Español
Português (Brasil)
Français
Deutsch
Italiano
Nederlands
Polski
Svenska
Türkçe
Русский
日本語
한국어
简体中文
Tiếng Việt
Bahasa Indonesia
繁體中文
Čeština
Dansk
Norsk (bokmål)
Suomi
Ελληνικά
Română
Magyar
Українська
ไทย
हिन्दी
العربية
עבריתWhat Is Invalid Traffic (IVT)?
Invalid traffic (IVT) is any ad impression, click, or conversion that doesn't come from a genuine, engaged human viewer, from simple bots to sophisticated fraud.
Invalid traffic (IVT) is any ad impression, click, or conversion that doesn't come from a genuine, engaged human viewer, ranging from known bots and data center traffic to more sophisticated fraud like click injection and domain spoofing. The IAB splits it into General Invalid Traffic (GIVT), easily filtered patterns like known crawlers and spiders, and Sophisticated Invalid Traffic (SIVT), harder-to-detect fraud that requires active detection rather than simple list matching.
GIVT vs SIVT#
| Type | Examples | Detection method |
|---|---|---|
| GIVT | Known bots, search engine crawlers, pre-fetch requests | Static blocklists, known signatures |
| SIVT | Click farms, device ID spoofing, ad stacking, domain spoofing | Behavioral analysis, statistical anomaly detection |
GIVT is largely a solved problem, reputable ad servers and networks filter it automatically. SIVT is the ongoing fight, since it's specifically designed to look like real traffic.
How IVT shows up in native advertising#
Native networks route through layered supply chains, an ad exchange or supply-side platform sitting between the publisher and the advertiser, which creates more surface area for fraud than a direct buy. Common native-specific IVT patterns include domain spoofing, fraudulent inventory disguised as a premium publisher's ads.txt-listed domain, and click flooding on app-install campaigns, a fraud pattern covered in more depth under cost per install.
Signals that suggest IVT in your own campaigns#
- Conversion rate that's suspiciously consistent across placements that should behave very differently.
- Click-to-session ratios where a large share of clicks never generate a single additional pageview.
- Traffic spikes concentrated in narrow time windows or unusual geos relative to your targeting.
- A publisher or site ID generating volume wildly out of proportion to its known traffic size.
Verification and prevention tools#
Most large advertisers run third-party verification, through ads.txt and sellers.json checks and independent measurement vendors, alongside their own conversion tracking anomaly monitoring. These files don't stop fraud directly, but they make it much harder for a fraudulent seller to impersonate a legitimate publisher's inventory undetected, since buyers can cross-check the declared supply chain against the SupplyChain Object passed with the bid request.
Why IVT matters even if your network already filters it#
Networks filter GIVT and a meaningful share of SIVT before you ever see the bill, but no network claims to catch all of it, and the exact filtering rate isn't independently auditable by an individual advertiser. This is a big part of why experienced media buyers watch their own traffic source and placement-level conversion data rather than trusting a network dashboard alone, the same discipline that shows up in disputes over click fraud specifically.
How OpenAdLibrary helps#
Comparing your own placement performance against how a creative is performing broadly across the index is one way to sense-check whether a specific publisher is behaving normally. OpenAdLibrary's ad intelligence tools show where and how long creative runs across networks, useful context when a placement's numbers don't add up.
