MediaMath filed for Chapter 11 bankruptcy protection on June 30, 2023. The company is 16 years old. It helped build the programmatic advertising industry in the early days of real-time bidding, pioneered audience management at scale, and attracted over a billion dollars in investment over its lifetime. Its collapse is not a management failure story, though there were management failures. It is a structural story about what happened to the independent middle layer of the programmatic supply chain when the forces reshaping the industry converged simultaneously.

The three forces that killed an independent DSP are worth naming clearly, because they are the same forces that will continue to reshape the competitive landscape in the years ahead, and because the industry deserves an honest accounting of why a well-capitalized, technically sophisticated independent platform could not survive them.

MediaMath, like every independent DSP, built its core identity and audience management infrastructure on third-party cookies. This was not shortsightedness — it was the only available infrastructure when MediaMath was building its core systems, and cookies were the industry standard that all demand infrastructure was built to use.

When Google announced cookie deprecation in 2019 — originally for 2022, then extended to 2023, then to 2024 — the investment requirement became clear: to remain competitive in a post-cookie environment, DSPs would need to build or acquire identity resolution infrastructure that could operate without browser-based identifiers. This meant either buying into a universal ID solution (UID 2.0, ID5, or similar), building proprietary identity infrastructure anchored in first-party data relationships, or developing clean room integration capabilities that could activate first-party publisher and advertiser data without cookie-based bridging.

For a DSP with The Trade Desk’s scale and capitalization, these investments are expensive but financeable. The Trade Desk has invested hundreds of millions in UID 2.0 infrastructure and OpenPath (a direct publisher integration initiative) specifically to own the post-cookie identity infrastructure layer. For a DSP at MediaMath’s scale — meaningful but not platform-scale — the same investment represented a proportionally larger operational burden relative to revenue, at a time when cookie deprecation timelines were shifting in ways that made the investment urgency unclear.

MediaMath’s public communications had consistently acknowledged the identity infrastructure investment requirement. The company was working on post-cookie solutions. But the timeline uncertainty created repeated decisions about when to commit capital — and each delay in Google’s deprecation timeline created justification for deferring the investment while the cookies continued to work.

Force Two: Walled Garden Consolidation of Performance Budget

The walled gardens — Google, Meta, Amazon — have consistently grown their share of digital advertising revenue over the past decade. This growth has been asymmetric in its impact: it disproportionately affected lower-funnel performance advertising, which is the budget category where independent DSPs competed most directly.

Meta’s Facebook and Instagram, Google’s Search and YouTube, and Amazon’s Sponsored Products all offer direct response advertising with closed-loop attribution — you can measure purchases and conversions against ad spend within the platform. Independent DSPs access open-web inventory with more fragmented attribution. As measurement became more central to budget allocation decisions, the performance advertising budgets that should have flowed to independent programmatic increasingly stayed in walled gardens where measurement was cleaner and optimization was tighter.

This trend accelerated after Apple’s ATT framework degraded Meta’s cross-app attribution in 2021, but the reallocation after the ATT impact was not primarily to independent DSPs — it was to Google Search (where intent signals are strong) and Amazon (where purchase data enables direct attribution). Independent DSPs competing in open-web display and video got a smaller share of the reallocation than the fundamental disruption to Meta might have suggested.

The macroeconomic environment in 2022-2023 made this worse. When advertising budgets contracted under economic pressure, the first budget lines to be cut were typically the less-defensible performance channels — open-web prospecting, mid-funnel display, and audience extension campaigns. These are the campaign types where independent DSPs were strongest. Budget concentration in measurable, justified channels (walled garden performance, search intent) at the expense of less-measured open-web budget directly impacted independent DSP revenue.

Force Three: The Clean Room and Identity Infrastructure Cost Independents Couldn’t Absorb

The premium programmatic market in 2023 requires data clean room integration, authenticated identity resolution, and direct publisher relationship infrastructure. This is not a future state — it is the current expectation from both sophisticated advertisers and premium publishers.

Amazon Marketing Cloud, Disney Select, and NBCUniversal’s One Platform each require DSP-level technical integration to activate their first-party data assets. Building and maintaining these integrations requires dedicated engineering resources and business development relationships that scale with platform size. A DSP that is one-tenth the scale of The Trade Desk cannot allocate one-tenth the engineering resources to clean room integrations and maintain competitive parity — the integrations have fixed costs that don’t scale proportionally.

The result is that large DSPs can build and maintain clean room integrations across Amazon, Disney, NBCU, and dozens of other publisher clean rooms, while independent DSPs face a prioritization problem: which integrations do you build, knowing each requires sustained engineering investment and that you may not have the scale to justify the integration to the data owner?

The Trade Desk’s customer base gives it leverage in these negotiations that MediaMath’s customer base did not. Premium data owners want DSP integrations with the platforms that control the most advertising spend. The identity infrastructure advantages of scale create a winner-take-most dynamic in DSP market position.

What the DSP Market Looks Like After MediaMath

The realistic DSP competitive landscape in 2023 is effectively three-player at the independent level: The Trade Desk, DV360 (Google), and Amazon DSP. MediaMath’s bankruptcy reinforces that the independent DSP market has structurally consolidated around one dominant independent platform.

This has real implications for advertisers. Single-DSP programmatic strategies are more exposed to platform risk and pricing power than two-DSP competitive structures. The DSP market’s consolidation reduces negotiating leverage for major programmatic buyers who depend on competitive tension between buying platforms to maintain pricing discipline.

The specific programmatic budgets that MediaMath managed will migrate to The Trade Desk, DV360, and Amazon DSP over the coming weeks and months. What does not migrate easily is the institutional knowledge, agency relationships, and specific technical integrations that MediaMath built over 16 years. The transition costs are real, and the receiving platforms will compete for the business.

FAQ

What happened to MediaMath’s advertiser campaigns after the bankruptcy filing? MediaMath’s Chapter 11 filing requires it to either reorganize under court supervision or wind down operations. Active campaigns required migration to alternative DSPs immediately, as ongoing campaign delivery depends on functional DSP operations. Advertisers with active campaigns needed to initiate emergency DSP transitions to The Trade Desk, DV360, or other platforms.

Why couldn’t MediaMath raise additional capital to invest in post-cookie infrastructure? MediaMath had raised over $1 billion in investment over its history. The issue was not simply capital availability — it was the investment thesis: at a time when the privacy-driven restructuring of programmatic was creating deep uncertainty about the future value of cookie-based targeting infrastructure, and when walled gardens were consolidating performance budget, the returns required to justify large-scale infrastructure investment in an independent DSP were increasingly difficult to model convincingly for institutional investors.

Is The Trade Desk now effectively a monopoly in independent programmatic buying? The Trade Desk is the dominant independent DSP after MediaMath’s bankruptcy. DV360 (Google) is a major competitor, and Amazon DSP is growing rapidly. The Trade Desk’s independence from walled garden ownership is its primary differentiating claim, and it remains the primary alternative to Google and Amazon for advertisers committed to independent programmatic infrastructure.

What should advertisers with MediaMath contracts do right now? Immediate priorities: identify alternative DSP platforms and begin transition planning, audit programmatic campaign setups that will need to be rebuilt in the receiving DSP, and review any data management or identity infrastructure that was integrated with MediaMath’s DMP capabilities. MediaMath’s DMP assets are a specific area of uncertainty — the data and audience model integrations need to be evaluated for migration or rebuilding.