Platform Concentration Is a Time Bomb
Three platforms control most digital ad spend. That efficiency is also fragility, and nobody's building backup systems.
Run this thought experiment. Meta’s advertising platform experiences a 72-hour outage tomorrow. Not a planned maintenance window. Not a minor glitch. Just gone. What happens to Q4 revenue targets?
If that question causes discomfort, it should. Most marketing organizations have engineered themselves into corner they don’t fully understand. The platforms that made digital advertising efficient have also made it brittle. And unlike enterprise software, where vendor risk is scrutinized in procurement, media platform dependency gets waved through because “everyone uses them.”
The OpenAI situation offers a useful parallel. The company expects to lose $10 billion this year yet signs deals worth hundreds of billions with chipmakers and cloud providers. When it announced partnerships with Oracle, Nvidia, and AMD, those companies’ market values jumped $470 billion combined. That’s a lot riding on one unprofitable entity. If OpenAI stumbled, the damage wouldn’t crater the tech industry—other AI companies would absorb capacity—but the ripple effects would be real.
Now examine marketing infrastructure. Three platforms—Meta, Google, TikTok—probably account for 70% or more of digital spend at most organizations. These platforms are wildly profitable, unlike OpenAI. But they’re also under constant pressure from regulators, algorithm updates, privacy restrictions, and platform policy shifts. They could be fine for years. They could also upend entire strategies with a single policy change announced on a Friday afternoon.
The difference: chip companies can pivot to new customers faster than marketing teams can rebuild media strategies. When a platform changes attribution models, restricts targeting, or shifts algorithmic prioritization, marketing teams are stuck. Infrastructure was built on that platform’s infrastructure.
How concentration happened
This occurred gradually, then suddenly. Ten years ago digital marketing was fragmented—dozens of ad networks, multiple social platforms, a chaotic landscape of targeting options. It was messy, inefficient, and diversified. Then the big platforms got better at targeting, simpler to use, and more effective at spending budget. Consolidation followed economic logic.
By 2025, the martech landscape includes nearly 10,000 solutions, yet usage concentrates. The average marketing stack runs 162 to 269 applications, but teams rely heavily on 6 to 10 tools for critical functions. That’s the trap—the illusion of diversity masking actual dependency.
UN Trade and Development data shows digital market concentration worsening. Seven of the world’s ten most valuable companies are now digital platforms. Between 2017 and 2025, the combined sales share held by the top five digital multinationals more than doubled from 21% to 48%. Asset share increased from 17% to 35% in the same period. Globally, governments took 153 actions to restore digital competition in 2024, up from just 14 in 2020. That regulatory pressure signals recognized systemic risk.
The CrowdStrike outage in 2024 cost an estimated $5.2 billion across affected organizations. That was an error by a benign actor. Control Risks warns that in 2025, a deliberate attack on concentrated cloud services could be catastrophic. State actors are foregoing previous norms in digital space. Concentrated technology stacks provide the biggest returns for emboldened threat actors.
Marketing faces the same concentration dynamics without the same regulatory scrutiny. Three platforms control distribution. If one makes a major policy shift, entire business models break. When Facebook prioritized “meaningful social interactions” in 2018, brand reach cratered overnight. When iOS 14 broke mobile attribution in 2021, it took a year to sort out the damage. These weren’t minor adjustments. They fundamentally changed how advertising worked.
What platform risk actually looks like
Platform dependency manifests in ways that don’t show up in quarterly business reviews until it’s too late. Four patterns keep recurring:
Algorithm changes that destroy performance overnight. Platforms optimize for platform goals, not advertiser goals. When those priorities diverge—and they do regularly—performance collapses. The problem isn’t malice. It’s misaligned incentives. Platforms need engagement. Advertisers need conversion. Sometimes those align. Sometimes they don’t.
Privacy restrictions that eliminate targeting capabilities. Third-party cookies are dying, have died, or exist in zombie states depending on who you ask. Either way, targeting capabilities built into strategies are disappearing. If entire acquisition models depend on pixel-based retargeting, the foundation is eroding.
Rising costs with declining efficiency. CPMs on major platforms have climbed steadily while proving performance has gotten harder. Attribution is mess. Multi-touch modeling gives different answers than last-click, which gives different answers than incrementality testing. Platforms keep launching new ad products promising better results if marketers just cede more control to automated buying systems.
Regulatory risk that’s intensifying. TikTok’s ownership saga might resolve fine, but it demonstrates that platforms can become political footballs. GDPR, CCPA, and whatever comes next create compliance complexity that affects how platforms can be used. Every new privacy regulation makes platform-based advertising more complex and less effective.
Digital advertising spending hit $790 billion in 2024, with digital now representing 72.7% of worldwide ad investment. That share jumped from 54.3% in 2019. Absolute digital spend has more than doubled since 2019. Most of that flows through a handful of platforms. The efficiency gains are real. So is the concentration risk.
Why diversification stays theoretical
Standard advice says diversify, but few organizations do it. Not because of inertia—because of rational behavior. If Meta delivers 3x ROAS and an emerging platform delivers 1.5x, money keeps flowing to Meta. The platforms that work keep working until they don’t. By then entire operations have been built around them.
Vendor concentration risk research shows familiar patterns. Businesses that rely heavily on few suppliers for critical functions face several dangers: suppliers know they’re indispensable and can extract better terms, single points of failure can halt operations, informal relationships don’t transfer to new ownership, and lenders flag concentration as destabilizing.
Marketing teams face identical risks but rarely frame them as vendor risk. That’s a mistake. Digital platforms are vendors. Critical ones. The services they provide—audience access, targeting capabilities, measurement infrastructure—are as vital as any enterprise software. Yet procurement processes that scrutinize SaaS vendors for redundancy and exit clauses don’t apply the same rigor to media platforms.
Smart marketers are hedging differently than conventional wisdom suggests:
Building owned audience infrastructure. Email and SMS aren’t exciting, but they’re not platform-dependent. Forty-three percent of consumers don’t trust AI-generated ads, and 44% of marketers can tell when ads used AI. Building direct relationships through owned channels matters more than it did five years ago. If a platform disappeared tomorrow, could the organization still reach customers? Most can’t answer yes.
Investing in experiences that don’t need platforms. The experiential marketing boom isn’t just about creating memories—it’s about creating brand connections that exist independent of buying ads. When 51% of companies are increasing experiential budgets, it’s partly because they’re recognizing platform dependency as business risk.
Actually building for multi-homing. Platform-dependent entrepreneurs figured this out: everything can’t sit on one platform. Research on platform entrepreneurship shows that PDEs employ multiple strategies to reduce dependency—operating on multiple platforms simultaneously, establishing direct customer relationships, and diversifying income sources. Brands should think the same way. That means testing emerging channels before they’re needed, maintaining presence on multiple platforms even when inefficient, and building creative and measurement systems flexible enough to work across different environments.
Getting serious about first-party data. Everyone says this. Few do it well. First-party data strategy means more than installing a CDP. It requires creating value exchanges that make customers want to share data, building systems that activate data without platform intermediaries, and doing the boring work of data hygiene and integration. Organizations that get this right reduce platform dependency for targeting and measurement.
The truth nobody wants
Here’s what’s uncomfortable: organizations probably won’t diversify enough to eliminate platform risk. Platforms are too good, too efficient, too embedded in how digital marketing works. But exposure can be reduced.
Think about cloud infrastructure. No competent CTO would run the entire business on one cloud provider, even if that provider had the best uptime and features. The risk is too concentrated. Yet marketing organizations do exactly this with media budgets. The analogy holds because both are infrastructure. Both are critical paths. Both have single points of failure.
The CrowdStrike incident revealed something about modern infrastructure—dependence on a few providers creates systemic vulnerabilities that cascade across entire ecosystems. Deloitte research shows that 61% of organizations experienced a third-party data breach or security incident in the past 12 months, and 98% of companies have at least one vendor with a documented security breach. Marketing’s vendor concentration creates similar exposure profiles.
Control Risks warns that concentration risk intensifies in 2025 because the traditional model of operational control is becoming obsolete. Digital concentration forces companies to relinquish control over critical assets. Few have the ability to protect their networks; fewer still can manage critical technology dependencies. Marketing faces the same dynamic. Platforms control distribution, targeting, measurement. When they change rules, options are limited.
What gets built instead
The brands that weather platform disruption won’t be the ones that saw it coming. They’ll be the ones that built systems resilient enough that major platform changes become manageable rather than existential. That’s less exciting than finding growth hacks, but it’s what keeps businesses running when platforms inevitably change the rules.
This requires different thinking about what marketing infrastructure means. It’s not just the stack of tools. It’s the dependencies that determine whether business can continue when one piece breaks. Organizations that map those dependencies and build redundancies around critical paths will survive. Those that optimize only for efficiency won’t.
Platform concentration isn’t going away. Returns are too good for platforms and marketers. But recognizing it as infrastructure risk rather than just media mix optimization changes the calculus. The question isn’t whether to use these platforms. It’s what happens when they stop working the way they do now. And whether the organization has built backup systems to avoid finding out the hard way.