Invalid traffic: how to spot it and protect your ad account

Invalid traffic can affect your earnings and your ad account at the same time. It usually arrives through a traffic source you paid for, a layout that leads to accidental clicks, or a habit such as clicking your own ads to see where they go. This guide explains what counts, what Google does about it, and how to check your own data calmly.

What invalid traffic is

Google defines invalid traffic as any clicks or impressions that may artificially inflate an advertiser's costs or a publisher's earnings, whether deliberate or accidental. In plain words, it is ad activity that does not come from a real person with genuine interest in the ad. Google's examples include:

• Publishers clicking their own live ads.
• Repeated ad clicks or impressions from one or more users.
• Publishers encouraging clicks, including wording that asks readers to click and ad implementations that cause many accidental clicks.
• Automated clicking tools or traffic sources, robots and other deceptive software.

Google's Ad Traffic Quality pages add incentivized activity, where users are asked or even paid to click on ads. Bought traffic can carry the same problem: Google warns that some traffic services generate clicks and impressions using click bots, or by giving users incentives.

One point matters most. Google accepts that a third party may generate invalid traffic on your ads without your knowledge, but says it is ultimately the publisher's responsibility to make sure the traffic on their ads is valid.

General and sophisticated invalid traffic

The industry standard for detecting and filtering invalid traffic comes from the Media Rating Council (MRC), a non-profit industry body. Its Invalid Traffic Detection and Filtration Standards Addendum, updated in June 2020, sets out two categories:

• General invalid traffic (GIVT): caught by routine filtering using lists or standard checks. Examples include known invalid data-center traffic, bots, spiders and crawlers, and software that does not present itself as a normal browser.
• Sophisticated invalid traffic (SIVT): harder cases that need advanced analytics, several corroborating signals or significant human review. Examples include hijacked devices, bots pretending to be real users, organized human clicking such as click farms, and tricking users into clicking.

Google's Ad Manager documentation says Google tries to identify and filter both general and sophisticated invalid traffic. This is why some invalid traffic is removed at once and some only later.

What Google does when it finds invalid traffic

Google uses automated systems and human reviews to analyze clicks and impressions, and says advertisers are not charged for the invalid traffic it filters. For publishers there are four possible outcomes.

1. Filtered and not paid. AdSense removes invalid traffic from your earnings at three stages. Real-time filtering removes things like double clicks and crawlers as they happen, which is why reported earnings can move around during the day. Clicks judged invalid can still appear in reports, but publishers don't receive any revenue for invalid clicks.

2. Deductions. Estimated earnings are not final. During month-end finalization, further invalid activity can be removed and shown as an invalid activity line item on your Transactions page. In rare cases Google deducts activity it has already paid for, from future earnings. The deductions FAQ says deductions are final and cannot be appealed, Google does not share the dates, channels or domains involved, and most deductions relate to traffic from the past 60 to 90 days.

3. Ad serving limits. Google may limit the number of ads your AdSense account can show, either while it assesses traffic quality or because of invalid traffic concerns. It says a limit typically lasts less than 30 days, though it can take longer. Your account stays accessible and the details appear in your Policy center. Ad Manager has an equivalent limit that applies to Google demand and third-party bidding, not to direct-sold campaigns, house ads or third-party waterfall sources.

4. Account action. With high levels of invalid traffic, Google may suspend or close an AdSense account, or suspend or disable an Ad Manager account.

Causes you control

• Buying traffic from low-quality sources. Google reports problems for publishers who partnered with low-quality ad networks, search engines or directory sites. Its traffic provider checklist suggests starting with a small test, asking where your promotion will appear, and being cautious if prices seem too good to be true.
• Placements that invite accidental clicks. Keep ads clearly separate from menus, navigation and download links. On phones, an ad right beside a button is easy to tap by mistake.
• Asking for clicks. Any wording that encourages readers to click ads counts, even a friendly "support us by clicking".
• Clicking your own ads, even to see where one leads. To check the ads on your site, use the AdSense Ad review center instead.
• Implementation errors. Google advises checking that custom ad code has no programming errors and works as expected in different browsers.
• Bots reaching your pages. Google cannot block IP addresses for you, but you or your site administrator can block suspicious addresses at your site.

Warning signs in your own data

Google does not explain how it detects invalid traffic, but you can notice patterns in your analytics and ad reports. Treat each as a reason to investigate, not as proof. A viral article can produce similar patterns with valid visitors.

Google Analytics automatically excludes known bots and spiders, so the most obvious bots may not appear there at all. Analytics and ad platforms also count differently (see why AdSense and GA4 revenue don't match), so compare each source with its own normal pattern.

If you are explaining a revenue fall rather than a spike, start with why did my page RPM drop?, which covers the more common causes.

What to do

1. Segment your traffic. Use URL channels, custom channels and ad units to segment AdSense traffic, and combine this with analytics data on where visitors come from.
2. Pause suspicious paid sources. Google recommends stopping or pausing a source that shows suspicious activity. Pause first, investigate afterwards.
3. Review placements on a phone, as a reader would. Move units that sit too close to navigation or buttons, and favor placements readers actually see (see ad viewability).
4. Tell the people around you. Staff, family and friends should not click your ads, even to help. Viewing your own pages is fine, but Google asks publishers not to reload them excessively.
5. Block clear abuse at the server, firewall or hosting level if logs show one address or range hitting pages unnaturally.
6. Report it. AdSense publishers can report suspicious activity through the invalid clicks contact form, linked from this help page (sign-in required).
7. Keep records of dates, sources, what you paused and what you changed. If you ever appeal, Google asks for as much information as you can provide and the changes you have made.

What not to do

• Do not click ads to test, investigate or "fight back" against a suspected bot or competitor. It only adds invalid clicks from your side.
• Do not buy cheap traffic you cannot verify. If a provider cannot show where your promotion runs or explain its low price, walk away.
• Do not use your own AdSense for Search box for your searches. Google advises against it because of the risk of accidental clicks.

This guide summarizes Google's published documentation. It is not legal advice, and no step here guarantees a particular outcome in a review or appeal.