If you have watched ad tech for the last decade, the sensible bet was consolidation. The industry was supposed to collapse toward a handful of scaled platforms — The Trade Desk, Google, Amazon — and a long tail of subscale players quietly disappearing. By that logic, launching a new demand-side platform (DSP) in 2026 would look like stubbornness at best, naïveté at worst.

And yet here we are. A new generation of demand-side and planning tools is proliferating, and buyers are paying attention. AdCritter positions itself as "the world's only AI-first DSP," billing faster plans, better proposals, and instant execution (AdCritter). Pontiac Intelligence built a transparent DSP with no minimum spend and a dedicated connected-TV bidder (Pontiac Intelligence; BusinessWire) and has since extended into streaming audio (Reuters). MOGL gives independent agencies direct access to The Trade Desk and enterprise-grade programmatic infrastructure without the enterprise-grade minimums (MOGL). A handful of still-newer entrants are pushing toward "next-gen buying platforms for sophisticated marketers" aimed squarely at the legacy DSP's accumulated friction.

The natural question: with The Trade Desk, Google, and Amazon already mature and deeply embedded, why would any buyer reach for something new?

Buyers aren't rejecting scale. They're rejecting rigidity.

The answer is not that the incumbents are weak. It is that buyer needs are changing faster than any single platform can absorb them. Signal loss, the deprecation of third-party cookies, the fragmentation of retail media, the shift of spend into CTV, and the collapse of planning and activation into a single continuous workflow — none of these are problems that a legacy UI, however powerful, was designed around.

What buyers are shopping for is intelligence, flexibility, and orchestration. They want software that can plan a campaign, forecast outcomes, and execute across inventory — not three disconnected tools stitched together by a human. They want transparency on fees and supply path. And increasingly, they want a planning surface that is not welded to one buying platform. The new entrants are growing not because they out-scale the giants, but because they are built for the workflow the giants grew out of.

There is also a more prosaic pressure underneath all of this: buyers want lower fees and more value out of their technology partnerships. After years of layered ad tech taxes — DSP take rates, data fees, measurement fees, platform minimums, the quiet margin extracted at every hop — the demand side is scrutinizing what it actually gets for the spend. The economics of contemporary programmatic partnerships are being audited, not just accepted. Buyers are asking whether a partnership earns its place by reducing total cost, improving outcomes, or both — and they are increasingly willing to move the work to the partner that can. This is not frugality for its own sake. It is a rational response to a market where the technology has matured enough that value, not access, should be the differentiator. Any new entrant, and any incumbent, that cannot answer the fee-and-value question credibly is going to find the buying community increasingly hard to convince.

The bias problem nobody built around

But rigidity is only half of it. The deeper driver is bias — and there are two kinds the mature platforms were never structured to solve.

The first is commercial bias. A scaled DSP is, at the end of the day, a business with its own revenue model, its own data assets, and its own inventory relationships. When the same company operates the buying platform, owns the identity graph, and takes a share of the media flowing through it, the optimization is never purely neutral. Spend has a way of gravitating toward the platform's own interests — toward owned data, owned inventory, owned measurement — whether or not that is the best outcome for the advertiser. Transparency on fees and supply path has improved, but the structural incentive to route value through the platform's own stack does not disappear just because the UI gets cleaner.

The second is algorithmic bias. Every DSP's bidding and optimization engine is a set of assumptions, encoded. Which conversions count, which audiences are reachable, which supply is "quality," how frequency is calculated, how bid shading is applied — these are decisions made inside a black box shaped by the platform's data and commercial priorities. Two DSPs given the same brief will produce meaningfully different plans, not because one is wrong, but because each is optimizing toward a slightly different definition of success. When a buyer runs everything through a single mature platform, they inherit that platform's biases wholesale, with no counterweight.

This is precisely where new entrants and independent layers earn their place. A challenger can compete on transparency and neutrality precisely because it has less incumbent stack to protect. And an objective, interoperable intelligence layer can sit above the DSPs, compare their outputs, and surface where one platform's bias is distorting a plan. In a market this concentrated, the absence of an independent vantage point is not a neutral state — it is a quiet transfer of decision-making to the platforms. Innovation here is not a luxury. It is the mechanism by which buyers recover agency.

The twist: AI strengthens the challengers and the incumbents

The engine behind this proliferation is unambiguous: AI-native and agentic capabilities. The new entrants are not winning on inventory access — they are winning on the experience of using them. AdCritter's pitch is literally speed: faster plans, better proposals, instant execution (AdCritter). MOGL markets planning, forecasting, and optimization features built for independents who were previously locked out (MOGL).

But here is the part most commentary gets wrong. The same forces fueling the challengers are about to strengthen the incumbents, not weaken them. The Trade Desk introduced Kokai Zuma — its latest Kokai release — bringing agentic AI, simplified measurement, and a more intuitive experience to market (The Trade Desk Investor Relations). Industry reports describe The Trade Desk building a conversational AI interface routing buyers to a growing roster of specialized "Koa Agents" (Adweek), with Digiday reporting the campaign agent is powered by Anthropic's Claude (Digiday). Amazon is leaning the same direction: Performance+ and Brand+ are AI-managed, goal-based campaign types within Amazon DSP, with a predictive model at their core (Amazon Ads), and the company recently shipped ten new capabilities giving advertisers more control over those automated campaigns (ppc.land). Viant launched "Outcomes," its first fully autonomous, open-internet campaign solution (Viant / BusinessWire), and its AI Bid Optimizer is reporting up to 42% lower CPMs (Viant).

In other words, this is not a story of disruptors toppling giants. It is a story of a software-layer rethink that lifts the entire category. The smarter the agents get, the more every platform — incumbent and challenger alike — is forced to compete on the quality of the intelligence they wrap around the same underlying inventory. That competition is the point.

An agentic layer, not another walled garden

This is precisely the gap MCG set out to close. At Media Consulting Group, we have built an agentic layer that sits nested on top of the supply side — on top of the SSPs — rather than competing with the DSPs beneath it. The design choice is deliberate: we did not want to ask buyers to rip and replace. The layer orchestrates supply shaping and planning across the existing DSP landscape, integrating into the platforms buyers already own and the workflows they already run.

That position is also what gives us our perspective. Because MCG is integrated with and across the largest SSPs, we sit at a vantage point most buyers never reach — the supply side of the market, where inventory, pricing, and path quality actually originate. That breadth of integration gives us unusual visibility into the media market: where value is concentrated, where supply paths are inefficient, where inventory is being arbitraged, and where better media value is available to advertisers willing to look past a single DSP's default. It is the difference between reading a market through one platform's report and reading it from the infrastructure underneath all of them.

The reason this matters is structural. The programmatic market does not lack for powerful buying platforms. It lacks for an objective, interoperable intelligence layer that can sit above them — one that can evaluate supply quality, shape inventory, and guide planning without being captive to any single platform's commercial incentives. Nesting that layer on the supply side, and integrating it across DSPs, is what lets a buyer keep their investment in The Trade Desk or Amazon or Viant while still gaining a smarter, independent view of how to deploy it — and, just as importantly, access to improved media value the platforms have no commercial reason to surface on their own. The chain is straightforward: visibility into the market, access to better media value, and the ability to deploy both across whichever DSP best fits the brief — resultingly, performance. The value of an independent supply-side layer is not theoretical. It shows up, eventually, in the numbers that actually matter to a buyer.

Why fragmentation, against the grain, is good for advertisers

There is a temptation to read this proliferation as regression — the market re-fragmenting just as it was finally consolidating. I read it the opposite way. Consolidation optimizes for the platforms. Proliferation optimizes for the buyers.

A market with more intelligent, more specialized, more interoperable options is a market where no single operator can dictate workflow, pricing, or data strategy. It is a market where an independent agency can access the same programmatic infrastructure as a holding company (MOGL), where a transparent no-minimum bidder can compete on CTV (BusinessWire), and where an AI-first DSP can reframe planning itself (AdCritter). The smarter the market becomes, the more innovative and diversified it gets — and that dynamism is what justifies the new entrants. They are not a sign of inefficiency. They are a sign that buyers finally have the leverage to demand better.

The role MCG intends to play

This is also why MCG does not pick a side among platforms. We provide objective guidance across the DSP landscape and integrate seamlessly into existing technology workflows — because we believe the diversification of technology providers, established and emerging alike, serves the buying community. Our job is not to sell a walled garden. It is to make the increasingly intelligent, increasingly fragmented market navigable — to give advertisers an independent agentic layer that works with the stack they have, not against it.

The winners of the next chapter of programmatic will not be defined by who hoards the most inventory. They will be defined by who delivers the most useful intelligence — whether that intelligence comes from a scaled incumbent or a four-person startup. The market is finally rewarding the right thing. We should let it.

Donald Williams is Chief Strategy Officer and Executive Advisor at Media Consulting Group (MCG).