DV360 and the Integrated DSP Dilemma

Context

There are two kinds of demand-side platform. There are the independents, who own no media and sit purely on the buyer’s side of the table. And there is Display & Video 360, which is a buying tool sold by the company that also owns the largest ad exchange, runs the auction, and owns some of the most valuable supply on the internet. DV360 is not the biggest DSP by accident. It is the biggest because it is plugged directly into Google.

I work on the data behind a DSP and an ad exchange, so DV360 is directly adjacent to what I do every day. And the thing I keep coming back to is that DV360’s greatest strength and its greatest liability are the same fact.

What I noticed

DV360’s edge is not that it has a cleverer bidder than The Trade Desk or Amazon. Its edge is access. Because it lives inside the Google Marketing Platform, it inherits Google’s demand relationships, Google’s identity and audience data, and Google’s supply — including inventory a competing DSP either cannot buy at all or can only buy on worse terms. The clearest example is YouTube. If you want to buy YouTube programmatically at scale, DV360 is the front door. That is not a feature you can replicate; it is a consequence of ownership.

Stack that up and the reach argument writes itself. One buying tool, first-class access to the world’s dominant video platform, Google’s audience signals, and the largest exchange to source the rest of the open web. For a large advertiser trying to spend a nine-figure budget efficiently, DV360 is the path of least resistance. It is the default at scale precisely because the integration removes friction that independents have to work around.

The product thinking

Here is the structural tension. The same vertical integration that makes DV360 powerful — owning the DSP, owning the AdX exchange, and owning dominant supply — puts Google on every side of the transaction at once. It is the buyer’s tool, the seller’s platform, and the auctioneer. When one company controls the buy side, the sell side, and the referee, the buyer can never fully verify that the tool is routing spend to what performs rather than to what Google also happens to own, or that the auction is not being shaded in the house’s favour. I do not say this as an accusation about intent. I say it because I know the structure: when you own both ends, the temptation to favour your own supply is constant, and the only way to credibly prove you are not doing it is to not own the other end. DV360, by design, owns the other end.

That structure is not a side issue. It is the exact thing the United States government took Google to court over.

The antitrust liability

In 2025, a federal court found Google liable for illegally monopolising key parts of the ad-tech stack, and the case moved into a remedies phase focused on what to do about it. As of mid-2026 the remedies picture is still unresolved and moving quickly, so treat specifics as provisional. Based on reporting, the DOJ has pushed for structural relief — forcing Google to divest AdX and potentially open-source or divest parts of its publisher ad server — while Google has argued for behavioural remedies like sharing bid data and has warned that a forced sale is unworkable and would break the technology (vendor / party positioning). Judge Brinkema reportedly raised practical concerns during closing arguments, including that there is no obvious buyer for AdX and any sale would face its own antitrust review (court-reporting, not a final order). A remedies ruling was expected imminently as of early 2026; verify the current status before relying on any specific outcome, because appeals will likely stretch this out for years.

The point for a product analyst is not to predict the ruling. It is to notice that the remedy the government wants attacks the seam that makes DV360 what it is. The integration between the buy side, the exchange, and the supply is both the moat and the target. You cannot fully cleave the exchange from the stack without changing the economics and the data flow that make DV360 the frictionless default.

The buyer’s dilemma

So here is the dilemma every large advertiser is quietly holding. On one side: unmatched reach, first-class YouTube access, and Google’s data, all in one tool. On the other: you are buying through the company that also runs the auction you are bidding into and owns a chunk of the inventory you are bidding on, and that company is under an active antitrust remedy that could reshape the very integration you are relying on.

Most large buyers resolve this by not choosing. They run DV360 for the reach and the YouTube access they cannot get elsewhere, and they run an independent DSP alongside it for the neutrality and the leverage. Multi-DSP is not indecision; it is a hedge against exactly this structural conflict.

The PM lesson

DV360 is the clearest case I know of a product where the source of the moat and the source of the existential risk are literally the same asset. Vertical integration gave it default status at scale — and the same vertical integration is the thing a court has found unlawful and may force apart. You cannot un-bundle the risk from the strength, because they are one design decision viewed from two directions.

If I were the PM on the buy side, I would not bet the roadmap on the integration surviving intact. I would build for the world where AdX is separated and the privileged data flows get thinner, and treat DV360’s current reach advantage as a lead to be defended rather than a permanent moat. The most durable version of DV360 is the one that earns its default status on execution, not on owning the auction — because owning the auction is exactly what is now on the table.