Google Ads card now shows invalid traffic savings and credits

Google Ads is rolling out an Invalid Traffic Overview Card that shows advertisers an estimate of the spend Google's protections saved them before billing and how much it credited back afterwards. Google announced the card on 12 November and says the global rollout began this week.

Keerat Sharma, Google's VP and general manager of ads privacy and safety, wrote that the card lets businesses see, for the first time, how much Google saved them and what it credited back automatically for invalid activity. Until now that information was split between campaign reports and monthly billing statements.

Google does not charge for invalid traffic it catches before an invoice is issued, and it says its defences catch the vast majority of it at that stage. Traffic found after a billing cycle closes is credited back, in Google's words, "where appropriate and possible".

Estimated savings is the card's new figure

Open the Campaigns menu, go to the Overview page and look for the "Invalid traffic" card. Google's help page says the card is rolling out gradually, so it may not be in every Australian account yet.

The card reports two figures for whatever date range you have selected:

• Estimated savings: the cost Google estimates it saved you by filtering bots, click farms and accidental clicks before billing.
• Credited cost: the amount credited back for invalid activity Google found after a monthly billing cycle had closed.

A "View details" button opens a daily trend chart and a campaign-by-campaign table, with links to the invoices where credits were applied. The reports can be downloaded as CSV files.

The estimated savings figure is the new part. Advertisers could already add an Invalid clicks column to the campaigns table, and credits already appear as "Invalid activity" adjustments under Billing. What was missing was a dollar value for traffic that never reached the bill.

Bigger campaigns can show higher estimated savings

The campaign table will tempt people to rank campaigns by estimated savings and treat the top one as the problem account. A campaign that spends more generates more traffic, so it will usually show higher estimated savings than a smaller campaign with the same share of invalid clicks. Read the figure alongside each campaign's spend.

Google's help page adds context on the Invalid clicks column. Invalid traffic fluctuates, so the numbers move around on their own, and a rise means the filters caught more interactions and removed them from your metrics and your bill.

The card also only counts what Google's systems identified. It gives no figure for invalid traffic that went undetected, so it cannot be read as a measure of how much fraud is still reaching the account.

Campaign cost figures ignore credits issued after billing

Google freezes campaign metrics at the end of each month so they match the invoice. Invalid traffic caught before that point is stripped from clicks, cost and conversions. Anything found later is credited on a future invoice, and the Cost and CPA figures in the main Campaigns table stay as they were.

That is why an account can show thousands of filtered invalid clicks in its campaign reports and only a small credit on its bill. Google's help page says the two numbers are not directly related: the column counts pre-billing filtering, while the credit covers detection after billing.

To see performance with credits factored in, open Report editor, go to the Template gallery and choose the Invalid Activity Credit: Search & PMax report. It carries 10 columns, including Adjusted cost, Adjusted cost per conversion and Credited clicks. Google notes that credits for clicks near the end of a month can land in the following month, so expect small timing mismatches. As the name says, it covers Search and Performance Max.

The Invalid traffic card does not measure lead quality

Google's filters run after an ad serves, so they can stop you paying for a bot click, but they cannot stop that visitor landing on your site or submitting your form. Real people also submit fake contact details.

Ginny Marvin, Google's Ads Liaison, made the same point in Google's Ads Decoded newsletter: lead capture happens on your site after the click, where bots can hit forms directly or people may enter fake information for gated content. An account can show very little invalid traffic in Google Ads and still be drowning in junk enquiries.

Google's help page calls lead quality a shared effort and lists what it expects advertisers to do on their side:

• Use the Leads objective with goals such as qualified lead, converted lead, book appointment or request quote.
• Protect forms with reCAPTCHA and server-side validation, and consider double opt-in by email.
• Install the Google tag and use enhanced conversions so bidding learns from real customer actions.

Google's blog post says its system upgrades have cut advertiser exposure to invalid leads by an average of 30% so far in 2026. That is a platform-wide average from Google, and the card does not report a lead-level figure for your account. The lead scoring on custom questions that Google added to lead form assets in October is a closer fit for that job, alongside matching each lead back to the CRM.

Column names and real figures can drift apart on other platforms too. In one Meta Ads account, a column with leads in its name (offsite_content_view_add_meta_leads) reported 260,849 against 2,003 actual leads, which is the same lesson from the other direction.

Investigation requests still need server logs and GCLIDs

The card does not replace the manual route. If you think invalid traffic slipped past the filters, you can file an invalid traffic investigation request through the Click Quality Form, covering activity in the last 60 days. Google says specialists review the data manually and typically take several business days, after which the account is emailed the findings and told of any credits.

Before you file, Google asks for the 10-digit customer ID, exact dates, the campaign and ad group names, suspicious keywords, and web server logs showing repeating IP addresses, devices and GCLIDs. For lead problems it wants a spreadsheet with a GCLID for each lead and your assessment of it, such as the outcome of a call or email.

One limit is set in advance: the investigation will not confirm whether specific clicks were marked valid or invalid. Google says revealing that would expose how it detects invalid traffic.

Google's help page also warns that a traffic spike or a budget that empties quickly does not always mean invalid traffic. Seasonal demand, a competitor pausing ads, or a change to budget, keywords, audiences or locations can all do it, and the Change history is the first place to check. Clicks that look automated have been showing up on the organic side too: the Search Console click spikes seen on some sites in early November looked like bot traffic while impressions stayed flat.

Add the Invalid Activity Credit report to monthly reviews

For an Australian advertiser the card changes what a monthly review should include:

• Open the Overview page, note estimated savings and credited cost for the month, and open View details to see which campaigns carry the most filtered traffic. Read those figures against each campaign's spend.
• Run the Invalid Activity Credit: Search & PMax report and use Adjusted cost per conversion in any CPA discussion, since the Campaigns table will never include post-billing credits.
• Download the campaign CSV if a client or manager asks whether click fraud is eating the budget. It shows what Google caught and nothing about what Google missed.
• If you run lead campaigns, keep scoring leads in the CRM and keep the form protections in place. A clean invalid traffic card is not evidence that the leads are clean.

If the card has not reached your account yet, Google says the rollout is gradual, so the Invalid clicks column and the Billing summary remain the places to look in the meantime.