Key takeaways
When is an invalid traffic review worth requesting, and when not?
When you have a concrete suspicion about clicks that Google did not filter and you have ruled out other explanations. A spike alone is not enough. Google lists causes of traffic increases that have nothing to do with invalid traffic, and the "Invalid clicks" column already covers what Google filtered.
The Google Ads help page on invalid traffic names these reasons for a sudden rise in clicks:
- Seasonal trends or news events: a genuine spike in interest.
- A competitor stops its ads: then your ads may show more often and get more clicks.
- Campaign changes: a higher budget, new keywords, or changed audience and location settings can bring more impressions and interactions.
Google points to your Change history for this. Compare the dates of the spike with the dates of your own edits.
Two patterns in your analytics look worse than they are, says Google. A zero-second visit does not always mean the visitor left immediately, because many analytics tools need a second click to record the duration. And the same IP address on many visits does not always mean one device: with Network Address Translation (NAT), hundreds or thousands of devices can share one public IP address.
The "Invalid clicks" column shows the traffic Google already filtered and did not charge you for. For those clicks you need no request. How the three kinds of invalid clicks differ is in our article on invalid clicks and credit.
How does the 60-day window work?
You can have activity of the last 60 days investigated. Google does not look at anything older. The window moves forward by one day every day, so a conspicuous day drops out of it 60 days later, whether you have finished your evidence or not.
A made-up date example: a spike happened on October 1. Around the end of November that day sits at the edge of the window. The help page says only "activity within the last 60 days" and does not say whether the count starts on the day of the activity or on the day of your request. So plan with a margin and submit well before the edge.
So do not wait for the perfect proof. A request with a clear period and the logs you already have beats a perfect one that arrives after the window closed.
Which details does Google ask for, and where do you get them?
Google asks for your Customer ID, the exact date range, campaign and ad group names, suspicious keywords, web server logs and a short explanation of why the traffic looks suspicious. The right-hand column is our advice, not Google's instruction.
| Detail according to Google | What it means | Where you can get it |
|---|---|---|
| Customer ID | Your 10-digit Google Ads account number. | Copy it from your own Google Ads account. |
| Specific date range | The exact days on which you noticed the activity, no more than 60 days back. | Look at your campaign table by day and note the first and the last conspicuous day. |
| Campaign and ad group names | The campaigns and ad groups where you saw the spike or the suspicious behavior. | Look in your campaign table and compare the days you noted. |
| Suspicious keywords | Keywords that drive the unusual traffic. | Compare your keyword figures for the same days with the days before. |
| Web server logs: IP addresses | IP addresses that produce repeated clicks or clicks without conversions. | Look in your server or shop logs and group the visits by IP address. |
| Web server logs: platforms and devices | Browser or device information from the suspected activity. | Look in the same log lines, which usually carry the browser identifier. |
| Web server logs: GCLIDs | The Google click identifiers of the suspicious clicks. According to Google, they help map a visit back to a specific ad interaction. | Look at the landing page URLs in your logs. With auto-tagging the GCLID is attached to the URL. |
| Trend indicators | A brief explanation of why the traffic is suspicious. | Write two or three sentences from your own comparison. Google's own example: a 300% spike in clicks without more conversions, or clicks from outside your targeted locations. |
According to the English help page, there is one more detail if your problem is invalid leads: a spreadsheet with the GCLID for each lead and your own assessment of each lead, for example the outcome of an email or a call.
Why does your logging decide before you submit the request?
Because Google asks for web server logs with IP addresses and GCLIDs for the period, and you cannot add what was never recorded.
Check two things now, before you suspect anything:
- Does your shop or server log each landing page call with the GCLID in the link? That needs auto-tagging, which is what attaches the GCLID to the URL. Our article on invalid clicks and credit mentions the click ID.
- Are the logs kept long enough to cover the 60 days?
If you only start logging once you have the suspicion, the weeks before it are missing. The trap is described briefly in our article on rising click prices: the logs are the hard part.
One note on data protection: IP addresses are personal data. How long you may store them is a question for your data protection adviser. This is not legal advice from us, and we name no retention period.
How do you submit the request, and what happens afterwards?
You collect the details and send them through the Click Quality Form. Google specialists then evaluate your traffic of the last 60 days and write to you by email. Go through these steps in order:
- Check the Change history. Look for edits to budget, keywords, audience or location around the days in question. If one explains the spike, you may not need a request.
- Narrow down the period. Note the exact days, all of them inside the last 60 days.
- Collect the details from the section above. Customer ID, campaign and ad group names, suspicious keywords, web server logs with IP addresses, platforms and devices and GCLIDs, and your short explanation.
- Fill in the Click Quality Form. Enter what you collected and keep a copy for yourself.
- Wait for the email. Google writes to you with its findings when the investigation ends.
After the request, Google evaluates the traffic of the last 60 days. If it finds additional invalid traffic that the automated filters missed, it notifies you of any credits. Google makes that decision, and a credit is given where appropriate and possible.
What can the request not do?
It confirms no single click and promises no amount. Google says an investigation cannot confirm whether specific clicks were marked as valid or invalid, because sharing that would reveal how detection works and help bad actors avoid it.
Google also does not provide a click-by-click comparison with third-party reports. According to Google, such tools often flag traffic that Google already filtered and did not charge you for. A gap between a tool report and Google's figures is therefore no proof that you paid for the difference.
So do not plan with an amount, and treat the result as an answer, not as a claim. A review is a look back and does not replace measuring your own traffic on an ongoing basis.
How does seacashback help you with this?
seacashback checks the clicks on your landing pages, excludes detected bots at account level and documents invalid clicks (IP address, time, click ID), so that you can report them to Google. You submit the request, and Google decides. The documentation does not prove that Google owes you anything.
The records exist only from the start of the recording, so earlier days stay without evidence. Click protection changes neither the auction price nor your bids. To the extent detected clicks would have been billed, it dampens on paper what a real click costs you in the end. Read how it works under How it works. For data on invalid clicks in shops, see the Click Protection Report 2026. It measures German shops, so treat it as a reference for that market only.
Frequently asked questions
How far back can I request an invalid traffic review?
Google limits investigations to the last 60 days. Older activity is not reviewed, and the window moves forward by one day every day.
Will Google tell me which clicks were invalid after my request?
No. According to Google, an investigation cannot confirm whether specific clicks were marked as valid or invalid. Google keeps this confidential so that bad actors cannot learn how detection works. You get an email with findings when the investigation ends.
Do I get a credit if I submit the request?
No promise. Google evaluates your traffic. If it finds additional invalid traffic that the automated filters missed, it notifies you of any credits. A request is no promise, and Google decides.
Is the report of a click protection tool enough as evidence?
Google names web server logs with IP addresses, platforms and devices, and GCLIDs. A tool report is not on that list, and Google says it cannot provide a click-by-click comparison with third-party reports. Use your own logs as the evidence and the tool report as a pointer to where to look.