For the first time, CTV upfronts beat primetime linear TV

eMarketer projects that 2026 U.S. CTV upfront commitments will top primetime linear TV upfront spending for the first time. Total U.S. CTV ad spend is projected at roughly $38 billion this year, up about 14% year over year. The money has moved. The plumbing is now catching up.

The milestone matters because of what it ends. For the entire history of the upfront market, primetime linear television was the anchor: the biggest commitments, the highest-status inventory, the line item that set the market. The projection says that era is over in 2026. CTV upfront commitments, money pledged ahead of the season for streaming inventory, will exceed primetime linear upfront spending. Not total CTV versus total linear. Upfront versus upfront, the negotiated core of each market. When the planning money tips, the balance of the industry tips with it.

And the scale behind the milestone is what makes it structural rather than symbolic. Roughly $38 billion in U.S. CTV ad spend this year, up about 14% year over year. A 14% growth rate on a $38 billion base is about $4.7 billion of new money flowing into streaming in a single year, and most of that money is arriving with programmatic expectations attached: efficiency, measurement, accountability. The buyers spending it are not asking for GRPs and a handshake. They want to know where every dollar lands.

What Viant actually announced

On September 24, Viant expanded Direct Access, its supply path optimization product, to deal-based buying: private marketplace and programmatic guaranteed, on top of open market. The company claims coverage across the largest CTV publishers, representing 94% of programmatic demand per Jounce Media, with no Direct Access fee for buyers or sellers to set up or execute a deal.

Unpack what that combination means in practice. Supply path optimization is the discipline of getting your budget to the publisher with as few intermediaries as possible. Historically in CTV, that discipline was mostly available on open market buys, where buyers could pick clean paths to inventory. The deal-based world, private marketplace and programmatic guaranteed, was where buyers went for premium access and price certainty, and it came with messier, longer supply paths. Viant is now saying its optimization layer covers all three: open market, PMP, and PG. The pitch is that you no longer have to choose between clean supply paths and premium deal access.

The two claims doing the heavy lifting in the announcement are the coverage number and the fee number. Coverage across the largest CTV publishers, representing 94% of programmatic demand per Jounce Media, is a reach claim: near-universal access to where the money actually flows. And no Direct Access fee for buyers or sellers to set up or execute a deal is a pricing claim designed to remove the standard objection to SPO products, which is that the optimization costs as much as the waste it removes.

The named routes complete the picture: FreeWheel, Magnite's SpringServe, Publica, and custom oRTB integrations. Those are the actual pipes. FreeWheel and SpringServe are the ad servers sitting inside major publishers and broadcasters. Publica is the CTV ad platform inside streaming apps. Custom oRTB integrations cover everything else that speaks open RTB. A named route is a concrete path you can ask your team about. It is not a vibe.

Two caveats worth keeping

First, the 94% figure is Viant's framing of Jounce data, not an independent finding. Vendor math deserves vendor-grade skepticism. Second, the named routes are FreeWheel, Magnite's SpringServe, Publica, and custom oRTB integrations, which is not a promise that every deal clears this path. Coverage claims describe the map, not your specific route.

On the first caveat, the distinction is worth more than it seems. Jounce Media is a real research source, and Viant citing it is better than citing nothing. But "94% of programmatic demand" is Viant's interpretation of Jounce's numbers applied to its own coverage claim. The question to ask is not whether the number is real but what it measures: 94% of what, precisely? Programmatic demand across all of CTV, or across the publishers Viant counts as "the largest"? Does your inventory live in that 94%, or in the 6% that sits outside it? A coverage number is only useful once you know whether your buys are inside the circle it draws.

On the second caveat, the route list is where the announcement gets concrete and where it gets limited at the same time. If your CTV deals run through FreeWheel or SpringServe or Publica, you are plausibly inside the Direct Access footprint. If your buys flow through other paths, or through resellers who bundle inventory across routes, the announcement says nothing about your situation. The honest read: this is a real capability on four named pipes, not a market-wide fix. Check your pipes.

Why this matters beyond one vendor

The bigger signal is structural: TV dollars are not returning to linear. They are renegotiating how they get to streaming. Supply path optimization in CTV means fewer hops between your budget and the publisher, which means less reseller tax and cleaner measurement. Every intermediary in the chain takes a cut and muddies frequency and attribution a little more. In a $38 billion market, those little cuts add up to real money.

Work the arithmetic on the market size. At roughly $38 billion in annual CTV spend, every percentage point of supply-chain waste is worth about $380 million industry-wide. If the average buy loses even a few points to reseller fees, duplicate auctions, and data leakage across hops, the aggregate waste runs into the billions. That is why SPO is not a finance-department hobby project. On a base this large, the plumbing is the margin.

The measurement half of the argument matters just as much. Every hop between your budget and the publisher does two things: it takes a cut, and it degrades the signal. Frequency capping is the clearest victim. A cap of three exposures per household only works if the system counting the exposures sees all of them. When impressions for the same household flow through multiple resellers and paths, each one counts what it sees, and the cap leaks. The result is the CTV complaint every media buyer knows: the same household sees your ad eight times while your dashboard says three. Cleaner paths mean truer frequency and cleaner attribution, which means the performance data you optimize against stops lying to you.

The practical move: audit your video supply path

The practical move is unglamorous: audit your video supply path before upfront season locks in. Know how many hands touch your impressions, what each one costs, and whether your frequency caps survive the trip.

1. Map the path. For your top CTV line items, trace the route from your DSP to the publisher: which SSPs, which resellers, which ad servers. If the answer is "we don't know," that is the first finding.
2. Count the hops and their costs. Every intermediary between your budget and the publisher takes a fee or an auction margin. List them, price them, and add them up. Compare the all-in working-media percentage on your CTV deals against your open-market buys.
3. Test your frequency caps. Pull household-level exposure data where you can get it and compare real exposure counts against your capped targets. If the same households are blowing past caps, your supply path is leaking.
4. Ask about deal-based SPO on your routes. The Viant announcement covers FreeWheel, Magnite's SpringServe, Publica, and custom oRTB. Ask your partners which of your deals touch those pipes and whether SPO is applied. Get the answer in writing, not in a slide.
5. Lock it in before the upfront. Upfront season is when commitments harden and paths get set for the year. A supply path audit done in Q4 is leverage. The same audit done in Q2 is archaeology.

The counterpoint

Not every intermediary is waste. Some resellers provide real value: aggregated reach across long-tail publishers, deal curation, data enrichment, guaranteed delivery terms. The goal of a supply path audit is not zero hops. It is known hops, priced hops, and hops that earn their keep. A short path to the wrong inventory is not better than a longer path to the right one. Audit for value, not for hop count alone.

How many hops does your video budget take before it reaches a publisher?

Data over opinions.