Click fraud

Click fraud is clicks on ads that do not come from people with a real need, generated instead by bots, hired workers or software to drain an advertiser’s budget or to collect commissions. Google calls these clicks invalid clicks.

What is click fraud?

Fraudulent clicks show up in reports like real clicks but never turn into orders or leads. You pay for them, and the algorithm learns the wrong lesson: it assumes the placement works and pushes more money into it.

The Media Rating Council splits invalid traffic into two groups. GIVT (general invalid traffic) can be filtered with fixed lists and rules, for example by blocking data center IP addresses. SIVT (sophisticated invalid traffic) imitates real users and can only be detected by analyzing many signals together (MRC, June 2020).

How big is the damage?

Juniper Research estimated that ad fraud took $84 billion in 2023, about 22% of the $382 billion spent on online advertising, with mobile at around 30% (Marketing Brew, October 2023). The report was produced with a company that sells anti-fraud tools, so treat the figure as a rough reference only.

AppsFlyer’s State of Fraud 2026 report analyzed more than 100 billion installs across 246,000 apps. Fraud rates in 2025 were roughly flat, but the gap between the affiliate channel and large self-reporting networks (SRNs) such as Google and Meta reached 36x (AppsFlyer, April 2026).

Types of click fraud

• Bots: software that mimics a person browsing the web and clicking ads.
• Hired humans: “click farms” that use hundreds of phones to click, which makes them harder to filter than bots.
• Competitors: rivals click your search ads to burn through your daily budget.
• Fraudulent publishers: they click their own ads, pay people to click, pile ads on top of each other (ad stacking) or hide ads under other buttons (clickjacking).
• In-app fraud: click flooding sends masses of fake clicks to “claim the spot” in advance, while click injection slips in a click just before an app finishes installing. Both aim to take credit for installs they did not drive.

How to detect click fraud

• An unusually high CTR with a conversion rate close to 0.
• Clicks concentrated on a handful of websites or apps, or in late-night hours.
• Time on page near 0 seconds, and lead forms full of meaningless phone numbers.
• Click-to-install times that are far too short (a few seconds) or spread suspiciously evenly over many days.
• In Google Ads, add the “Invalid clicks” column to see how many clicks Google has already filtered out.

How to prevent it

Google filters invalid clicks automatically before charging you. Invalid clicks it catches later are refunded as a credit on a future invoice (Google Ads Help, About invalid traffic). The rest is up to you:

• Set location targeting to “Presence” (people physically in the area) instead of “Presence or interest”.
• Add negative keywords, and exclude low-quality websites and apps from Display campaigns.
• Put reCAPTCHA or a verification step on lead forms.
• Pay affiliate partners on CPA or revenue, never per click.
• For apps, turn on your MMP’s fraud protection features.

Hypothetical example

An electronics shop in Da Nang runs search ads for “air conditioner repair.” Over one week, clicks between 10 p.m. and 2 a.m. rise 40% while the number of phone calls stays the same. The shop limits its ad schedule to 7 a.m. through 9 p.m., switches location targeting to “Presence” and adds negative keywords.

Click fraud vs install fraud and accidental clicks

Install fraud fakes app installs, while click fraud fakes clicks. The two often go together. Accidental clicks caused by ads placed right next to navigation buttons are also classed by Google as invalid clicks, even though nobody meant to cheat.

Frequently asked questions

Are CPM campaigns safe from fraud?

No. Impression fraud still happens, and fake clicks still push the system to optimize in the wrong direction.

Should you buy a click fraud protection tool?

Check Google’s Invalid clicks column first. Many of the clicks that third-party tools flag as fraud have already been filtered by Google and never billed.

If you suspect fake clicks are eating into your ad budget, Adtek’s measurement team can go through your account data with you. Get in touch via our contact page.