How to get a Google Ads refund

Google Ads has a refund policy for invalid clicks, but it works in two different ways. Most advertisers only know about one and miss money as a result.

Prefer to watch?

This 4-minute video walks the whole claim end-to-end — spotting the fraudulent traffic burst, exporting click-level evidence, reading the CSV report, and building and submitting your investigation request to Google.

The two refund paths

1. Automatic credits

Google's systems continuously monitor traffic. When they detect clear invalid activity — obvious bots, double clicks, known crawler IPs — they automatically credit your account. You don't do anything. The credit appears in your billing transactions and reduces a future charge.

Google catches automatically: double-clicks within seconds, known crawlers and spiders, blatant bot signatures.

Google misses: sophisticated bots that mimic humans, VPN and proxy traffic, click farms, coordinated competitor sabotage, rotating-IP datacenter traffic.

This is the gap ClickFortify fills — and the gap where a manual claim is worth the effort.

2. Manual investigation (credit request)

For fraud Google didn't catch, you can request a manual investigation. Google's click quality team reviews your evidence and, if confirmed, applies a credit.

The credit goes to your Google Ads account balance — not your bank or card. It reduces your next invoice.

Step 1 — Check what Google already credited you

Before filing, see what you've already received.

Note the amount. When you file a manual claim, request only the gap — Google won't double-credit.

Step 2 — Add invalid-click columns to your campaign view

Google's own invalid-click data shows what it filtered. Adding the columns gives you a clear picture of what slipped through.

Compare Google's invalid count against ClickFortify's bad-clicks count — the difference is the fraud that slipped through.

Step 3 — Identify the fraud period in ClickFortify

• Dashboard → Threat Summary chart: spikes in bad clicks show you the fraud-burst dates.
• Click Traffic → filter Status = Bad to isolate confirmed fraud.
• Note start and end dates.
• Cross-reference with the Google Ads data from Step 2 — do the spikes align?

Submit separate claims per fraud period if needed. There's no limit on how many you can file.

Step 4 — Generate the ClickFortify evidence report

The CSV contains per-click: timestamp, IP, fraud classification, bot/VPN/datacenter flags, country/city, campaign, cost. Summary header shows total bad clicks and wasted spend.

Step 5 — Build the evidence package

Open the CSV in Excel or Google Sheets:

• Filter Status = bad.
• Sum Cost on filtered rows — that's your wasted-spend figure.
• Sort by IP Address — repeat offenders are your strongest signal.
• Note the top 5–10 fraudulent IPs.
• Record the exact date range and affected campaigns.

Cross-reference against:

Step 6 — Submit the manual investigation request

Claim description template

I am requesting a manual invalid click investigation for:

Campaign: [CAMPAIGN NAME]
Period: [START DATE] to [END DATE]
Ad account: [ACCOUNT ID]

Our third-party click fraud protection platform (ClickFortify) detected
[TOTAL BAD CLICKS] invalid clicks during this period that do not appear to
have been filtered by Google's automatic systems, representing approximately
$[WASTED SPEND] in charged ad spend.

Evidence summary:
- [X] clicks classified as bot traffic
- [X] clicks from VPN or proxy sources
- [X] clicks from data-center IPs
- [X] unique IPs responsible for [X] repeat clicks
- Top fraudulent IPs: [IP 1], [IP 2], [IP 3]

I confirmed $[ALREADY CREDITED] has already been credited for this period.
I am requesting review of the remaining $[REMAINING AMOUNT] in unaddressed
invalid activity.

Attached: full click-level CSV export with timestamps, IPs, bot/VPN/datacenter
flags, and cost per click.

Step 7 — What to expect

Google rarely sends proactive updates during investigation. If you haven't heard back in two weeks, contact support with your case reference number.

What helps your chances

• Submitting within 60 days.
• Providing specific IPs, not general descriptions.
• Showing a clear pattern: same IPs repeating, time clusters, one campaign heavily targeted.
• Quantifying the dollar amount and separating it from existing credits.
• Noting that Google's automatic protections didn't catch this.

What hurts your chances

• Traffic older than 60 days.
• Date range with no supporting data.
• Claiming what Google already automatically credited.
• No CSV attachment.
• Vague language ("I think a competitor is clicking my ads") with no data.

Prevention is worth more than refunds

If you're not yet blocking actively, switch Action Mode to Warn or Strict. See Launch protection for the recommended setup.

What happens next

• Make this part of your Monthly review.
• Get to active blocking — Launch protection.