Google filters most invalid activity before you pay. Here are the definitions, the reporting columns, the credit process and what Google will and will not tell you.
Google Ads invalid clicks are clicks that Google judges not to reflect genuine interest in an advertiser’s business, including accidental clicks, clicks generated by automated software and deliberate manual clicking. Google filters most of them before they are billed, reports the filtered volume in dedicated columns, and credits invalid activity that it discovers after an invoice has already been issued.
This guide explains Google’s definition and examples, how Google says it detects invalid traffic, where the invalid click data sits in your account, and how credits show up in billing. It also covers what Google will and will not disclose, when an investigation request makes sense, and how Google’s categories relate to the industry distinction between general and sophisticated invalid traffic. A short FAQ closes the article.
Google’s glossary defines invalid clicks as “clicks on your ads that don’t represent a genuine interest in your business” and groups them into three kinds: accidental clicks caused by poor ad placement, clicks from automated tools, bots or other deceptive software, and manual clicks that increase your ad costs or inflate publisher profits (Invalid clicks: Definition).
Google’s longer help page uses the broader term invalid traffic, which covers impressions as well as clicks, and gives four example categories (About invalid traffic):
Google notes that the list is not exhaustive. The important point is that “invalid” describes the traffic, not the motive. An accidental tap on a mobile ad is invalid but not malicious. Deliberate abuse is a narrower subset, covered in our guide to deliberate click fraud on Google Ads.
According to Google, its systems use automated filters, machine learning and manual reviews, and a specialized Ad Traffic Quality team conducts manual investigations. Google says the team “often detects this traffic as it happens or shortly after”, and that it also analyzes traffic over longer time periods to find more complex patterns (About invalid traffic).
Google’s Ad Traffic Quality site adds detail on timing. It says its filters “stop the vast majority of invalid traffic in real time or soon after”, but that “in some cases, it may take up to several weeks to recognize a suspicious pattern.” It also says that “because no filter is perfect”, staff manually review issues flagged by advertisers, publishers and automated systems (How we prevent it).
Those three layers explain why invalid clicks can surface at different points in your reports and bills.
| Detection layer | When it acts | What you see |
|---|---|---|
| Automated filtering | As the click happens or shortly after | The click never enters Clicks or Cost; it is counted in the invalid click columns |
| Longer-term analysis | Days to weeks later | Removed before billing if found in time, otherwise credited on a later invoice |
| Manual review | When activity is flagged by systems, publishers or advertisers | Can lead to credits; investigation findings arrive by email |
Google does not publish its detection criteria. It explains that investigation findings will not confirm whether specific clicks were marked valid or invalid, and that it keeps this information confidential so that bad actors cannot learn to work around its defenses (About invalid traffic).
Google describes the Invalid clicks and Invalid click rate columns this way: they “indicate the number and percentage of clicks that were classified as invalid and automatically filtered from your account. You aren’t charged for these clicks, and they don’t affect your account statistics.” The same page notes that the column is available on the Campaigns page only (About columns in your statistics table). To add it, go to Campaigns, select the Columns icon above the table, search for “Invalid clicks” and apply the change (About invalid traffic).
The calculation matters when you read the rate. Google’s API reference defines the underlying metric as “the percentage of clicks filtered out of your total number of clicks (filtered + non-filtered clicks) during the reporting period”, and lists it as available at the campaign and account (customer) level (metrics, Google Ads API). The denominator includes the filtered clicks themselves.
As an illustration, imagine a campaign that shows 900 clicks in the Clicks column and 100 in the Invalid clicks column. Total clicks are 1,000, so the invalid click rate is 10 percent, not 100 divided by 900. You pay only for the 900, and Cost, CTR and average CPC are all calculated from them.
Because the data is campaign-level, you cannot tie invalid clicks to a specific keyword, ad group or search term. Use it as a campaign health signal and compare it with the same campaign’s history rather than with other accounts.
For formats whose main action is not a click, Google refers to an Invalid interactions column alongside Invalid clicks, explaining that these columns “show traffic Google filters before you were billed” (About invalid traffic). If you run video or other interaction-based campaigns, check both.
Google also summarizes protection on the Overview page within the Campaigns menu. The Invalid traffic card shows estimated savings and credited cost, and selecting View details opens a more granular breakdown (same source). It is the quickest way to see filtering and post-billing credits side by side.
For advertisers who report through the Google Ads API, Google also exposes general invalid clicks, described as “a subset of your invalid clicks that are detected through routine means of filtration (such as known invalid data-center traffic, bots and spiders or other crawlers, irregular patterns, etc.)”. As with the main metric, you are not charged for these clicks and they do not affect account statistics (metrics, Google Ads API). The wording echoes the industry’s definition of general invalid traffic, discussed below.
Timing determines whether invalid activity is filtered or credited. Google states that invalid traffic caught before the end of the billing cycle is automatically removed from billing and campaign reports, so no refund is needed. Invalid traffic detected after billing results in credits “in the form of a billing adjustment for the value of the invalid activity”. These appear as “Invalid activity” under the Adjustments drop-down in the Month card of the Billing summary, and campaign metrics are not retroactively changed when a credit is issued (About invalid traffic). Google adds that you may see many filtered invalid clicks in campaign reports but small credit amounts in billing, and that this is normal because the two numbers are not directly related: the columns show what was filtered before billing, while credits cover what was found afterward.
The Billing activity page, reached from the Billing icon, lists invalid traffic credits as adjustments and adds “an additional line stating the original month of service in which the invalid traffic occurred”. You may see more than one line if credits relate to different months, and View more breaks the credits down by campaign (About the Billing activity page).
For campaign-level detail, Google offers the Invalid Activity Credit Report, available for Search and Performance Max campaigns. It sits in Report editor, under the Template gallery, as “Invalid Activity Credit Report: Search & PMax”, and breaks credits down by campaign and network (About the Invalid Activity Credit Report). Its columns include:
Google notes two quirks on the same page. A click is attributed to the month in which it occurred, while its credit is applied to a later month, which can cause “a slight timing mismatch between interaction reporting and credit issuance”. Rounding of monthly totals can also produce minor discrepancies.
| When the activity is identified | Campaign reports | Billing |
|---|---|---|
| Before the billing cycle closes | Excluded from Clicks and Cost; shown in invalid click columns | Never charged |
| After the invoice is issued | Original metrics stay as reported; the credit report shows adjusted figures | Credited as an “Invalid activity” adjustment, with the original month of service |
Knowing the boundaries saves time when you suspect a problem.
Google provides:
Google does not provide:
The safeguards that remain in your hands, from exclusions to form protection, are covered in our guide to practical invalid click protection.
Request an investigation only after you have checked Google’s own invalid click data and ruled out ordinary explanations such as setting changes, seasonality, new competitors or location options. A structured way to do that is set out in our method for building an evidence-based click fraud case, and our guide to reading suspicious ad sessions helps you avoid common false positives.
Google accepts investigation requests through the Click Quality Form for activity within the last 60 days (About invalid traffic). Before you submit, gather the information Google asks for on the same page:
Google says the review “typically takes several business days” and that you will receive an email with the findings. Do not expect a click-by-click verdict. If credits are due, they appear in billing as described above; Google’s credit documentation says credits are issued only “where appropriate and possible” (About the Invalid Activity Credit Report).
The Media Rating Council, an industry body whose accreditation audits cover invalid traffic filtering, sets out two categories of invalid traffic in the June 2020 update of its Invalid Traffic Detection and Filtration Standards Addendum (MRC IVT Standards Addendum):
| MRC category | How it is detected | Google examples that fit |
|---|---|---|
| GIVT | Lists and standardized checks, applied routinely | Known crawlers and bots, data-center traffic; the API’s “general invalid clicks” |
| SIVT | Advanced analytics, corroboration across signals, human review | Clickjacking, ad stacking, paid or organized human clicking, malware-driven automation |
The distinction helps explain the timing described earlier. List-based checks can act almost instantly, while sophisticated patterns may only become visible after analysis over days or weeks, which is when credits rather than filtering come into play. Google states that the MRC accredits its invalid traffic defenses for specific services and metrics (About invalid traffic). For the wider landscape of non-human activity around ads, including benign crawlers, see our map of automated ad traffic.
Not necessarily. The rate measures clicks that were caught and not charged. A rise means more invalid activity was filtered, which can follow anything from a new placement to a burst of crawler activity. It becomes worrying when it coincides with other signals, such as poor engagement on the clicks you did pay for. Compare the rate with the campaign’s own history rather than with figures published by vendors.
Google’s approach is filtering and credits rather than refunds. Activity removed before the billing cycle closes is never charged, and activity found later is credited as a billing adjustment (About invalid traffic).
Filtering happens in Google Ads reporting and billing. The visit itself may still have reached your website and been recorded by your analytics tool, which is one reason clicks and sessions differ. Our guide to how Google Ads counts clicks explains the other reasons.
No. Google says an investigation will not confirm whether specific clicks were marked valid or invalid. You can see aggregate counts, rates and credits, but not a list of individual decisions.
They may be measuring different things. Google bills on its own data and says it cannot evaluate third-party reports without knowing their criteria. Many clicks a tool flags may already have been filtered and never charged. If you request an investigation, send the raw material Google asks for, such as server logs with timestamps, IP addresses and GCLIDs, rather than the tool’s verdict, and do not expect Google to reconcile the two numbers.
Yes. Google notes that reports may occasionally show more conversions than clicks when a click that led to a conversion is later marked invalid and removed; the conversion itself remains valid (Invalid clicks: Definition).
Keep reading
Google Ads
Click fraud is the deliberate part of invalid traffic. Learn who is behind it, how exposure differs by network and how to report a case to Google without overreacting.
Click fraud
No setting makes a campaign immune to bad clicks. See how Google's filtering, your own campaign controls and outside tools fit together, and where each one stops working.
Google Ads
The phrase covers everything from click bots to worried advertisers. Here is how a click is really recorded, filtered and charged, and how to audit your own traffic.
