Invalid Traffic (IVT)

Invalid traffic (IVT) is the industry's umbrella term for any ad activity that doesn't come from a genuine user genuinely viewing or interacting with an ad: data-center bots, automated fraud, hijacked devices, accidental and incentivized clicks, and everything in between. The MRC's invalid-traffic standards split it into two tiers that decide how it's caught — GIVT (general invalid traffic, the routine, list-filterable kind — the GIVT-GENERAL-INVALID-TRAFFIC entry) and SIVT (sophisticated invalid traffic, the disguised fraud requiring forensic detection) — a distinction that exists precisely because filtering the polite bots tells you nothing about the impolite ones.

The two tiers and why the split matters operationally: GIVT is the self-announcing half — declared crawlers (GOOGLEBOT honestly naming itself), data-center IP ranges, known-bot user agents, pre-fetch — removed by routine list-and-parameter checks that any MRC-accredited measurement applies before numbers reach the buyer (hygiene, table stakes). SIVT is the engineered remainder — the CLICK-FRAUD and CLICK-INJECTION schemes, DOMAIN-SPOOFING's laundered inventory, BOT-TRAFFIC rotating residential IPs to fail list-matching, hijacked devices, and ad stacking and pixel stuffing — detection of which requires forensic analytics (timing distributions, render verification, behavioral fingerprints, supply-path audits), not lists. What IVT costs, beyond the obvious wasted spend: it corrupts every metric downstream (inflated impressions and clicks, flattered CTR, fake conversions that train SMART-BIDDING toward the fraud, polluted ENGAGED-SESSION and analytics data), it distorts attribution (bots that 'convert' steal credit from real channels), and it erodes the trust the open programmatic market runs on. The defense stack the sophisticated buyer runs: GIVT filtration assumed and audited (which lists, applied where), third-party verification vendors (the IAS/DoubleVerify-class SIVT detection), supply-path transparency (ads.txt / sellers.json to cut spoofed inventory), allow-lists and direct deals over open-exchange long-tail, and incrementality testing as the ultimate backstop (a HOLDOUT-style test reveals whether 'conversions' were ever real — fraud can't survive a clean counterfactual). The honest framing this glossary keeps: 'we filter IVT' deserves the follow-up 'general, sophisticated, or both, and verified by whom?' — because clean-GIVT reporting is hygiene mistaken for verification.

Invalid traffic matters in every programmatic and performance buy — most acutely in open-exchange, long-tail, and incentivized inventory where SIVT concentrates, and in any account where metrics look too good (suspiciously high CTR, conversions from odd geographies, traffic spikes without campaign changes). It matters as measurement hygiene beyond ads (unfiltered IVT pollutes site analytics) and as a vocabulary-precision test in vendor conversations. The discipline is the two-tier defense — lists for GIVT, forensics and verification for SIVT, supply-path hygiene to cut spoofed inventory, and incrementality as the backstop that prices what any of it was actually worth — with clean-GIVT reporting never mistaken for fraud-free traffic.

The IVT vocabulary and the GIVT/SIVT split come from the MRC's invalid-traffic detection guidelines (2015 onward), which standardized what list-based filtration must remove and named the sophisticated remainder - turning 'we filter bots' from a slogan into an auditable, accredited claim as programmatic's fraud problem scaled.

Etymology: source.

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Areas of marketing where invalid traffic (ivt) is a core concern: