Invalid traffic refers to ad impressions that are not generated by real users with genuine interest. This includes activity from bots, automated scripts, ad hiding, and other deceptive software. Essentially, it's traffic that artificially inflates ad costs or publisher earnings without bringing in real customers.

This article will help you to identify the main types of IVT and possible solutions to the most common ones.

In case your website, mobile app or CTV channel is IVT flagged, please reach your CSM out for more information.

Content

IVT Categories

Invalid traffic can be divided into two categories: General IVT and Sophisticated IVT.

General IVT

General IVT represents the more common forms of invalid traffic that can be detected through conventional filtration methods. This includes traffic originating from data centers, automated bots and crawlers, and irregular user activities.

Data Center traffic

Impressions that are generated by traffic originating from known data centers and hosting IPs.

Common Causes of IVT:

• IPs on the TAG Data Center IP list.

• IPs on data center IP lists from other industry sources.

Bots and Known Crawlers

Impressions generated by bots and automated tools that are legitimate or self-declared automated services. This includes creative previews, creative scans, and industry-vetted spiders and bots (such as those on the IAB Tech Lab list). For CTV this also includes indirect impressions reported by vetted and known SSAI vendors, but these are not counted towards IVT.

Common Causes of IVT:

• Bots included on the IAB International Spiders and Bots list.

• Automated ad-tech services such as creative previews and malvertising checks.

Irregular Activity

Impressions generated by users or devices that have a history of irregular usage (high-frequency), or generate a high rate of invalid events.

Common Causes of IVT:

• Unusual click or impression frequency or behavior.

• Devices responsible for a substantial volume of invalid events.

• In the context of CTV, extended user sessions missing any user interaction.

Non-rendered Ads

Impressions where an ad was served but never rendered to the user or was impossible to view.

Common Causes of IVT:

• Creatives that are served in a 0x0 placement, a common cause is a responsive ad placement that fails to fully expand after loading a creative (due to requiring user interaction to expand or otherwise).

• Environments without rendering capabilities.

Sophisticated IVT

Sophisticated IVT involves complex scenarios that demand advanced analytics, multi-point verification, and intervention to accurately identify. This category encompasses instances such as proxy traffic, automated and emulated activities, inventory spoofing, and more.

Proxy Traffic

Impressions that were forwarded through a proxy connection, masking the true origin of the user and device.

Common Causes of IVT:

• Devices connected using transparent proxies.

• Anonymous proxies (such as those through residential IPs) that are identified by the IP being present on an industry proxy list.

• Proxies identified using other signals signifying proxy usage.

Automated and Emulated Activity

Impressions generated by bots, viruses, and automated tools that mimic user behavior and falsify user and device info (such as emulating mobile and connected TV devices).

Common Causes of IVT:

• Events generated by virtual machines (e.g., Android VMs that are running on Desktop).

• Automation tools (e.g., Selenium, Headless Chrome, Puppeteer, and misc custom tools and JS scripts) but are purported to be real devices and users. A common non-malicious source is web crawlers designed to get around conventional scraping detection.

• Signals that suggest an illegitimate device or user environment.

Inventory Spoofing

Impressions where the inventory has been misrepresented, such as: serving on a property different than reported or obfuscating content for purposes of fraud.

Common Causes of IVT:

• Discrepancies in Bid Request Data:
  Occurrences where bid request information contradicts the data identified within the impression, such as disparities in domain, bundle ID, user IP, geographic data, and more.

• Invalid Inventory Identification Signals:
  Instances involving faulty signals utilized to identify inventory, including supply chain components with invalid IDs or the absence of expected device-related data.

Falsified or Manipulated

Impressions generated by hijacking a user’s sessions to generate payable events or by reporting false events. Examples include an unintended click, hijacking legitimate ad opportunities for fraud, and falsifying attribution or measurement data.

Common Causes of IVT:

• Occurrences where the source of traffic is manipulated to appear legitimate, distinct from the actual publisher data.

• Clicks generated through scripts or automated processes, lacking genuine user interaction.

• Sessions generated without genuine user intent, often resulting from automated redirects or similar mechanisms.

• Instances in which data, such as device information or viewability signals, is deliberately falsified or misrepresented.

Incentivized, Malware, or Out-of-Store

Impressions that were generated by traffic from incentivized sources, ads injected by malware or adware, and mobile apps that were removed or are unavailable from any official app stores.

Common Causes of IVT:

• Activities generated by users who are rewarded for viewing or interacting with ads (e.g., points towards a gift card for viewing a video or display ad, or access to a video for clicking an ad).

• Sessions initiated by web browser extensions that inject ads on new pages or create pop-ups containing ads.

• Events originating from Android applications not available on the Google Play Store due to violations of the terms of service, such as apps that offer pirated content.

• Activities associated with bundle IDs that have been linked to families of malware or viruses, indicating a potential threat to the integrity of the advertising ecosystem.

Obstructed Ads

Impressions in which a creative is rendered but obstructed by another element or ad. This method also identifies inventory where there is a historical lack of opportunity for an ad to be in view.

Common Causes of IVT:

• A creative that is larger than the placement it is being served into.

• A creative that has other elements/creatives rendered over it.

How to improve your IVT rate

In case your website, mobile app or CTV channel has a high IVT level, the first thing to do is to identify which IVT type is causing the problem.

Please make sure to do the following to keep your IVT rate low:

Monitor and Audit Traffic Sources Regularly

Why it matters: A significant portion of IVT comes from low-quality traffic sources, including paid campaigns, suspicious referrals, or third-party traffic vendors.

Actions:

• Use analytics tools (like Google Analytics) to break down traffic by source and behavior.

• Flag and block IPs or referrals domains that show suspicious patterns (e.g., high bounce rates, short session durations).

• Avoid low-cost, non-transparent traffic providers.

Implement Bot Detection and Mitigation Tools

Why it matters: Bots and scrapers inflate pageviews and clicks, directly contributing to IVT.

Actions:

• Use services like Cloudflare to detect and filter bots.

• Apply CAPTCHA or reCAPTCHA v3 for user verification on high-value interactions (e.g., signups, form submissions).

Improve Ad Placement and Design

Why it matters: Poorly placed ads can lead to accidental clicks, which are a major source of IVT.

Actions:

• Avoid placing ads too close to navigational buttons or in areas with high click potential.

• Use responsive ad formats that align with screen sizes and user behavior patterns.

• Follow the Better Ads Standards to prevent intrusive ad experiences.

Avoid Hidden Ad Placements

Why it matters: Ads designed for desktop can be partially hidden in mobile due their size.

Actions:

• Make sure to correctly define the sizes of your ad placements on desktop and mobile to avoid this problem.

• Ensure that ad placements are not obstructed by other page elements. Please check this on different resolutions and devices.