The Great CTV Mirage: Why "Completion Rates" Are Failing Advertisers

In the high-stakes arena of Connected TV (CTV) advertising, a 98% completion rate is often touted as the gold standard of campaign performance. It is a number that looks impeccable on a PowerPoint slide, suggesting that nearly every viewer sat through an entire advertisement. However, according to industry veterans and independent analysts, this figure is not a testament to creative resonance or audience engagement—it is a structural artifact of the medium itself.

As advertisers pour billions into streaming environments, a sobering reality is emerging: the metrics used to justify these massive investments are largely decoupled from human behavior. With CTV spending projected to reach $38 billion by the end of 2026, the reliance on "completion rates" as a primary Key Performance Indicator (KPI) is increasingly being viewed as a systemic failure, one that risks turning a premium advertising channel into a black hole of wasted budget.

The Structural Illusion: Why 98% Is the Baseline

To understand why completion rates are functionally meaningless in modern CTV, one must look at the mechanics of the format. Unlike the skippable pre-roll ads found on YouTube or social media, where a viewer can bypass an advertisement after five seconds, the vast majority of CTV inventory is non-skippable.

When a user streams content on a FAST (Free Ad-Supported Streaming TV) channel or a subscription-supported platform, the ad server forces the commercial to play to completion. The impression is logged, the video finishes, and the system dutifully records a 100% completion rate. As Vlad Chubakov, Associate Director of Programmatic at Delve Deeper, noted in late 2025, these rates "almost always reach 99.99%."

For a marketer, this is not a measurement of quality; it is a measurement of software execution. Treating this metric as a performance indicator is akin to a retail store manager claiming "success" because the lights were on, without noting that the front door was locked and no customers were inside.

Chronology: The Escalation of the Measurement Crisis

The debate over CTV transparency has moved from whispered concerns among programmatic traders to a full-blown industry crisis.

• September 2023: The Video Advertising Bureau (VAB), utilizing FreeWheel data, highlights a 94% completion rate for long-form premium video, setting a high industry benchmark.
• November 2025: Industry expert Vlad Chubakov warns that completion rates are being misused as a proxy for performance, sparking internal debates among agency heads.
• October 2025: Teads launches deterministic CTV measurement to bridge the gap between exposure and real-world sales, moving beyond vanity metrics.
• February – March 2026: A wave of industry activity, including the launch of pre-bid attention targeting by OpenX and TVision, signals an industry pivot toward "attention" over "completion."
• March 2026: PPC Land reveals that over one-third of CTV impressions are served to screens that are powered off, highlighting the failure of current verification stacks.
• May 2026: DoubleVerify reports a 140% surge in CTV fraud schemes, revealing that 34% of monitored impressions were served in non-video environments like screensavers and fitness apps.
• June 2026: A viral LinkedIn thread initiated by media director Austin Lake brings the issue of "vendor self-reporting" to the forefront, exposing how small and mid-sized businesses (SMBs) are being misled by default reporting dashboards.

Supporting Data: The Disconnect Between Tech and Reality

The most damning evidence against current CTV reporting lies in the disparity between the "device" and the "screen." Most modern CTV measurement tools rely on the streaming device (such as a Roku stick or a smart TV app) to report that an ad was played. However, these devices rarely communicate with the television’s power state.

According to data cited by PPC Land in early 2026, more than one in three CTV impressions are delivered to screens that are switched off. Because the advertising stack—including the Demand Side Platform (DSP) and the verification vendor—relies on the streaming device’s internal logs, the system blindly records a "completion."

Furthermore, the quality of the inventory remains highly questionable. DoubleVerify’s 2026 Global Insights report found that nearly a third of impressions were served in environments that were never intended for premium video, such as background music apps or teleconferencing platforms. When an advertiser buys "CTV," they often believe they are reaching a lean-back, captive audience; in reality, they are often bidding on programmatic clutter that provides zero actual visibility.

Official Responses: The Conflict of Interest

The reliance on vendor-supplied reports is the industry’s "dirty little secret." Dr. Augustine Fou, a leading independent analyst, has been vocal about the fundamental conflict of interest inherent in the current model. In a recent discussion, Fou noted that advertisers should never trust reports generated by the very vendors profiting from the spend.

"What else would they say other than the campaigns are performing great, keep spending?" Fou wrote.

This sentiment is echoed by agency professionals like Austin Lake, who manages over $100 million in ad spend. Lake points out that the vast majority of advertisers, particularly those in the SMB and mid-market sectors, do not realize that independent, third-party verification is an option. They operate under the assumption that the dashboard provided by their platform is the objective truth. This "blind trust" has allowed platforms to continue reporting completion rates as the primary metric, shielding themselves from demands for more granular, outcome-based data.

The Implications: Why This Matters for Future Budgets

The failure to move beyond completion rates has profound implications for the future of the advertising ecosystem.

1. The "Race to the Bottom" for Inventory

When algorithms are optimized for completion rates, they are effectively programmed to find the cheapest, non-skippable inventory available. This incentivizes a "race to the bottom" where quality publishers and premium streaming services are penalized in favor of low-cost, high-volume FAST channels or even fraudulent apps. The goal of the algorithm is to fulfill the budget at the lowest cost-per-completed-view, not to find the most attentive viewer.

2. The Erosion of Trust

The IAB Spain study from May 2026 revealed that 44% of advertising professionals have little to no confidence in current CTV measurement systems. When nearly half of the industry admits they don’t trust the data they are using, the long-term sustainability of the channel is at risk. If CTV cannot prove it is more effective than traditional linear TV, advertisers may begin to pivot back to channels where attribution is clearer and fraud is more easily managed.

3. The Shift to Outcome-Based Measurement

The industry is beginning to recognize that "delivery" is not "performance." The push by the IAB for standardized Conversion APIs and the rise of deterministic measurement tools from companies like Teads suggest a shift in priorities. Advertisers are beginning to demand proof of business outcomes—site visits, leads, and store visits—rather than simply a log file confirming that a video played to its end.

Conclusion: A Call for Accountability

The "98% completion rate" is a vanity metric that serves the interests of the platform, not the brand. As the CTV market matures, the industry must pivot from measuring what is easy to count to measuring what is meaningful.

True performance in CTV requires a multi-layered approach: identifying if the screen is actually on, verifying the content environment, measuring attention-weighted exposure, and linking those impressions to real-world business results. Until advertisers stop accepting completion rates as a proxy for success, they will continue to pay a premium for a service that—far too often—is not being watched by anyone at all. The future of CTV depends on this transition from blind trust to rigorous, outcome-driven verification.