Plenty of clicks, almost no leads, and the thought comes up: someone is clicking your ads on purpose, bots or competitors. Google does not charge you for part of such clicks, and for some it credits the money back after you have paid. Below we look at where to see this in your account, when to write to Google yourself and what Google does not promise.
We checked the menu names and the rules against the Google help pages on October 15, 2026. The labels in your account may differ slightly.
What Google counts as an invalid click
According to Google’s help, invalid clicks are clicks without genuine user interest. There are three kinds: manual clicks meant to raise your costs; bots and automated clicking tools; accidental clicks, such as a second click on a double tap.
You are not charged for invalid clicks and impressions: the system removes them from your reports and your bill on its own. How many clicks it filters out, Google does not publish. There is no such figure in the help.
Step 1. The “Invalid clicks” column
Open “Campaigns”, click the Columns icon above the table, choose “Modify columns”, add “Invalid clicks” and click “Apply”. These are clicks Google filtered out automatically, and you were not charged for them.
Look at each campaign separately: if one has a noticeably higher share of such clicks, there is something to look into. A side effect from the help: in rare cases the system removes a click but keeps the conversion that followed it, so a report may show more conversions than clicks.
Step 2. What was credited after payment
Google catches most invalid clicks before the billing period ends, and they never reach your bill. If they are found later, Google credits the money in the next billing cycle, but only “where appropriate and possible”: there is no guarantee.
Where to look: the Billing icon → “Summary” → open the month. Look for the line “Invalid activity (or Invalid clicks)”. The same appears on the “Transactions” page.
Step 3. The Invalid Activity Credit Report
How much was credited for each campaign is shown by a separate report: the “Campaigns” menu → “Report Editor” → “Template gallery” → “Invalid Activity Credit Report: Search & PMax”. It adds the columns “Credited clicks”, “Credited interactions” and “Credited amount”. The report covers only Search and Performance Max campaigns.
There is a conflict over the date. A trade publication wrote that Google documented this report in its help on June 1, 2026, while one service claims the report appeared back in April. Whether the report is new or only newly documented is unclear from the sources. Practitioners also add a caveat: the report shows what Google credited, not what invalid traffic cost you.
When to write to Google yourself
Practitioners name these warning signs: spend jumped sharply in a day for no reason; clicks from cities where you do not advertise; the same keywords bring many clicks and not a single call. This is a reason to check, not proof.
If the suspicion remains, there is the Click Quality Form. What to know:
- Google reviews only the last 60 days. Do not delay.
- A review usually takes several business days, and the reply comes by email.
- Prepare your account number, exact dates, campaign and ad group names, suspicious keywords and your web server logs with IP addresses and click IDs (GCLID).
- Google does not say which exact clicks were judged invalid: that protects its methods. You get a conclusion, not a click-by-click breakdown.
Practitioners add that the form may open with a 404 error if you sign in with an email that has no access to the account. The help says nothing about this.
Not every click without a lead is a bot
More often the cause is simpler: irrelevant search terms, a weak page, a form that does not submit. For a first look, read the article on three checks of Google Ads.
Two traps that look like click fraud:
Clicks are not sessions. Google’s help on the clicks and sessions discrepancy lists many causes: they are different metrics, Google Ads filters out invalid clicks while Analytics shows all data, auto-tagging may be off, the tracking code may fail to load. Google gives no acceptable percentage for the gap and only says some minor discrepancy is expected. Practitioners have their own thresholds in percent, and they differ from author to author, so we give no figures.
A one-day spend spike is not always bots. For most campaigns the daily spending limit equals two average daily budgets, which is normal.
Junk leads: what is on your side
Invalid traffic filters work mostly after an ad is shown, so they do not stop fake form submissions by themselves. In Google’s help on invalid leads (“Prevent invalid leads”) lead quality is called a shared responsibility. Google advises adding reCAPTCHA (an “I am not a robot” test) to the form, double opt-in confirmation by a link in an email, and server-side validation of the data before it goes to your CRM (a customer records system).
Practitioners add a hidden field that a bot fills in and a human cannot see, and SMS confirmation. The sources give no figures on effectiveness, and a form that is too complicated can scare off real customers. Whether Google removes such leads from the data an automated bidding strategy learns from, the help does not explain.
What to do yourself: open the month’s list of leads and check whether emails, phone numbers and texts repeat. Protection for the form should be added by whoever maintains your site.
IP addresses: block or not
The help describes IP exclusions: up to 500 addresses per campaign, and an asterisk in place of the last three digits blocks a range. For the whole account: “Admin” → “Account settings” → “IP exclusions”. For a campaign: “Campaigns” → “Settings” → “Additional settings”. For Performance Max campaigns only the account level is available.
Opinions differ here. In the help Google advises excluding addresses that generate suspicious clicks, yet in another article it warns that many networks share one public address, so an address alone proves nothing. Some practitioners consider blocking weak protection. One agency showed in 2026, on its own account, how to get around it with two addresses; this is not independently confirmed, and we found no response from Google.
Addresses are taken from your site’s server logs, which are provided by whoever runs your hosting.
Click fraud protection services
Practitioners often install separate services to see more than Google shows. Google warns that such tools can flag as invalid clicks that it has already filtered out and not billed, and that it cannot assess third-party reports because it does not know their criteria. So the “losses” total in such a service is not yet money you overpaid. Compare it with the “Invalid clicks” column and the line in “Billing”.
We do not give figures on the share of bots published by the services themselves: these are commercial studies, and Google gives no such share.
The order of checks we recommend
- Add the “Invalid clicks” column and open “Billing” → “Summary”.
- See which words people clicked on. Exclude irrelevant queries with negative keywords (words for which your ad is not shown).
- Compare clicks with sessions in Analytics and check that the form submits.
- Scan your leads for junk.
- Only now collect dates, campaigns and logs and write to Google while the 60 days have not passed.
Summary
Google filters out invalid clicks on its own: you do not pay for some, and it credits others in the next period, if possible. The “Invalid clicks” column, the line in “Billing” and the Invalid Activity Credit Report show how much. If you suspect more, write to Google within 60 days and do not expect a breakdown of every click. But first make sure the cause is not in queries, the page or the form.
How we run search campaigns is in the section on search advertising.
Want a free audit of your Google Ads account? Send us a request: we will look at your account and tell you what is worth fixing.