Google Ads targeting tactic slashes invalid clicks by 50%

Advertisers are projected to lose a staggering $172 billion annually to ad fraud by 2028, a threat that hits hardest in hyper-competitive sectors with sky-high cost-per-click (CPC) rates. One of our clients operated squarely in that danger zone, where rampant invalid click activity was destroying campaign performance. By deploying a novel Google Ads targeting tactic, we slashed invalid clicks by 50% and restored profitability.

Our client offered book editing and ghostwriting services. The search terms triggering their ads were both relevant and high-intent, yet the traffic refused to convert at a viable rate. The warning signs of click fraud were unmistakable:

• Google reported a staggering 60% to 80% invalid click rate.We tested third-party click fraud tools, but they delivered no measurable improvement. Next, we filed an investigation with Google. The company acknowledged suspicious activity but claimed it had already caught and refunded all invalid clicks. We were not convinced.Taking matters into our own hands, we added 540 Google-defined audiences set to “Targeting” within our Search campaigns. The result was immediate: the invalid click rate dropped by half, and the conversion rate climbed to profitable levels.Here is why we tested this approach and why it worked.First, a quick refresher: invalid clicks are clicks Google deems not driven by genuine user interest , including intentional fraud, accidental double-taps, or duplicate clicks. Google does not charge for these, and it issues credits for any initially charged clicks later flagged as invalid.But Google’s detection system, while robust, is not flawless. That gap has spawned an industry of third-party tools designed to block fraud-prone IP addresses. Unfortunately, fraudsters cycle through IP addresses using VPNs, staying ahead of monitoring software. A tool may block one IP, but the fraudster simply moves to another. Google also limits advertisers to a maximum of 500 IP address exclusions per campaign, a constraint these tools cannot bypass.Our insight: what distinguishes fraudulent traffic from legitimate users? We focused on Google’s predefined audiences , hundreds of segments based on demographics, search behavior, and browsing patterns. If someone researches private jets and Rolex watches, Google may classify them as a luxury shopper. We hypothesized that fraudsters, cycling through IP addresses, rarely invest time in building normal-looking online profiles that fit these audience signals.So we added most available audiences to our Search campaigns, not limiting ourselves to those directly related to our target. We used the audiences as a filter. Crucially, we selected the “Targeting” setting, not “Observation.” With Targeting, Google restricts ads to users who trigger your keywords and belong to the selected audiences. Observation merely reports engagement without limiting delivery.This tactic is only recommended for accounts with high invalid click rates. Potential downsides include blocking legitimate users who fall outside Google’s predefined audiences.To test this in your own account, navigate to a Search campaign, select Audiences > Edit audience segments > Targeting > Browse, choose your audiences, and save.Common questions about fighting click fraud:
• Will Google refund clicks it identifies as invalid? If identified at the time of the click, you are not charged. If identified later, you receive a credit toward future advertising.By using Google’s predefined audiences as a filter, we cut this account’s reported invalid click rate in half, blocking activity Google claimed it was already catching. The result: failing campaigns turned profitable.