The Ad Tech Board

Strategic Sale / Absorption — Bridgepoint exits MiQ to a strategic, ending independence35–43%

On Leased Land

MiQ, the independent programmatic trading desk, and its CTV/ATV bet

Persisting Independent — Recap, Secondary Buyout, or Continuation; Aggregator Survives, Margins Bleed24–32%

Erosion / Distress — Agentic Commoditization + Supplier Seam Break → Margin Collapse or Distressed Sale15–22%

Software Re-Rating — Sigma Becomes a Real Margin Story (bull)10–16%

Absorbed Front Door, Gap Narrows on Rate Not Level (modal)38–46%

The Missing Front Door

Walmart's $1.4B acquisition of Vibe.co — completing a retail-media + connected-TV ad stack, and the long catch-up to Amazon

Catch-Up Stalls: Moat Unaccredited, Channel Conflict and SMB Churn Bite24–32%

Standalone Onramp Persists12–18%

Breakout: The Closed Loop Actually Converts10–16%

Deal Slips, Stalls, or Breaks (regulatory / diligence / walk)8–14%

Openthrough 2028 (directional to 2030)

Programmatic Infrastructure 4

Who Keeps the Spread

Programmatic CTV's DSPs and SSPs — and where the margin really sits for a platform-agnostic intermediary

Curation eats the stack (sell-side ascendancy)35–45%

The pincer (margin compression from both sides)30–40%

Walled-garden gravity (bifurcation, narrowed)15–25%

Agentic routing commoditized10–20%

Layered Coexistence on an MCP Substrate (negotiation + transport stack, governance split)36–44%

The Protocol War

How AI agents are being wired into the ad buy — MCP, AdCP, the IAB's ARTF, and what agentic buying does to CTV

Garden + Model-Layer Enclosure (the open stack governs a shrinking remainder)26–34%

Agentic Plateau — pilots don't scale; the standards war stays pre-volume14–20%

Material Disintermediation — agents compress the open-market CTV stack by 202910–16%

Regulatory / Substrate Shock (wildcard — antitrust rupture OR MCP fragmentation)8–14%

Absorbed / Bounded Plumbing (MODAL) — containerized bidding becomes real, multi-fabric open-web infrastructure but a commoditized SSP feature, not a new margin-capturing seat; at-scale workload is…40–46%

Openthrough 2030

Closer to the Metal

Containerized bidding — when the buyer's bidder moves inside the exchange's data center, and who captures the margin it relocates

Margin Relocation with a Real Seat — in-fabric execution forms a DEFENDED independent-SSP hosting/curation rent and/or an incumbent-scale DSP becomes a tenant; the standalone DSP-infrastructure tier…20–28%

AdCP Topology Bypass — direct buyer-agent/seller-agent negotiation routes around the impression auction, shrinking the problem containerized bidding solves; relocated margin flows to publishers and…11–17%

Stall / Abandonment — the category stays pilot-bound and never escapes beta, or collapses into one dominant proprietary fabric; not even absorbed as a standard feature8–14%

Incumbent-Fabric Capture — containerization scales but a hyperscaler/walled garden (Amazon via AWS RTB Fabric + DTE, or Google), not an independent SSP, keeps the relocated spread; open-web…8–13%

Squeezed but Standing38–46%

The Seller's Agent

The supply-side platforms that carry programmatic CTV — and their bet on attention, outcomes and AI

Absorbed into the Platforms18–26%

Up the Stack15–23%

Structural Break12–20%

Managed Degradation (Dual-Stack Equilibrium — IP Demoted, Not Dethroned)48–58%

Identity & Addressing 2

Openthrough 2030

The Address Knows You

IPv4/IPv6 and the global CTV programmatic landscape

Accelerated Displacement (Authenticated Takeover of Premium CTV)20–30%

Obfuscation / Regulatory Shock8–15%

IP Renaissance (Verified Linkages & IPv6 Prefix Standardization)7–15%

Residual Tail (Unmodeled Discontinuities)3–5%

Messy Continuity — durable plural multi-signal, multi-graph stack38–48%

The Signal Layer

Identity in programmatic CTV — hashed emails, MAIDs, IP, ACR, clean rooms, and what survives to 2029

Walled-Garden Tilt — gardens gain decisively (gradual grind, not a fast tip)28–38%

Regulatory / Litigation Reset — a privacy step-change reshapes the stack10–16%

Open-ID / Standards Consolidation — a universal ID and currency converge toward parity8–15%

Residual tail — exogenous shock, signal cliff, or non-modeled break4–9%

Fragmented Floor (base case)35–45%

Trust, Politics & Society 3

Openthrough 2030

Safe Beside What?

Brand safety in programmatic Connected TV — and the global fight over who sets the floor

Content-Level Resolution18–25%

Walled-Garden Consolidation12–18%

Fraud-and-Fragmentation Drift12–18%

Re-coordination Under Legal Cover8–14%

Openthrough the 2028 presidential cycle (early 2029)

The Measured Screen

Political advertising migrates to connected TV — the smart-TV ACR and voter-file targeting stack, who captures the spend, the streaming regulatory gap, and whether any of it actually moves votes.

On-Curve Continuation — CTV climbs the flattening shoulder of its S-curve; the federal streaming gap stays open34–44%

Constrained Plateau — political dollars keep rising but CTV's SHARE stalls in the low-to-mid 20s%18–26%

Inventory Unlock / Re-Acceleration — a continuity-breaking unlock pushes CTV share materially higher (low-30s%)13–19%

Regulatory / Platform Break — a regulatory, judicial, or structural shock bends the trajectory (incl. structural wildcard)13–19%

Effectiveness / Credibility Deflation — the persuasion case breaks and share growth stalls below continuation7–12%

Bifurcation / Split Verdict — LinkedIn durable, the other four a workaround being disintermediated36–46%

The Living-Room Handoff

How social platforms — LinkedIn, Pinterest, Reddit, TikTok, Meta — extend their audiences onto Connected TV, and which ones credibly make the jump through 2029

Near-Total Disintermediation / Full Inversion — even LinkedIn compresses to a thin licensing fee22–32%

Durable Multi-Platform Category — extension sticks for 2–3 players (LinkedIn + owned-rail Pinterest, ± a Meta pipe)16–24%

Meta Front-Runs — a shipped Meta CTV product (owned Reels-TV land-grab and/or demand network) reshapes the category12–18%

Regulatory Rupture — VPPA/state-privacy/DMA breaks the household-IP plumbing itself9–15%