The channel walls came down.

Digital media in 2026 is not the stack it was in 2022. The boundaries that organized planning then, with CTV in one column, programmatic display in another, and retail media in its own swim lane, have dissolved into a single addressable ecosystem where identity, signals, creative, and outcomes travel across channels by default (MediaPath Research, The State of Digital Media 2026, May 2026). US digital ad revenue reached $294.6 billion in 2025, up 13.9% year over year (IAB/PwC, 2026), and digital now crosses two-thirds of US adult media time. Audiences made the move first. Brands now run six to nine channels by default, and the viewer carries the same identity across all of them.

The convergence runs along four vectors at once: retail media selling CTV inventory, social video living on TV screens, search splitting into AI surfaces, and digital out-of-home consolidating into telecom (MediaPath Research, 2026). Each vector has dated, public proof behind it. The planning unit shifted from channel selection to cross-channel orchestration years ago. Most org charts have not caught up, and the rest of this piece is about what that lag costs.

Off-site retail media, meaning retailer data activating impressions on CTV, social, and the open web rather than on the retailer's own properties, is growing 42.1% year over year against 15.3% for on-site inventory, nearly three times faster (eMarketer, 2026). Off-site placements now account for roughly 20% of total retail media network spend. The fastest-growing fifth of the category's budget runs somewhere other than the retailer's storefront. Retail media in 2026 works as a data layer that prices impressions across the wider ecosystem.

Concentration inside the category is extreme. Amazon Ads holds 79.7% of US retail media spend and Walmart Connect 8.0%, and the two will capture 89% of 2026 incremental retail media dollars (eMarketer, 2026). Walmart's acquisition of Vizio is the canonical case of what the second-place network did about it. The $2.3 billion deal closed on December 3, 2024, and Walmart Connect now sells CTV inventory on Vizio's SmartCast operating system with Walmart transactional data serving as both targeting and measurement. The pairing opened to advertisers as a beta in April 2025 (Walmart Connect, April 2025). The result reads directly in the earnings: Vizio contributed 28 percentage points to Walmart's ad revenue growth in the second quarter of fiscal 2026 (Walmart Q2 FY26 8-K, 2025). A retailer is now a television distributor, selling retail data and CTV inventory as one stack.

The clearest evidence that the walls are down is where television now happens. YouTube captured 13.4% of total US TV viewing time in July 2025, the largest lead any single media distributor has held since Nielsen's Gauge launched (Nielsen Gauge, July 2025). The largest distributor on American television that month was a social video platform, and the biggest screen in the house is where social video increasingly lands. The channel label on the insertion order no longer tells you where the audience is sitting.

Television is also bought the way display is bought. Programmatic accounts for 84% of US CTV video ad spend, roughly $27 billion of $32 billion (eMarketer, 2025), and 2026 is the first year CTV upfront commitments exceed primetime linear TV upfronts, at $17.73 billion against $16.98 billion (eMarketer, 2026). Daily media time tracks the same dissolution: US adults spend 3 hours 12 minutes with digital video and 2 hours 18 minutes with social, against 1 hour 56 minutes with linear TV (eMarketer, 2026; Nielsen Total Audience, Q4 2025). A plan that assigns television to one team and video to another is describing its own org chart.

Fourteen months of deal flow made the convergence structural. T-Mobile acquired Vistar Media for $600 million in February 2025, taking on 1.1 million screens, 370 media owners, and 3,000 advertisers, and folding digital out-of-home into a telecom (T-Mobile, February 2025). Broadsign acquired Place Exchange in November 2025 (Broadsign, November 2025). Inside nine months, two of the largest independent DOOH supply-side businesses had moved under bigger owners.

The platform moves point the same direction. Amazon DSP added Spotify's biddable audio inventory in October 2025, placing Spotify on an off-site roster that already included Netflix, Disney, Warner Bros, Roku, and SiriusXM (Spotify 6-K, October 2025); walled-garden DSPs now activate audio and CTV inventory well beyond their own properties. The Trade Desk answered from the open-web side in February 2026 with Ventura, assembled from OpenAds, OpenPath, UID2, OpenPass, V (formerly VIDAA, with more than 50 million connected devices), and Nexxen, and designed as the connective layer between CTV inventory and identity (The Trade Desk, February 2026).

Every one of these transactions joins a data or identity asset to an inventory pool that sat in a different planning column two years ago. The buyers are not betting on a channel. They are betting on the connections between channels.

Org charts are the last thing to move. Most brands still run channel-led teams, channel-led budgets, and channel-led metrics against an ecosystem that stopped honoring those borders. The cost is measurable: audiences bought twice across overlapping channels, attribution that weakens at every channel boundary, and walled-garden pricing that faces no cross-channel pressure. Call that cost the fragmentation tax (Dentsu, 2025). Channel-specialist structures cannot price the tax their own structure creates, because each team sees only its own column. Three structural moves close the gap.

The first move is measurement-first planning. MMM, incrementality, and attention set the outcome architecture before any channel is assigned dollars, bypassing last-touch attribution entirely. The IAB/MRC Attention Measurement Guidelines v1.0, published in November 2025, made attention metrics defensible at the brief stage (IAB/MRC, November 2025). The second is orchestration-led planning. With six to nine channels running by default, cross-channel audience deduplication now does the job frequency capping used to do. Treat overlap as a planning variable and deduplicate before the buy.

The third is identity-anchored planning. UID2, ID5, LiveRamp ATS, and first-party data clean rooms are foundational infrastructure for a cross-channel plan, and the industry kept building all four even after Google retired most of Privacy Sandbox in October 2025 (MediaPath Research, 2026). The report's closing thesis holds for this excerpt: plan to the ecosystem, anchor on the open web, and measure the whole. Every quarter, the gap widens between the brands organized for one addressable ecosystem and the brands still planning like it is 2022.