Özgül Yavuz Kaddoura
Updated Date:
Desktop Click Fraud in Q2 2023: Key Findings and What They Mean Today
Key Findings: Desktop Click Fraud in Q2 2023
Pixalate's Click Fraud Report for Desktop Web analyzed clicks on programmatic desktop ads worldwide in the second quarter of 2023. These are its main findings:
| Finding | Data |
| Invalid click rate on global desktop web traffic | 21% |
| Share of invalid clicks from click farms and data centers | 57% |
| Country with the highest rate | Colombia, 45% (followed by Singapore and Peru) |
| Ad size with the highest rate | 970x250, 41% |
| Category with the highest rate | Health, 44% |
These figures come from open programmatic traffic and describe a single quarter. Your own rate depends on your channels, targeting and industry.
What These Findings Mean for Advertisers Today
A 21% invalid click rate means roughly one in five desktop clicks in the report didn't come from a genuine visitor. For advertisers, the findings point to a few practical lessons. Newer data shows the problem has not gone away. In Q1 2025, 21% of US programmatic web traffic and 26% of mobile app traffic was invalid (Pixalate). On Google Ads, invalid traffic averaged 12.02% on Display against 5.21% on Search (Advanced Television, 2026). For our own 2026 findings across 718,590 clicks, see the Click Fraud Study 2026.
First, more than half of invalid clicks came from click farms and data centers. Click farms use real devices and data center traffic can rotate IP addresses quickly, so blocking individual IPs alone isn't enough; detection needs to look at devices and behavior as well.
Second, rates varied widely by country and category. If you advertise in markets or industries with higher fraud rates, such as health, it's worth checking your location targeting and reviewing your traffic more often.
Finally, invalid clicks are paid for from your budget and distort the data your bidding relies on. Combining careful campaign settings with automated click fraud protection helps keep that budget focused on real customers.
Signs of Click Fraud in Your Campaigns
Early detection makes click fraud much easier to contain. Watch for an unusually high click-through rate (CTR) without matching conversions. In the retail sector, a common sign is visitors who click high-value product ads but never add anything to their cart. Also look out for more clicks from locations you don't target, consistently low conversion rates despite high click volumes and unusual click timing, such as clicks concentrated at specific hours.
Technical Clues to Watch
Pay close attention to suspicious user agents or IP addresses that show automated or repetitive patterns. More organized fraudsters use click farm software to manage many devices at once, making their fake traffic look human. Keep track of click farm and bot patterns you've identified before, and watch for sudden increases in clicks from those sources.
How Click Farms Operate
Click farms use banks of real smartphones and computers to click on ads, which helps them get past simple filters. According to Pixalate, click farms and data centers together accounted for 57% of invalid desktop clicks in Q2 2023.
When this traffic targets competitive keywords, it can use up a daily budget quickly. Catching it takes more than an IP list: it means combining signals such as device recognition, click patterns and on-site behavior.
Other Warning Signs
Other warning signs include your daily budget running out quickly without matching conversions, and unusual behavior on your site, such as high bounce rates and very short visits. Watch for repeated clicks from the same sources, frequent clicks on irrelevant keywords and sudden click surges on low-volume keywords. These patterns often point to bots or competitors trying to use up your daily budget.
Why Ongoing Monitoring Matters
Reviewing these warning signs regularly helps you catch click fraud early. If any of them appear, investigate before they affect more of your budget and your campaign data.
Automated protection makes this easier. ClickSambo checks every click on your Google Ads and Meta Ads campaigns in real time and blocks suspicious sources automatically, so more of your budget reaches the people who actually drive your business.
$84 Billion
Ad fraud cost in 2023 (Juniper Research)
21%
Invalid desktop clicks in Q2 2023 (Pixalate)
57%
Of invalid clicks from click farms and data centers (Pixalate, Q2 2023)
45%
Invalid click rate in Colombia, the highest (Pixalate, Q2 2023)ClickSambo checks every click on your Google Ads and Meta Ads campaigns in real time and blocks suspicious sources automatically.
See How It Works 
Frequently Asked Questions
What share of desktop ad clicks were invalid in Q2 2023?
According to Pixalate's Click Fraud Report for Desktop Web, 21% of clicks on global desktop web traffic were invalid in the second quarter of 2023. More than half of these invalid clicks (57%) came from click farms and data centers.
Where was desktop click fraud highest?
Among countries, Colombia had the highest invalid click rate at 45%, followed by Singapore and Peru. Among categories, ads on health websites had the highest rate at 44%, and the 970x250 ad size had a 41% invalid click rate.
Are these figures still relevant today?
They describe a single quarter of open programmatic desktop traffic, so they aren't a forecast for your account. But the problem hasn't gone away: Juniper Research estimated that ad fraud cost advertisers $84 billion in 2023. Newer data puts the yearly cost of invalid traffic at $63 billion (Advanced Television, 2026). The best way to know your own exposure is to check your campaign data for the warning signs described above.
How can I protect my campaigns from click fraud?
Start with your settings: target only the locations you serve, count only meaningful actions as conversions and review your placements and invalid click rates regularly. Then add automated protection. ClickSambo checks every click on your Google Ads and Meta Ads campaigns in real time using device fingerprinting, VPN and proxy detection and behavioral AI, and blocks suspicious sources automatically.
