The most defensible numbers on ad fraud in 2026 come from sources that do not sell click-fraud detection. Juniper Research put the cost near $84 billion in 2023, about 22% of all online ad spend, and projected $172 billion by 2028. Imperva measured automated traffic at 51% of the web in 2024, with malicious bots at 37%. The Association of National Advertisers found that only $439 of every $1,000 of programmatic spend reaches a real consumer. None of the three is your number, but together they bound the problem.
Click fraud is bots and click farms clicking paid ads with no genuine interest in what they advertise. You pay for the click; nobody buys. That definition is narrow on purpose, and it matters for reading the figures below, because each one measures something different.
The $84 billion figure, and what it counts
Juniper Research’s estimate is a market-level number: roughly $84 billion lost to ad fraud in 2023, around 22% of all online ad spend, on a path to $172 billion by 2028. It tells you the scale of the problem in dollars, aggregated across advertisers, platforms, and verticals. It does not tell you what share of any single account is affected, because the analysis spans everything from tiny local campaigns to large brand accounts, and the rate varies enormously between them.
Read it as a ceiling on complacency, not a forecast for your spend. An $84 billion industry-wide loss means the mechanism is real and large. It says nothing about whether your particular campaigns sit above or below the average, and not all of it is recoverable: some is filtered by the platforms before billing, some is credited back, and some is never caught.
Most of the web is automated now
Imperva’s 2025 Bad Bot Report put automated traffic at 51% of the web in 2024, the first time bots have outweighed humans, with malicious bots at 37%. That is prevalence, not cost: it counts what share of traffic is software rather than people.
One honest caveat. Imperva measures web traffic broadly, not paid-ad traffic specifically. But the same automated traffic that fills the open web also lands on your campaigns, and on the conversion forms behind them. When more than a third of all internet traffic is malicious automation, a clean assumption about your own funnel is the optimistic one.
Less than half of a programmatic dollar reaches a person
The ANA’s 2024 Programmatic Benchmark Study found that only $439 of every $1,000 of programmatic ad spend reaches a real consumer. The rest goes to fees, made-for-advertising sites, and invalid traffic. This is the advertisers’ own trade body measuring the supply chain, not a detection vendor selling a fix, which is what makes it hard to wave away.
It is also a figure that has been moving in the right direction: the ANA reported spend on made-for-advertising sites falling from 15% to 6.2% year over year. The waste is large and being worked on, both at once.
What the numbers do and don’t prove
They prove the problem is large and documented by independent measurers: a research firm, a security vendor, and the advertisers’ own trade body, none of which sells click-fraud protection. They do not prove your account loses 22%, or 37%, or any single figure. Averages hide vertical-by-vertical and platform-by-platform spread. A clean lead-gen account on tight exact-match keywords can run far below them; a display-heavy account with loose placements can run well above.
Hold the three lenses apart. Juniper is dollars and a projection. Imperva is a share of all web traffic. The ANA is the efficiency of the programmatic supply chain. Stacking them into a single scary percentage would misrepresent all three.
How to find your own number
The only figure that should drive your spending decisions is the one from your own traffic. That means scoring sessions on signals the visitor’s browser cannot honestly fake: automation markers, fingerprint consistency, network reputation, and behaviour. ClickLens grades each session human, suspect, or bot from those signals, and every deduction has a named reason you can inspect.
Take the three published figures as the reason to look, then measure. Run a free audit to see how your own traffic scores against the same signal categories, or read how the detection methodology is built and tested before you trust any percentage, including ours.
Sources
- Juniper Research (reported by Search Engine Land), “$84 billion of ad spend lost due to ad fraud in 2023”, September 2023. Accessed 25 June 2026. https://searchengineland.com/ad-spend-lost-ad-fraud-2023-432610
- Imperva (Thales), “2025 Imperva Bad Bot Report”, 2025. Accessed 24 June 2026. https://www.imperva.com/blog/2025-imperva-bad-bot-report-how-ai-is-supercharging-the-bot-threat/
- Association of National Advertisers (via WFA), “ANA’s 2024 Programmatic Benchmark Study”, January 2025. Accessed 24 June 2026. https://wfanet.org/knowledge/item/2025/01/21/ana-s-2024-programmatic-benchmark-study-progress-but-challenges-remain