Why CTV CPMs Are Slumping – And What the Ecosystem Must Do About It

Despite the explosive growth of Connected TV (CTV) consumption and inventory, CPMs are stagnating—or even declining. This may seem paradoxical at first glance: more premium screen time, higher viewer engagement, and rising brand demand should lead to higher prices. But as our own analysis at Adlook shows, the issue is not about demand or screen size—it’s about signal fidelity and ecosystem maturity.

‍Let’s unpack the current CTV landscape, the reasons behind softening CPMs, and what must change to restore value and advertiser confidence.

Our recent analysis of global CTV bidstream data, based on direct connections to virtually all major CTV SSPs worldwide, reveals the structural challenges at the heart of CTV monetization.‍

• Only 69.3% of bid requests include any content info (2025 YTD data)
• Key attributes like genre (56.5%), content rating (52.5%), and livestream status (60.1%) are inconsistently populated.
• High-value signals like title (12.4%), season (2.5%), episode (2.7%), and producer ID (0.6%) are barely present.
• Over 90% of bid requests originate from the US, with the supply heavily concentrated in a few top apps—while the longtail remains underutilized but steadily growing.

Without robust, consistent signal coverage, brands are very often buying into a black box. And in programmatic, opacity kills value.

a. A legacy of premium pricing without precision
CTV was initially priced at a premium because of its “TV-like” experience. But the quality of impressions varied widely. Many players in the ecosystem capitalized on this arbitrage opportunity, offering bundles of variable-quality inventory under a uniform CPM. Now that buyers are demanding accountability, that model is collapsing.

‍b. Weak incentives and signal fragmentation
In the open web, cookies (for all their flaws) created a unified targeting layer. In CTV, signal ownership is distributed:

• Publishers control metadata and content signals.
• SSPs pass it through (or not) with little standardization.
• DSPs receive inconsistent or incomplete inputs, limiting targeting and optimization.

No one actor in the supply chain is fully incentivized—or empowered—to improve this. As a result, even top-tier inventory lacks the precision data advertisers expect, or data and inventory is walled gardened.

‍c. Technical and commercial hurdles
Developing a true precision-targeting engine for CTV is still expensive. And with CPMs falling, the ROI for innovation isn’t there—yet. Moreover, many transactions still occur via programmatic guaranteed (PG) or private marketplace (PMP) deals, sidelining the potential of open exchange optimization.

‍d. Walled gardens and monopolistic instincts
Wherever user-level IDs are unavailable, large players naturally gravitate toward closed ecosystems to consolidate data and extract rents. CTV is no exception. Instead of building open, interoperable standards, we see a splintered landscape where data and access are traded like commodities.

a. Better Signals, More Granularity
Our findings are clear: when content metadata is richly populated, advertisers can align their buys with precise audience moments.
Examples:

• Genre & content rating: Ideal for aligning tone and brand suitability.
• Title, series, episode, season: Enables franchise-level targeting (e.g., advertise during kids’ content, but exclude horror).
• Livestream: Great for sports, events, and moment-based targeting.
• Contextual use case: Targeting premium scripted drama during evening hours, using content metadata combined with timestamp.

We are working directly with SSPs and publishers to increase this signal population and standardization. The result: more relevant impressions, higher advertiser confidence, and rising CPMs.

‍Unlocking Value through Precision Targeting: A Hidden Advantage for Advertisers
While the industry races to enhance bidstream signal quality and adoption, savvy advertisers already leveraging these signals are seeing outsized returns. Campaign data from our platform shows that bid requests enriched with content-level bidstream attributes—such as genre—consistently yield 30% to 40% higher advertiser willingness to pay compared to those lacking such metadata. However, the current distribution of these signals is uneven, and quality varies across supply partners and inventory pockets. At Adlook, we’ve invested heavily in normalizing and harmonizing this fragmented data to ensure consistency and reliability across our buying engine. This effort not only enables more accurate targeting but also reveals undervalued inventory segments—areas with lower bid density where advertisers can still secure premium placements at favorable CPMs. These are extreme-ROI opportunities that will likely diminish as signal adoption becomes more standardized in the coming quarters. Now is the moment to act.

‍b. FAST Channels: A Rising Opportunity
Free ad-supported streaming TV (FAST) channels offer a structure closer to linear, with EPG data and defined programming. If signal standards improve, FAST inventory will become a key lever for precision targeting and high-value audience segments.

‍c. Optimization Engines & Measurement Maturity
Today’s CTV buying lacks the feedback loops common in open web display:

• No real-time conversion feedback
• Sparse incrementality studies
• Limited attribution signals

That’s changing. Emerging partnerships between DSPs, supply owners and data partners are unlocking performance measurement at scale—combining ACR data, panel insights, and probabilistic models. We’re still in the early innings, but the direction is clear.

Five Key Barriers:

• Inertia from high-margin legacy models: Easy money from mixed-quality inventory removed the pressure to innovate.
• Signal passivity: No single player “owns” signal maximization—leaving data gaps and inefficiencies.
• Tech debt vs. ROI: Building precision CTV tech is costly and hasn’t yet been commercially rewarded.
• Fragmentation: Commercial and technical fragmentation lengthens the supply chain and reduces transparency.
• Control vs. collaboration: Walled garden instincts prevail, undermining ecosystem efficiency.

‍

Our Strategy at Adlook:

• Leverage our battle-tested buying engine (10+ years of outperforming KPIs) to bring true optimization to CTV.
• Secure deep SSP integrations to access and enrich bidstream data.
• Partner with publishers directly to demonstrate that better signals lead to better targeting—and better yield.
• Show the impact: the more granular and accurate the data, the higher the advertiser’s willingness to pay.

The CTV transformation didn’t begin yesterday—but it feels like we’re still in the airport, heading to our gate. The journey ahead involves massive acceleration, driven by better signals, smarter measurement, and performance-optimized buying.

‍The winners will be those who control access, utilize it effectively, and have the buying intelligence to turn fragmented signals into precision outcomes.

‍At Adlook, we believe we’re uniquely positioned to do exactly that.

‍Let’s board.