37 Digital Ad Fraud Statistics, Trends, and Data Points for 2026

Digital ad fraud is the diversion of advertising money to invalid traffic, spoofed inventory, and bots that mimic human viewers. This briefing compiles verified estimates of global losses, invalid-traffic (IVT) rates by channel, and prevention benchmarks from Juniper Research, the Association of National Advertisers (ANA), TAG (Trustworthy Accountability Group), Imperva, Pixalate, and Statista. Every loss forecast below is an estimate whose value depends heavily on methodology, so figures from different firms are not directly comparable and should be read as ranges, not settled facts.

Executive Summary

• Global digital advertising spend lost to fraud reached an estimated $68 billion in 2022, up from $59 billion in 2021, according to Juniper Research (worldwide).
• Juniper Research forecast that global ad-fraud losses would exceed $100 billion in 2026 and reach roughly $172 billion by 2028 (worldwide, estimate).
• Statista, citing Juniper Research, forecast that North America would account for about 42% of the roughly $173 billion in global ad-fraud losses projected for 2028 (worldwide).
• Automated bot traffic reached 51% of all web traffic in 2024, the first time bots outnumbered humans in a decade, with bad bots at 37%, per Imperva’s 2025 Bad Bot Report (global).
• The ANA’s 2023 Programmatic Media Supply Chain Transparency Study found that only about 36 cents of every programmatic dollar effectively reached a consumer, leaving as much as $20 billion in potential efficiency gains (United States, 2023).
• Made-for-advertising (MFA) sites commanded roughly 15% of programmatic spend and 21% of impressions in the ANA study (United States, 2023).
• Pixalate estimated $1.5 billion in programmatic mobile-app ad spend and $1.4 billion in programmatic CTV ad spend were lost to IVT and fraud in Q3 2024 alone (global).
• TAG Certified Channels held their invalid-traffic rate below 1% (0.98% in the first half of 2022), far under blended industry benchmarks, showing prevention can work at scale (United States).

Key Findings

• Global ad-fraud losses were an estimated $68 billion in 2022, up from $59 billion in 2021, per Juniper Research (worldwide). Source
• The United States accounted for an estimated $23 billion, or about 35%, of global ad-fraud losses in 2022, per Juniper Research (United States). Source
• The top five countries (US, Japan, China, South Korea, UK) accounted for roughly 60% of global ad-fraud losses in 2022, per Juniper Research (worldwide). Source
• Juniper Research forecast global ad-fraud losses would exceed $100 billion in 2026 (worldwide, estimate). Source
• Statista, citing Juniper Research, forecast North America would represent about 42% of the roughly $173 billion in global ad-fraud losses projected for 2028 (worldwide, forecast). Source
• Automated (bot) traffic reached 51% of all web traffic in 2024, per Imperva’s 2025 Bad Bot Report (global). Source
• Bad bots represented 37% of all internet traffic in 2024, up from 32% in 2023, per Imperva (global). Source
• The ANA found that only about 36 cents of every programmatic dollar effectively reached consumers, with as much as $20 billion in efficiency gains available (United States, 2023). Source
• MFA sites captured about 15% of programmatic spend and 21% of impressions in the ANA study (United States, 2023). Source
• Pixalate estimated global CTV IVT at 23% and mobile-app IVT at 23% in Q3 2024, versus 14% for web (global). Source
• Pixalate estimated $1.5 billion in mobile-app and $1.4 billion in CTV programmatic ad spend were lost to IVT in Q3 2024 (global). Source
• Pixalate estimated $1.14 billion in global CTV open-programmatic ad spend was lost to IVT in Q2 2024 (global). Source
• TAG Certified Channels recorded an IVT rate of 0.98% in the first half of 2022, remaining below the 1% milestone for a second consecutive year (United States). Source
• Earlier ANA and White Ops Bot Baseline research projected roughly $5.8 billion in global economic losses to bot fraud in 2019 (worldwide). Source

Global Loss Estimates: Why the Numbers Diverge

Loss forecasts vary widely because each firm defines fraud differently and models it against different spend bases. Read these as estimates, not audited totals.

Juniper Research estimated $68 billion in global ad spend was lost to fraud in 2022, up from $59 billion in 2021 (worldwide). Juniper Research subsequently forecast losses would surpass $100 billion in 2026 and approach $172 billion by 2028 (worldwide, estimate). Statista, republishing Juniper Research data, placed the 2028 global figure at roughly $173 billion, with North America the single largest region at about 42% (worldwide, forecast).

These figures share a common lineage in Juniper Research modeling, so they should not be treated as independent confirmations of one another. A separate widely circulated figure of $84 billion in losses for 2023 also traces to Juniper Research, but the underlying methodology and spend base differ across publications, so the year-over-year path is not perfectly continuous. The practical takeaway is directional: losses are large, measured in the tens of billions of dollars, and forecast to keep rising with total digital ad spend.

Source: Juniper Research, “Digital Advertising Spend Lost to Fraud to Reach $68bn”; Statista, “Global economic losses due to ad fraud by region 2028.” Juniper | Statista

Invalid Traffic and Bot Rates

Invalid traffic (IVT) covers both general invalid traffic (known bots, data-center IPs) and sophisticated invalid traffic (SIVT) engineered to evade detection. Bot-traffic measures from web-security firms and channel-level IVT measures from ad-verification firms answer different questions, so they are complementary rather than interchangeable.

Imperva reported that automated traffic reached 51% of all web traffic in 2024, the first time in a decade that bots outnumbered humans (global). Bad bots specifically accounted for 37% of internet traffic in 2024, up from 32% in 2023 (global). Imperva attributed part of the increase to AI tools lowering the barrier to launching simple bot attacks, with simple bad-bot traffic rising from under 40% in 2023 to 45% in 2024 (global).

These web-traffic percentages describe the open internet, not the fraction of paid ad impressions that are fraudulent. Ad-fraud rates on measured, brand-safe inventory are typically far lower than raw bot-traffic share, which is why channel-level IVT benchmarks (below) matter more for advertisers than headline bot percentages.

Source: Imperva, 2025 Bad Bot Report. Source

Fraud by Channel: CTV, Mobile, and Web

Connected TV (CTV) and mobile apps carry higher measured IVT than desktop web, largely because server-side ad insertion and app-based traffic are harder to verify. Pixalate’s quarterly open-programmatic benchmarks are among the most granular public data on this split.

In Q3 2024, Pixalate measured global CTV IVT at 23% and mobile-app IVT at 23%, versus 14% for web (global). Pixalate estimated $1.5 billion in programmatic mobile-app ad spend and $1.4 billion in programmatic CTV ad spend were lost to IVT and fraud that quarter (global). In Q2 2024, Pixalate estimated $1.14 billion in global CTV open-programmatic spend was lost to IVT, with North America showing the highest regional CTV IVT rate at 19.5% (global).

Pixalate figures cover open programmatic inventory analyzed across more than 100 billion impressions per quarter, so they exclude much walled-garden and direct-sold inventory. That scope makes them a useful floor for open-market risk but not a total-market estimate.

Source: Pixalate, Q3 2024 Global Invalid Traffic and Ad Fraud Benchmarks. Source

Programmatic Waste and Made-for-Advertising Sites

Not all wasted spend is classic fraud. Made-for-advertising (MFA) sites are low-value pages built to harvest ad dollars rather than serve readers, and they sit in a gray zone between fraud and inefficiency. The ANA quantified this waste in its 2023 transparency study.

The ANA found that only about 36 cents of every dollar entering a demand-side platform effectively reached a consumer (United States, 2023). MFA sites commanded roughly 15% of programmatic spend and 21% of impressions in the studied dataset (United States, 2023). The ANA estimated that following best practices could recover as much as $20 billion in efficiency across the open-web programmatic market (United States, 2023). The study analyzed $123 million in spend across 35.5 billion impressions from 21 member companies between September 2022 and January 2023.

MFA waste is distinct from bot fraud: impressions on MFA sites may be served to real humans but deliver little brand value. Advertisers that treat MFA reduction and IVT filtering as separate workstreams tend to capture more of the recoverable spend.

Source: ANA Programmatic Media Supply Chain Transparency Study, December 2023. ANA | Marketing Dive summary

Prevention: What Certified Supply Chains Achieve

Certification and verification measurably lower measured fraud. TAG’s benchmark studies compare IVT in TAG Certified Channels against blended industry benchmarks from MRC-accredited vendors.

TAG Certified Channels recorded an IVT rate of 0.98% in the first half of 2022, holding below the 1% milestone for a second consecutive year (United States). Prior-year TAG studies reported certified-channel IVT rates around 1% while blended industry benchmarks ran roughly ten times higher, with TAG reporting reductions in the range of 84% to 90% versus benchmark in various years (United States). Earlier ANA and White Ops Bot Baseline research projected that more bot fraud would be stopped than would succeed for the first time in 2019, with about $5.8 billion in projected global losses that year (worldwide).

These results show fraud is not intractable: verified, certified supply paths cut measured IVT to a fraction of the open-market rate. The limitation is coverage, since certification protects only the inventory routed through it.

Source: TAG US Fraud Benchmark Study 2022; ANA/White Ops Bot Baseline. TAG | ANA

Original Synthesis

The following insights combine the verified public datasets above. Each is a derived comparison, not a new measurement, and each carries the limitations noted.

1. The certified-channel fraud gap

Formula: blended industry IVT benchmark divided by TAG Certified Channel IVT rate. Using TAG’s own figures, a certified-channel rate near 1% against blended benchmarks TAG cited as roughly ten times higher implies certified paths carry on the order of one-tenth the measured fraud of the open market. Inputs: TAG US Fraud Benchmark Study 2022. Limitation: TAG measures only certified inventory and relies on participating-vendor data, so the comparison flatters certified channels and is not a randomized test.

2. Channel risk ratio, CTV/mobile versus web

Formula: channel IVT rate divided by web IVT rate, Q3 2024. At 23% for CTV and mobile against 14% for web (Pixalate, global, Q3 2024), CTV and mobile app inventory carried roughly 1.6 times the measured IVT of web that quarter. Inputs: Pixalate Q3 2024 benchmarks. Limitation: open-programmatic scope only; walled gardens and direct deals are excluded, so this ratio describes open-market risk, not all buying.

3. Loss-growth versus spend-growth framing

Logic: Juniper Research forecast losses rising from about $68 billion (2022) toward roughly $172 billion (2028), while Statista placed the 2028 total near $173 billion with North America at about 42%. Because these share Juniper modeling, the consistent signal is that losses scale with total digital ad spend rather than shrinking as a share of it. Inputs: Juniper Research and Statista. Limitation: single-source lineage means the two figures corroborate the model, not each other; all are estimates.

Charts to build

• Global ad-fraud losses over time. Data: Juniper Research annual global loss estimates ($59B 2021, $68B 2022, $100B+ 2026 forecast, ~$172B 2028 forecast). Source: Juniper Research. Insight: losses roughly triple across the forecast window. Citation-worthy because it frames the scale of the problem in one line.
• IVT rate by channel, Q3 2024. Data: CTV 23%, mobile apps 23%, web 14%. Source: Pixalate. Insight: CTV and mobile carry ~1.6x web IVT. Citation-worthy for media planners choosing channels.
• Bot versus human web traffic, 2023 to 2024. Data: bad bots 32% (2023) to 37% (2024); automated traffic 51% (2024). Source: Imperva. Insight: bots crossed the majority threshold. Citation-worthy as a milestone stat.
• Certified versus open-market IVT. Data: TAG Certified Channels ~1% versus blended benchmark ~10%+. Source: TAG. Insight: certification cuts measured fraud roughly tenfold. Citation-worthy for prevention ROI.
• Programmatic dollar flow. Data: 36 cents of every programmatic dollar reaches the consumer; ~15% of spend to MFA. Source: ANA. Insight: nearly two-thirds of programmatic value leaks before reaching audiences. Citation-worthy for budget accountability.

Inline chart: IVT rate by channel, Q3 2024 (Pixalate)

Source: Pixalate, Q3 2024 Global Invalid Traffic and Ad Fraud Benchmarks.

Methodology

Sources were selected for verifiability and proximity to primary data. Priority went to the originating research firms and trade bodies (Juniper Research, ANA, TAG, Imperva, Pixalate) and to Statista where it clearly republished a named primary source. Reputable trade journalism (Marketing Dive, GlobeNewswire press releases issued by the research firms) was used only to corroborate figures traceable to those primaries.

Inclusion rule: a statistic appears only if a named publisher, a year or quarter, and a geography could be attached to it. Exclusion rule: figures I could not tie to a specific publisher and period were dropped. Conflicting numbers were handled by attributing each to its publisher and flagging that loss forecasts share a common Juniper Research lineage and are not independent confirmations. Derived insights in Original Synthesis are labeled as ratios or framings built from the cited inputs, not new measurements. Loss forecasts are explicitly presented as estimates that vary by methodology and spend base. Bot-traffic percentages (share of web traffic) are kept distinct from ad-impression IVT rates because they measure different populations. Date of last update: July 2026.

Source Quality

Tier 1 (primary research bodies and trade associations): ANA (Programmatic Media Supply Chain Transparency Study; Bot Baseline); TAG / Trustworthy Accountability Group (US Fraud Benchmark Study); WFA (ad-fraud guidance and MFA coalition definition).

Tier 2 (credible market-research and verification firms): Juniper Research (loss forecasts); Imperva (Bad Bot Report); Pixalate (channel IVT benchmarks); Statista (data aggregation republishing named primaries).

Tier 3 (reputable trade journalism used for corroboration): Marketing Dive; GlobeNewswire press distributions issued by the research firms.

Most Quotable Statistics

• Bots outnumbered humans online for the first time in a decade in 2024, reaching 51% of web traffic (Imperva, global).
• Juniper Research forecast global ad-fraud losses would top $100 billion in 2026 and approach $172 billion by 2028 (worldwide, estimate).
• Only about 36 cents of every programmatic dollar reaches a consumer (ANA, United States, 2023).
• CTV and mobile-app IVT both hit 23% in Q3 2024, versus 14% for web (Pixalate, global).
• TAG Certified Channels held fraud below 1% (0.98%) in the first half of 2022 (United States).

Data Limitations

• Loss forecasts are estimates whose totals depend on how each firm defines fraud and which spend base it models; the widely cited $68B, $100B+, and $172B figures share Juniper Research lineage and are not independent.
• Bot-traffic share (Imperva) measures all web traffic, not the fraction of paid impressions that are fraudulent, so it overstates ad-specific risk if read literally.
• Pixalate benchmarks cover open programmatic inventory only and exclude much walled-garden and direct-sold spend.
• TAG benchmarks measure certified inventory using participating-vendor data, which favors certified channels and is not a randomized comparison.
• MFA waste (ANA) overlaps with but is not identical to fraud; some MFA impressions reach real humans.
• Quarterly and annual figures are not always directly comparable across publishers due to differing scopes and definitions.

Recommended Dataset Fields

Suggested columns for a downloadable CSV compiling these statistics: metric_name, value, unit (USD / percent), channel (CTV / mobile / web / all), geography, period (year or quarter), metric_type (loss estimate / IVT rate / bot share / waste), publisher, underlying_source, is_forecast (Y/N), methodology_note, source_url, date_accessed.

Press Summary

Digital ad fraud remains a multibillion-dollar drain on advertising budgets, though the exact scale depends heavily on methodology. Juniper Research estimated global losses of about $68 billion in 2022 and forecast they would exceed $100 billion in 2026 and approach $172 billion by 2028, figures that share a common model and should be read as estimates. Invalid-traffic pressure is concentrated in newer channels: Pixalate measured 23% IVT on both CTV and mobile apps in Q3 2024, versus 14% on web, with an estimated $1.5 billion in mobile and $1.4 billion in CTV programmatic spend lost that quarter. Automated bots crossed a milestone in 2024, reaching 51% of all web traffic per Imperva. Prevention works where applied: TAG Certified Channels held fraud below 1%. The ANA separately found that only 36 cents of every programmatic dollar reaches consumers, with up to $20 billion in recoverable efficiency. All loss forecasts are estimates and vary by source.

Suggested Headlines

• Ad Fraud Losses on Track to Approach $172 Billion by 2028, Juniper Research Forecasts
• Bots Now Outnumber Humans Online: 51% of Web Traffic Was Automated in 2024
• CTV and Mobile Ad Fraud Hit 23% in Q3 2024 as Newer Channels Lag on Verification
• Only 36 Cents on the Dollar: What the ANA Found Inside Programmatic Advertising
• How Certified Supply Chains Cut Ad Fraud Below 1%

FAQ

How much money is lost to digital ad fraud globally?

Juniper Research estimated about $68 billion was lost worldwide in 2022 and forecast losses would exceed $100 billion in 2026 (worldwide, estimate). Source.

What share of web traffic is bots?

Automated traffic reached 51% of all web traffic in 2024, per Imperva’s 2025 Bad Bot Report (global). Source.

What percentage of web traffic is bad bots?

Bad bots accounted for 37% of internet traffic in 2024, up from 32% in 2023, per Imperva (global). Source.

Which channel has the highest ad-fraud rate?

In Q3 2024, CTV and mobile apps both showed 23% IVT, above web at 14%, per Pixalate (global). Source.

How much CTV ad spend is lost to fraud?

Pixalate estimated $1.4 billion in programmatic CTV ad spend was lost to IVT in Q3 2024 (global). Source.

How much mobile ad spend is lost to fraud?

Pixalate estimated $1.5 billion in programmatic mobile-app ad spend was lost to IVT in Q3 2024 (global). Source.

What are made-for-advertising (MFA) sites?

MFA sites are low-value pages built to harvest ad dollars; the ANA found they took about 15% of programmatic spend and 21% of impressions in its 2023 study (United States). Source.

How much of programmatic spend reaches consumers?

The ANA found only about 36 cents of every programmatic dollar effectively reached a consumer (United States, 2023). Source.

Does certification reduce ad fraud?

Yes; TAG Certified Channels held IVT at 0.98% in the first half of 2022, below the 1% milestone for a second consecutive year (United States). Source.

Which region loses the most to ad fraud?

Statista, citing Juniper Research, forecast North America would account for about 42% of the roughly $173 billion in global ad-fraud losses projected for 2028 (worldwide, forecast). Source.

About This Research

This briefing was compiled by CO Consulting, a research-driven growth-consulting firm. For teams working to quantify and reduce fraud exposure in their own media budgets, we offer a consultation.

Related research

• Google Ads CPC by Industry
• Programmatic & Display Ads
• Retail Media
• Connected TV (CTV) Ads
• Out-of-Home (OOH) Ads