Advertising fraud in CTV is one of the industry’s major concerns and, as such, we have already addressed this topic in the past, explaining the main techniques used to generate fraudulent impressions and how different players across the ecosystem can help mitigate it.

This time, we focus exclusively on Non-Human Traffic (NHT) in CTV, its impact on the credibility of the medium, and the strategies required to combat it effectively. This phenomenon poses a direct risk to campaign performance and to trust in the underlying data.

What is “non-human traffic” in CTV?

Non-human traffic (NHT) refers to visits, impressions, or interactions generated by automated systems rather than real people with genuine interest. It includes Invalid Traffic (IVT) and advertising fraud (ad fraud)—related but distinct phenomena:

• Invalid Traffic (IVT): Timpressions or interactions that do not come from humans, generally without the intent to deceive advertisers. Examples include technical errors, internal app testing, or monitoring bots. While typically residual on platforms with robust controls, its volume should remain marginal and not materially affect campaign performance. The real issue arises when this traffic scales, distorting key metrics and undermining media credibility.
• Advertising fraud (ad fraud): deliberate manipulation of metrics for illicit financial gain. Its goal is to deceive advertisers or intermediaries, resulting in direct financial impact. In CTV, NHT translates into ads being served to simulated TVs or bots, rather than real viewers. Metrics may appear legitimate, but they do not represent human audiences—eroding trust in data and directly affecting campaign effectiveness.

Scope and relevance of the problem today

Although the digital advertising industry has been dealing with non-human traffic for years, CTV is not immune—and in some cases may be even more vulnerable than other digital channels. To understand the scale of the issue, several recent studies provide useful benchmarks:

• According to the Pixalate Q1 2024 Benchmark Report , 12% of programmatic CTV traffic was classified as invalid (IVT) in open programmatic advertising (including fraud), translating into approximately $528M in wasted ad spend during that quarter.
• The same report shows that in EMEA, IVT rates were even higher (30%), with popular devices such as Amazon Fire TV and Xiaomi exhibiting above-average invalid traffic rates.
• Another Pixalate reportindicates that up to 28% of open programmatic CTV traffic may come from unauthorized sellers, typically associated with higher IVT rates.
• DoubleVerify addresses this issue in its DV Global Insights 2025, estimating that up to 25% of impressions may fail to reach real humans without specific protections in place.
  There are no single “official” figures, but these findings consistently indicate that—depending on market, buying method, and applied filters—a significant share of CTV traffic may not be viewed by real people, particularly within open inventory lacking strict verification.

How to identify non-human traffic signals in CTV

In traditional web environments, one of the most basic indicators of non-human traffic was the click: unusually high click volumes or CTRs often signaled bot activity or automated interaction. In CTV, this logic no longer applies. Click-through rates in CTV are typically very low or even nonexistent, not because the traffic is low quality, but because of how content is consumed. Many creatives do not include clickable elements and, while some formats enable actions across the funnel, advertisers generally are not trying to drive users directly to a website from CTV.

As a result, CTR cannot be considered a reliable indicator of non-human traffic in CTV. Metrics must be interpreted contextually, taking into account the nature of each device, format, and creative.

More relevant indicators include:

• High impression volumes with low contextual quality
  If impressions do not correlate with inventory verification metrics, actual exposure time, or audience signals, IVT may be present.
• Unusual geographic or device patterns
  For example, traffic concentrations from data-center–associated IPs, suspicious locations, or a lack of reliable device information.
• Sudden unexplained spikes
  Impression increases not tied to content launches, events, or planned campaigns may indicate automated exploitation.

Following IAB standards, consistency checks across established processes can help identify issues early:

OpenRTB / supply chain signals (pre-bid)

• Supply chain validation
  Requiring sellers.json and the SupplyChain object (schain) to verify who is selling the inventory and how many intermediaries are involved—critical for reducing opaque reselling and supply spoofing.
• app-ads.txt / inventory sharing
  Verification of seller authorization for the app bundle (including CTV apps). In inventory-sharing scenarios, the use of inventorypartnerdomain allows partner validation without inflating (app-)ads.txt files to unmanageable levels.
• Semantic field consistency (plausibility checks)
  Coherence across app.bundle, device.make/model, UA, IP, geo, IFA (where applicable), schain, and seller. Spoofing farms often fail longitudinal consistency checks (e.g., the same “TV” appearing across multiple countries, ASNs, or impossible time windows).

Playback signals (post-bid)

Beyond the bid request itself, playback behavior provides valuable signals:

• Heartbeats and temporal cadence
  Real human traffic shows jitter, pauses, buffering, bitrate changes, and first-frame latency. Bots tend toward “perfect” timing or repetitive patterns.
• Household / IP concurrency
  Session concurrency levels that are impossible for a real household, particularly for long-form content.
• QoE and player telemetry
  Start/fail ratios, stalls, bitrate switches, and distributions by device or OS. Persistent anomalies often correlate with NHT or non-standard environments—for example, app bundles generating thousands of concurrent sessions from a small number of IPs.

Ads.cert and advanced authentication signals

Within the programmatic supply chain, Ads.cert (Signed Bid Requests) is also worth highlighting. This IAB Tech Lab specification is designed to harden programmatic transactions through signed requests, increasing trust in environments like CTV where device and supply spoofing can be particularly lucrative.

In practical terms: sellers.json and schain tell you who is selling; ads.cert helps ensure the request itself is not forged.

Cryptographic and manufacturer-native signals

A newer line of defense against this type of fraud is the emergence of cryptographic or manufacturer-native signals:

• Device Attestation via OM SDK
  The IAB Tech Lab has promoted device attestation capabilities within the Open Measurement SDK to combat spoofing, enabling privacy-preserving validation that the environment corresponds to a real, manufacturer-supported device.
• Watermarking / proprietary authentication
  Initiatives such as the Roku–DoubleVerify collaboration support traffic authentication (e.g., “advertising watermarking”) to reduce forged requests that mimic real devices.
  This approach is critical because much of CTV NHT does not look like bot traffic—it looks like a legitimate Smart TV until cryptographic proof or native authenticity signals are required.

Strategies to reduce the impact of NHT in CTV

Combating non-human traffic in CTV is not straightforward, but several proven practices can significantly mitigate risk:

• Use advanced detection technologies
  Specialized vendors can identify suspicious traffic patterns and automatically discard impressions likely to be invalid, preventing them from being billed.
• Buy verified and transparent inventory
  Favoring direct relationships with publishers or curated deals reduces exposure to invalid traffic compared to open exchange buying, where most NHT tends to originate. More controlled inventory—such as private marketplaces or direct deals—typically shows significantly lower IVT rates due to stricter partner oversight and delivery environments.
• Conduct regular audits and quality reviews Ongoing analysis of traffic, device patterns, IPs, and supply paths helps identify irregularities and remove low-value inventory.

Conclusion

Non-human traffic in CTV is not a marginal issue—it has become a real threat to the effectiveness, transparency, and credibility of advertising investment in streaming environments.

While many ecosystem players—streaming platforms, pay-TV operators, smart TV manufacturers, and broadcasters— have strengthened internal controls at the inventory source, the combination of rapid CTV investment growth and the complexity of the programmatic ecosystem continues to create opportunities for automated actors and invalid traffic.

Advertisers investing in streaming must remain especially vigilant. Filters, active verification, and quality analysis are essential to ensure campaigns actually reach real people. Investing in transparency and meaningful metrics is key to protecting ROI, improving performance, and ensuring genuine visibility and authentic audiences.