Advertising fraud is measured differently across campaigns and vendors, but recent benchmarks show that invalid traffic can be especially concentrated in connected TV, mobile apps, and particular devices or markets. In Q3 2024, Pixalate measured 23% invalid traffic (IVT) in both global programmatic connected-TV and mobile-app traffic, compared with 14% in programmatic web traffic. IVT is Pixalate’s source-defined measure and includes ad fraud; it is not identical to every vendor’s fraud-only rate.
Table of contents
- Global rates and estimated losses
- Desktop and mobile web benchmarks
- Mobile-app fraud by store and country
- Connected-TV fraud by region and device
- Open-programmatic regional comparisons
- Fraud spikes and operational scale
- Protection benchmarks and measurement notes
Global rates and estimated losses
Pixalate’s Global Q3 2024 Ad Fraud Benchmarks Report places connected TV and mobile apps above the global programmatic web average. The report measured:
- Global CTV programmatic traffic at a 23% IVT rate in Q3 2024.
- Global mobile-app programmatic traffic at a 23% IVT rate.
- Global programmatic web traffic at a 14% IVT rate.
The year-over-year direction was also unfavorable. Global CTV IVT was up 44% from Q3 2023, mobile-app IVT was up 30%, and web IVT was up 7%. These are comparisons between the stated periods, not forecasts for a later year.
The same Pixalate benchmark estimated $1.5 billion in programmatic mobile-app ad spend lost to IVT and ad fraud in Q3 2024, alongside an estimated $1.4 billion lost in programmatic CTV. The dollar figures are source estimates, not audited industry totals.
CTV-only open-programmatic estimates vary by quarter and methodology. Pixalate’s Q1 2024 CTV benchmark measured a 12% global open-programmatic CTV IVT rate and estimated $528 million in lost spend. Its Q2 2024 CTV benchmark measured a 19.4% rate and estimated $1.14 billion in lost spend. Because these results cover different quarters, they should be read as period-specific estimates rather than a simple annual trend.
| Global benchmark | Measurement period | Reported result |
|---|---|---|
| Programmatic web IVT | Q3 2024 | 14% |
| Programmatic mobile-app IVT | Q3 2024 | 23% |
| Programmatic CTV IVT | Q3 2024 | 23% |
| Open-programmatic CTV IVT | Q1 2024 | 12% |
| Open-programmatic CTV IVT | Q2 2024 | 19.4% |
Desktop and mobile web benchmarks
Pixalate’s Q1 2024 Desktop & Mobile Web Invalid Traffic Benchmarks recorded a 12% IVT rate for global desktop-web programmatic traffic and a 10% rate for global mobile-web programmatic traffic. The device and browser breakdowns show why a single web average can hide meaningful differences.
Among the five leading desktop browsers measured, Mozilla Firefox had the highest reported IVT rate at 22% in Q1 2024. In a later Q3 2024 comparison from Pixalate’s Global Q3 2024 report, desktop Safari traffic had a 30% IVT rate, versus 13% for Google Chrome. These browser figures belong to their respective measurement periods and reports.
Desktop-web rates by market in Q1 2024 were 13% in both Canada and the United States, 12% in Mexico, and 12% in the United Kingdom. On mobile web, Apple-device traffic had a 15% IVT rate and Samsung-device traffic had a 14% rate. Argentina’s mobile-web traffic also recorded a 15% rate.
Taken together, the supplied benchmarks show that desktop and mobile web are not uniform environments. A campaign’s observed rate can change with the browser, device, market, and quarter being measured.
Mobile-app fraud by store and country
Mobile-app risk was 23% globally for programmatic traffic in Q3 2024, according to Pixalate’s Global Q3 2024 report. The app-store split was also substantial: Google Play Store programmatic app traffic had a 25% IVT rate, while Apple App Store programmatic app traffic had a 16% rate.
Country-level results ranged from 14% to 38% in the markets reported. China recorded the highest rate at 38%, followed by Singapore at 37%. Japan measured 22%, India 20%, and Spain 18%. The United Kingdom and France each measured 15%, while Germany measured 14%.
| Mobile-app market or store | IVT rate | Period |
|---|---|---|
| China | 38% | Q3 2024 |
| Singapore | 37% | Q3 2024 |
| Japan | 22% | Q3 2024 |
| India | 20% | Q3 2024 |
| Spain | 18% | Q3 2024 |
| United Kingdom | 15% | Q3 2024 |
| France | 15% | Q3 2024 |
| Germany | 14% | Q3 2024 |
The differences do not establish that one store or country always has more fraud. They describe Pixalate’s measured programmatic populations in Q3 2024. In regional Q1 2024 data, Pixalate also reported that Apple App Store apps in Latin America had a 10% higher IVT risk than Google Play Store apps in that region; this is a relative comparison, not a percentage-point rate.
Connected-TV fraud by region and device
Connected TV produced some of the widest spreads in the benchmarks. Pixalate’s Q1 2024 CTV report measured EMEA open-programmatic CTV traffic at a 30% IVT rate. CTV-app traffic measured 30% in the United Kingdom, 25% in Canada, and 24% in Germany. Amazon Fire TV and Xiaomi had the highest global CTV device IVT risk in that report, at 32%.
The Q2 2024 CTV report gives a different quarter’s picture. North American open-programmatic CTV traffic measured 19.5% IVT. APAC CTV-app traffic measured 18.4%, while India measured 21.4%, Canada 19.7%, and the United States 19.4%. Vizio CTV devices recorded a 40% IVT rate, the highest device-specific figure in the supplied Q2 results.
These figures reinforce the importance of naming the device and inventory type. “CTV” can include open-programmatic traffic, app traffic, and device-specific populations, and the reported rate varies across those definitions and measurement periods.
Open-programmatic regional comparisons
Pixalate’s Q1 2024 regional invalid-traffic benchmarks compare channels across regions. North American mobile-app traffic measured 20% IVT, mobile web 15%, and CTV 12%. In North America, Amazon Fire TV apps measured a 31% CTV IVT rate.
Latin American mobile-app traffic also measured 20%, while LATAM CTV measured 16%. The regional report found that Apple App Store apps in LATAM had a 10% higher IVT risk than Google Play Store apps. The comparison is useful for channel planning, but it does not provide an absolute Apple or Google rate in the supplied data.
APAC had the highest regional mobile-app rate reported, at 30%, and APAC CTV measured 29%. EMEA mobile-app traffic measured 19%, the lowest among the regions reported for that channel. The results are Q1 2024 open-programmatic benchmarks and should not be generalized to every campaign or inventory source.
Fraud spikes and operational scale
Event-driven conditions can change the observed rate. IAS reported in its mobile ad-fraud coverage that U.S. mobile-phone ad fraud was 0.60% in 2021 and 0.70% in 2022. The same source reported that U.S. ad fraud rose 100% during the 2022 Winter Olympics, producing 77 million additional ad-fraud violations.
Operations can be large even when a benchmark is expressed as a rate. In its 2024 investigation, HUMAN said the Konfety mobile ad-fraud operation drove as many as 10 billion bid requests per day while representing fraudulent traffic as legitimate. HUMAN also reported that Konfety used more than 250 decoy applications available on Google Play to support its “evil-twin” operation.
The Konfety figures describe an investigation into one operation and are not an estimate of total mobile-app fraud. They do, however, illustrate how bid volume and application identity can be manipulated at operational scale.
Protection benchmarks and measurement notes
Not all reported rates describe the same population. IAS’s Annual Media Quality Report reported a 0.6% global ad-fraud rate on campaigns optimized against ad fraud using 2023 data reported in May 2024. IAS found fraud violations on unprotected campaigns were up to 14 times higher than on optimized campaigns.
That IAS benchmark should be kept separate from Pixalate’s IVT figures. Pixalate figures cover predominantly buy-side open-auction programmatic traffic, while IAS figures are campaign and media-quality benchmarks with separate optimized and unprotected populations. IVT includes ad fraud under Pixalate’s definition, but it is not identical to every vendor’s fraud-only rate.
For readers evaluating advertising performance, the most useful comparison is therefore a like-for-like one: retain the vendor, channel, inventory type, geography, device or browser, and measurement period. The dollar losses cited above are estimates, and older figures remain tied to their original periods rather than being relabeled as 2026 results.