What is Invalid Traffic (IVT)?

Invalid traffic (IVT) is ad engagement, such as clicks or impressions, that does not come from a real person with genuine interest, so it should not be billed as human attention. It covers bots, hidden ads, accidental double counts, and scripted activity. The Media Rating Council (MRC) splits IVT into general invalid traffic (GIVT) and sophisticated invalid traffic (SIVT) for measurement and filtering.

How IVT appears in campaigns

A valid ad interaction involves a person who can see or choose to click an ad under normal circumstances. IVT breaks that link. Sometimes the cause is benign automation, like a crawler hitting a tagged page. Sometimes it is malicious: a bot network firing bids, or a publisher loading creatives out of view.

Real-time bidding decides prices in milliseconds. Buyers rely on declared fields in the bid stream; sellers control the page or app that creates the request. When those fields lie or when the environment is engineered to count ads no one sees, IVT slips into reported performance. You pay or accrue stats before a human ever makes a choice.

IVT is not identical to ad fraud, though the Venn diagram overlaps heavily. All ad fraud produces invalid outcomes, but some IVT is accidental or policy-defined without a criminal motive. Platforms still reverse charges for many IVT events; advertisers still care because refunds rarely cover full loss and models train on dirty history first.

Measurement vendors and platforms sometimes disagree on labels for the same event. Treat IVT as a risk to manage with your own baselines: week-over-week click quality, assisted conversion paths, and refund rates compared to vertical benchmarks you track internally.

General versus sophisticated IVT

General invalid traffic (GIVT) is easier to spot with lists and simple rules: known data-center IPs, declared spiders, obviously impossible user agents. Sophisticated invalid traffic (SIVT) needs deeper analytics and often manual review because it mimics humans and hides origin.

Both categories inflate metrics. GIVT skews reach and frequency reports; SIVT can fool optimization because sessions look engaged. Triage starts by separating the two so engineering hours go to the expensive problems, not only housekeeping.

Impact on advertisers

Spend. CPM and CPC billing do not wait for audits. Say your blended CPC is USD 6 and 18% of clicks are invalid: on USD 30,000 monthly search spend, that is USD 5,400 in noise before credits, plus any smart bidding drift toward bad segments.

Measurement. IVT poisons lift tests and geo experiments. If a bot cluster concentrates in one region, you might conclude creative wins there when humans did not drive the result. Compare ad platform data with GA4 versus Google Ads views to spot disconnects early.

Downstream work. Invalid clicks that reach a site still hit analytics. Lead forms filled by scripts create junk leads. Sales effort burns on dead records even when the ad network later flags traffic.

According to ClickPatrol’s PPC fraud study, non-human PPC activity remains material; treating IVT as a rounding error underestimates risk in competitive auctions. Pair those benchmarks with your account’s geographic and device breakdowns to see if outliers deserve immediate exclusions or bid adjustments.

Detection and mitigation

Platforms apply baseline filters and may label events in reporting. Read Google’s default invalid click protection and six ways to reduce invalid clicks for starter controls. Add your own checks: placement reviews, IP caps, conversion quality thresholds, and third-party scoring that runs pre-bill.

Technical signals include datacenter and proxy footprints, VPN exit nodes, inconsistent device claims, and impossible click cadence. Behavioral signals include zero scroll depth, identical paths at scale, and bursts that do not match business hours. How we detect fraud explains layered scoring; suspicious clicks ties labels to operations.

At ClickPatrol, we analyze 800+ data points per click to classify sessions before they consume budget, which reduces IVT-driven learning noise. See how ClickPatrol blocks invalid traffic on paid search as a Google Certified Click Tracker, or request a demo for a walkthrough.

Where IVT enters the ad stack

On the web, a page issues ad calls as it renders. In apps, SDKs bundle device and session fields into exchanges. At each hop, resellers may rebundle supply. IVT can originate on the device (malware clicking in the background), on the page (hidden iframes), or in the exchange path (misstated domain or app IDs). Buyers see a normalized bid request; the hard part is knowing whether the described human is real.

Pre-bid filters try to drop bad requests before money moves. Post-bid measurement reconciles what served versus what humans could see. Post-click analytics on your own site closes the loop: if clicks exist but sessions are hollow, IVT likely reached billing even when viewability vendors disagree.

IVT and policy context

Networks publish definitions aligned with MRC and IAB guidance. Advertisers should know what counts as billable versus adjusted. Google’s terms on fake clicks matter when you file disputes. Keep exports of suspicious intervals with IDs, not only screenshots.

Click fraud is one high-cost flavor of IVT for paid search. High CPC niches concentrate abuse because each invalid event costs more. Teams under declining performance pressure should test whether IVT explains part of the drop before rewriting creatives.