-app inventory.

Why: The persistent disconnect between audience planning and campaign execution - where audience definitions shift between planning tools and buying platforms, and execution diverges from original intent - has remained unresolved despite years of DSP and data platform development. Ogury argues that by anchoring the entire workflow to its Persona Intelligence data model, from brief through activation, SONA eliminates the translation losses that introduce audience drift. The agentic framing compresses workflow time by an estimated 20 to 30 hours per campaign while maintaining full trader visibility at each recommendation stage.