GIVT traffic
Matches visits against known general invalid traffic: spiders, crawlers and known bot lists.
Action: Block
Rules
Rules define what "clean traffic" means for you, or what quality signals matter most. Pick them per tag or campaign. Each rule runs in one of two modes.
Matches visits against known general invalid traffic: spiders, crawlers and known bot lists.
Action: Block
Detects IPs sending an unusually high number of ad requests or clicks in a short window, typical of automated traffic.
Action: Block
Example: 20 requests per minute per IP
Detects traffic routed through commercial VPNs and proxies, which mask real location and are often used to fake geotargeted delivery.
Action: Monitor
Example: Block only if another rule also triggers
Scores each IP from its abuse history and blocks those above your chosen risk threshold, catching repeat offenders before they click.
Action: Block
Example: Block above 75/100, monitor 50-75
Matches visitors against AdPurify's continuously updated lists of IPs previously confirmed as sources of invalid or fraudulent traffic.
Action: Block
Blocks traffic from specific networks (ASNs), such as data centers and hosting providers, where real consumers rarely browse.
Action: Block
Example: AS16509 (AWS), AS14061 (DigitalOcean)
Lets you manually block or always allow specific IPs and CIDR ranges, such as competitors, offices or trusted partners.
Action: Block
Example: Block 203.0.113.0/24, allow 198.51.100.7
Compares the browser's timezone with the IP's geolocation and flags visitors whose two signals disagree, suggesting proxies or spoofing.
Action: Monitor
Example: Flag if offset differs by more than 3 hours
Flags visitors whose browser language is unlikely for their IP's country, a signal of spoofed location or non-genuine traffic.
Action: Monitor
Example: Flag zh-CN browsers on Brazilian IPs
Catches the same page being reloaded repeatedly within seconds, a common trick to generate extra ad delivery events or clicks.
Action: Block
Example: 5 reloads of the same URL in 30 seconds
Identifies traffic arriving from known Tor exit nodes, which hide the visitor's origin and are frequently used for ad fraud.
Action: Block
Detects ads layered on top of each other or hidden by size, opacity or position, producing ads nobody sees.
Action: Block
Flags ad requests or clicks arriving with no genuine page view, or visitors leaving instantly after the ad fires.
Action: Block
Example: Leaves within 1 second of the ad request
Detects webdriver flags and headless user agents left by Selenium, Puppeteer or Playwright, revealing scripted rather than human visitors.
Action: Block
Flags visitors reporting a screen or viewport of zero pixels or sizes no real device has, a common bot giveaway.
Action: Block
Flags sessions with no mouse, touch, scroll or keyboard activity before the ad loads, something real visitors almost always produce.
Action: Monitor
Example: No interaction within 5 seconds of page load
Detects visitors who block or reject cookies and local storage, making retargeting impossible and the click of little value.
Action: Monitor
Flags visitors with no referrer when combined with other risk signals, since bots often arrive without a traceable source.
Action: Monitor
Example: Escalate to block when 2+ other rules trigger
Only counts traffic coming from approved referrer domains, such as your own site and trusted partners, and rejects the rest.
Action: Block
Example: yourbrand.com, partner.com
Controls whether your ads can load inside iframes or on third-party embeds, preventing clickjacking and unauthorized placements.
Action: Block
Example: Allow framing only from yourbrand.com
Restricts ads to visitors whose browser language matches your target audience, filtering out irrelevant traffic unlikely to convert.
Action: Block
Example: Allow English and French only
Limits ads to chosen operating systems, useful for app campaigns or excluding outdated systems often linked to bot traffic.
Action: Block
Example: Allow iOS and Android only
Applies rules only during chosen hours or weekdays, such as blocking overnight traffic spikes or pausing ads outside business hours.
Action: Block
Example: Block 00:00-05:00 local time, Sat-Sun
Allows or blocks ads on specific pages, site sections or URL patterns, giving you precise control over where ads appear.
Action: Block
Example: Block /checkout/* and /account/*
Checks that the visit comes from the country you target. Flags or blocks every visit outside it.
Action: Monitor
Example: Target: Germany. Flag every visit outside DE
Checks your target at region or city level. Useful for local campaigns and store offers.
Action: Monitor
Example: Target: Texas. Flag visits outside TX
Allows or blocks listed countries. Publishers use it to protect premium geos.
Action: Block
Example: Allow US, CA and GB only
Measures MRC viewability: at least 50% of the ad pixels in view for 1 second.
Action: Monitor
Example: Report the viewable rate per placement
Measures how long the ad stayed in view. Spots placements with very short exposure.
Action: Monitor
Example: Flag placements under 5 seconds on average
Flags ads that load but sit below the fold and are never scrolled into view.
Action: Monitor
Example: Flag if never in view during the session
Detects the same IP clicking your ads repeatedly within a set period, a classic sign of competitor or click-farm fraud.
Action: Block
Example: Block after 3 clicks in 24 hours
Need help choosing? Ask us →
Rules are checks that run on each visit. In monitor mode, AdPurify records a match while the ad tag continues to run. In publisher blocking mode, a matching blocking rule prevents the wrapped ad tag from loading.
Yes. Each wrapped slot has its own rule set.
Lite allows 3 active rules per slot. Professional and Enterprise allow unlimited rules.
Traffic classification is probabilistic, so no signal should be treated as perfect proof on its own. Start every new rule in Monitor-only mode, review matched traffic, then tune that rule's sensitivity before enabling blocking. You can tune rules individually across 30+ checks and use narrow allowlists for trusted IPs, networks or known services.