Published May 28
Streaming TV advertising still carries a structural problem inherited from the web: the industry often treats television ad breaks as disconnected impressions instead of cohesive viewing experiences.
That mismatch is what’s driving the push toward dynamic ad podding — a programmatic approach that allows buyers and sellers to transact on the full ad break rather than fragmented individual requests.
The Trade Desk is now accelerating that transition, arguing that podding can reduce duplication, improve efficiency across the supply chain, and create better outcomes for both advertisers and viewers.
The Current spoke with Rob Hazan, GM of product management at The Trade Desk, about why the ecosystem tolerated inefficiency for so long, why transaction IDs matter and what streaming advertising could look like once podding becomes standard.
This conversation has been edited for clarity and length.
The industry spent years treating TV like the web. Why?
What’s happening is that streamers, broadcasters and app owners don’t naturally think in terms of discrete impression opportunities the way the web did.
On the web, you had squares on a page. Each square was an impression opportunity. You’d solicit bids for that slot and fill it.
TV is fundamentally different because it has ad breaks. You have a fixed amount of time dedicated to advertising within the content itself, but there’s an a priori unknowable number of impressions that could fill that break.
If you have a 90-second pod, that could be one 60-second ad and one 30-second ad. It could be three 30s. It could be a combination of 15s, 30s, 45s and 60s.
Those opportunities aren’t discrete. They’re related to each other because viewers experience them serially.
Everybody has seen what happens when this is done poorly — repeated ads back-to-back, the same creative shown multiple times in a single break or competing brands running one after another. None of those outcomes are desirable for consumers or marketers.
You can think of podding as being synonymous with the TV ad break.
The OpenRTB 2.6 protocol formalized support for podding in 2022, but adoption has been gradual because the ecosystem had already adapted itself around single-impression workflows.
So why did the market tolerate the inefficiency for so long?
Part of the persistence was inertia. Programmatic advertising came from the web, so the ecosystem simply applied web transaction mechanics to television while streaming matured.
Over time, though, streamers and content owners began pushing pod awareness further up the supply chain.
First, ad servers understood they were filling 90 seconds rather than managing isolated impressions. Then SSPs and exchanges started receiving podded supply instead of fanned-out requests.
Now the industry is at the point where DSPs increasingly want to receive podded requests directly rather than fragmented bid streams.
What changes when buyers evaluate the full pod instead of isolated impressions?
The biggest benefit is that buyers can optimize the creative experience coherently.
A buyer can now decide whether a 15-second, 30-second or 60-second creative makes the most sense inside a specific pod without worrying about competing against itself inside the same break.
That awareness improves the delivery of the advertiser’s message because the buyer understands the broader context of the pod — the user, the device and the stream itself.
As adoption evolves, more sophisticated capabilities become possible too.
The OpenRTB specification already allows for buyers to target specific positions within a pod. Research consistently shows viewers pay the most attention to the first and last ads in a break, which means those positions should theoretically command premiums.
Right now, a lot of that positioning happens manually through deals. Programmatic podding creates the possibility for that valuation to happen dynamically and at scale.
How does podding change pricing dynamics for publishers?
Initially, the pricing impact may not dramatically change because sellers still ultimately decide how to position ads inside the pod.
But the efficiency gains matter.
Instead of processing 12 bid requests for one ad break, the ecosystem processes one. That reduces compute costs across the entire supply chain.
Those savings should ultimately flow through the ecosystem — benefiting buyers while also increasing efficiency for sellers.
There’s another important effect too: cleaner signals around scarcity.
Under highly duplicative systems, truly scarce premium supply can get drowned out by noise created through excessive request generation.
With podding, buyers get a more accurate picture of actual inventory availability, which theoretically allows high-value scarce inventory to command appropriately higher CPMs.
Why are transaction IDs such a critical part of this transition?
Because podding without transaction IDs could actually make duplication worse.
If bad actors duplicated podded requests instead of single impressions, they’d be duplicating a much more efficient and valuable signal.
Transaction IDs provide a way to measure duplication rates across supply paths.
They’re not used to make impression-level bidding decisions. Instead, they’re used in aggregate to evaluate which supply paths are more efficient so spend can be steered toward cleaner inventory when buyers have flexibility.
That distinction matters because guaranteed deals and direct commitments won’t suddenly stop transacting differently. But where buyers have optionality, efficiency becomes a meaningful signal.
How will dynamic ad pods impact advertiser outcomes?
We use many signals within each bid request to determine which opportunities are most likely to drive advertiser outcomes. Starting Aug. 31, we expect that bid requests for long-form video ad breaks will be expressed via Dynamic Ad Pods with Transaction ID, and our goal-seeking algorithms will prioritize and value this added context more highly than the signals provided on fanned-out single impression requests.
That doesn’t mean we’ll stop buying traditional fan-out supply, and it won’t affect PGs or existing commitments. But wherever our valuation systems have the ability to steer spend between supply paths, all else being equal, we’ll prefer the more efficient path with the lower duplicated request rate.
The goal is to remove the “first-mover disadvantage” that has discouraged some publishers and SSPs from adopting podding earlier. Under the old dynamic, cleaner supply paths could lose out simply because competing paths generated more requests.
What we’re trying to do is encourage the more efficient behavior and reduce the duplication that has distorted streaming TV transactions for years.
Are there still holdouts resisting podding?
Not really.
High quality publishers, SSPs and ad servers recognize duplication as harmful even if the ecosystem historically felt forced to tolerate it.
Duplication creates enormous waste in compute costs and distorts market incentives by encouraging everyone to duplicate simply to remain competitive.
What’s changed is that the industry now has a coordinated mechanism to move away from it.
Major ad servers, SSPs and exchanges have largely aligned around supporting podding and transaction IDs ahead of the August transition.
What does streaming advertising look like in three years if this becomes standard?
My expectation is that the industry continues to value context, efficiency, and supply chain transparency.
There are various technology and business limitations that exist today, which result in sellers issuing multiple bid requests for a single opportunity — privacy constraints, packaging requirements and merchandising strategies, for example. But I think these are temporary limitations that we can work to address as an industry.
The broader trajectory is toward making programmatic advertising as efficient as possible.
That matters because the open internet competes directly with walled gardens for advertising budgets.
The more efficient programmatic becomes, the better it can compete on return on ad spend and advertiser outcomes.
The long-term goal is straightforward: move as close as possible to one request per ad break.
That’s the theoretical optimal state.
And for streaming TV, it’s probably the point where programmatic finally starts behaving like television instead of pretending television is the web.
The Current is owned and operated by The Trade Desk Inc. This information is provided solely for background and is not a representation or guarantee of any future performance.