Viant's TVision acquisition reveals how CTV's identity infrastructure is being assembled deal by deal — and what that means for marketers in Southeast Asia.
Viant’s $40 million acquisition of TVision is being reported as a CTV measurement play. It’s actually something more structural than that — it’s the second piece of an infrastructure thesis that most buyers haven’t fully mapped yet.
The Acquisition Logic Nobody’s Writing About
When Viant acquired IRIS.TV, it got contextual intelligence — a way to understand what content a viewer is actually watching, not just which app or device they’re on. TVision adds attention measurement: time-in-view, eyes-on-screen data, co-viewing signals. Separately, these are interesting capabilities. Together, they’re the beginning of a cookieless identity layer built specifically for the living room.
As AdExchanger reported, Viant has been explicit that one deal enabled the other — IRIS.TV gave them the content signal infrastructure that makes TVision’s attention data actionable at scale. That sequencing matters. It tells you Viant isn’t shopping for features; they’re assembling a stack. The question for every programmatic buyer right now is: which DSP’s acquisition roadmap most closely matches your targeting requirements in a post-cookie environment?
Why Publicis Isn’t Building a DSP — and Why That’s the Right Call
Digiday reported that Publicis CEO Arthur Sadoun ruled out building a Trade Desk rival, describing the ambition as too far off the company’s current roadmap. That’s a disciplined answer, and strategically it’s probably correct.
Building DSP-grade infrastructure from scratch in 2026 means competing against companies that have spent a decade hardening their identity graphs and supply-path relationships. The window for that kind of vertical integration closed around 2022. What Publicis actually has — Epsilon’s first-party data assets, Citrus Ads for retail media, and CoreAI — is a more defensible position: a holding company that doesn’t need to own the pipes because it controls the data that flows through them.
The harder version of this question isn’t whether Publicis should build a DSP. It’s whether any agency holding company’s first-party data moat remains meaningful as clean rooms commoditise and DSPs build their own identity resolution layers. That tension isn’t resolved yet.
What This Means for CTV Buyers in Southeast Asia
Southeast Asia’s CTV landscape is structurally different from North America in ways that make the Viant playbook both instructive and partially irrelevant. Streaming fragmentation here runs through regional players — Vidio in Indonesia, iQIYI, WeTV, and Viu across the region — alongside global platforms, none of which have the open contextual data partnerships that IRIS.TV’s model depends on.
Attention measurement in the region is also underdeveloped. The ACR and viewability infrastructure that TVision monetises in the US simply doesn’t exist at scale in most SEA markets. For buyers operating across Thailand, Vietnam, or the Philippines, the practical equivalent of the Viant stack is being assembled informally: a mix of platform-native audience tools, clean room pilots with telcos and super-apps, and increasingly, retail media networks from Lazada and Shopee that carry richer purchase-intent signals than anything CTV can currently offer.
The strategic takeaway isn’t to replicate Viant’s M&A logic in SEA — it’s to identify which signal types in your market are currently siloed and could be combined to produce the equivalent of contextual-plus-attention targeting.
The RAG Moment in MarTech Infrastructure
There’s a quieter infrastructure shift happening in parallel. Martech Zone’s Douglas Karr recently published a detailed breakdown of building a Retrieval-Augmented Generation assistant on Cloudflare Workers, using Vectorize and Llama 3.3 — a fully self-hosted, site-specific AI layer built without a single enterprise SaaS contract.
The cost and accessibility of that build would have been inconceivable 24 months ago. What it signals for adtech and martech teams is that the gap between enterprise AI tooling and what a competent internal team can deploy is collapsing. Clean room logic, audience segmentation queries, creative personalisation pipelines — the components that DSPs and platforms have been selling as differentiated products are increasingly replicable with open-source models and commodity cloud infrastructure.
For growth and media teams in Southeast Asia, where enterprise martech licensing costs have always been a barrier, this matters more than it does in mature markets. The playbook is being rewritten at the infrastructure layer, not the campaign layer.
Key Takeaways
- Viant’s sequential acquisitions — IRIS.TV then TVision — reveal how CTV’s cookieless identity stack is being built through compounding capability, not single-feature deals; map your DSP partners’ roadmaps with the same lens.
- Publicis’s decision to skip building a Trade Desk rival reflects a broader truth: in 2026, first-party data ownership is more defensible than pipe ownership, but only if clean room commoditisation doesn’t erode that moat.
- In Southeast Asia, the attention and contextual signals that North American CTV buyers take for granted don’t yet exist at scale — retail media networks on Lazada and Shopee currently offer the region’s strongest intent-signal equivalent.
The deeper question Viant’s acquisition trail raises isn’t about CTV specifically — it’s about which companies are building durable infrastructure versus which are assembling features. As cookieless targeting matures and open-source AI makes more of the stack replicable, the moats that actually hold will be data relationships and proprietary signal access, not software. The question worth sitting with: does your current media stack give you access to signals that your competitors genuinely cannot replicate?
At grzzly, we work with brands and media teams across Southeast Asia who are navigating exactly this kind of infrastructure transition — figuring out where to invest in owned data capability versus where to rely on platform partnerships, and how to build targeting architectures that don’t collapse the moment a platform changes its policies. If you’re rethinking your adtech stack or trying to make sense of what identity resolution actually looks like in your markets, Let’s talk
Sources
Written by
Rogue GrizzlyOperating at the contested frontier of cookieless targeting, clean rooms, and identity resolution. Comfortable where the infrastructure is shifting and the playbooks have not yet been written.