As the dust settles from the slow, then sudden, death of MediaMath, it’s clear that this was a one-off case of financial mismanagement rather than a canary in the coal mine for demand-side platforms (DSPs) or adtech.

The event resurfaces the interdependency of the ecosystem and recalls the last time that a major DSP turned off the lights. In that case, Amazon’s acquisition of Sizmek netted a key piece of IP for pennies on the dollar, nicely complementing its streaming, hardware, and sell-side platform (SSP) business. Sizmek was arguably at the height of its game when Amazon bought its ad server and DCO (Dynamic Creative Optimization) unit but was overextended and unable to flip from bold bets to steady profitability. Meanwhile, MediaMath’s stack has been in desperate need of investment. Now that the book of business is no longer part of the transaction (current clients had to shift their ad budgets to new systems over the Fourth of July holiday weekend), who would use cash to buy an adtech fixer-upper? It’s not certain that MediaMath will be acquired, even on the auction block.

The Middle Of The Market Has The Most To Lose

There are clear winners and losers here. The winners are the largest DSPs such as Google Display & Video 360 and The Trade Desk, adtech titans whose financials are laid bare quarterly, and smaller players that exist as key tech investments within a larger ecosystem — think FreeWheel, Samsung, and Roku. Ad spend will also likely go to competent upstarts such as MNTN that were able to raise capital even in this tough environment.

The losers are the independent DSPs serving the middle of the market. This incident will make advertisers skittish to invest in less capitalized partners, given the razor-thin margins that much of the industry runs on. Similarly, the $100 million of unsecured debts that MediaMath leaves behind will hurt midsize vendors and SSPs much more than the larger ones like Magnite, which can likely absorb a $10M write-down.

Lessons For The Entire Adtech Ecosystem

While MediaMath’s failure won’t trigger a marketwide sell-off of adtech holdings, there are lessons to glean from this seemingly inevitable collapse:

  • Marketers must consider financial health and risk as part of the RFP process for DSPs. Pay closer attention to macroeconomics such as hiring, capital raises and expenditures, quarterly financials, and leadership turnover. And consider the downsides that M&A might have for your business. Treat investment of media dollars akin to an institutional investment under your fiduciary stewardship. It’s a bet on retention of data and relationship expertise, as well as business and process continuity. Even if there were mitigating factors here — like end-of-quarter timing aligning with campaign end dates and the fact that many use multiple buying platforms anyway, nobody wants their brand tied to a sinking ship.
  • DSPs need to differentiate or race downmarket. MediaMath continued to sign up customers because there is always a no-frills value customer to be won and business will naturally flow to the lowest-margin vendor. But if margins get too low, the cycle of insolvency and incentivization to cut corners begins again. Many DSPs within larger enterprises stand to benefit from the loss of a competitor but don’t offer unique, differentiating assets that justify the higher premium. DSPs that exist as part of a hardware business should surface attestation signals. Those that are part of telecoms, e-commerce, and social platforms should integrate point-of-sale and foot traffic for attribution. Without differentiation, hungrier SSPs will eat your lunch.
  • Other adtech vendors should seize the moment to push further into the buy side. SSPs in particular have an opportunity to directly pick up these newly orphaned advertisers and capitalize on DSP margin pressure. Everyone in adtech must vet brand tie-ups and integration partners, forecasting over 3–5-year cycles rather than the shorter cycles that buyers would consider. These times demand strong due diligence on new deals and a top-down review of existing buying partners — if you’re floating someone eight-figure sums a month, you have leeway to check their credit.

The dollar amounts that MediaMath owed to even diminutive players in adtech reflects the “earned value” of its reputation and the near certainty that a profitable business at that scale couldn’t vanish overnight. Now that it has, our calculus needs updating.