On-Site Bot Evidence Generation Process: How to Prove Bot Clicks

The on-site bot evidence generation process is the method by which a website collects and records behavioral and technical signals from each visit, cross-checks them, and produces a report that can be used to prove a click or session was automated. In practice, it involves adding a small script to your site that captures mouse movement, click patterns, session timing, and other signals, then sends them to a detection engine that evaluates them against known bot behaviors.

This evidence is what you need to dispute invalid clicks with Google or Meta and claim a refund. Without it, ad platforms have little reason to believe your traffic was fraudulent.

Why On-Site Bot Evidence Matters

Bot clicks can steal up to 20% of your Google and Meta ad budget. That is a direct loss, but you can recover it if you can prove the clicks were not human. On-site evidence is the proof. It shows exactly why a visit was classified as a bot, with specific signals and timestamps.

Without this evidence, your refund request is just a claim. With it, you have a documented case that ad platforms can review and approve.

The Core Signals Used to Generate Evidence

Bot detection systems look for patterns that real humans rarely produce. The following signals are commonly collected on-site:

• Ghost click detection – Catches click activity that happens without the natural sequence of human intent.
• Honeypot trap interactions – Watches for bots that respond to hidden or intentionally deceptive page elements.
• Robotic linear mouse movements – Flags unnaturally straight pointer paths that rarely appear in real user sessions.
• Absence of humanlike mouse tremor – Looks for the tiny imperfections and jitter typical of human movement.
• Superhuman input speed – Identifies interactions that happen faster than a person could realistically perform.
• Grid-aligned movement patterns – Detects movement that snaps to precise lines or blocks instead of natural curves.
• Absence of clicks or scrolling – Highlights sessions that stay too static to match a real browsing journey.
• Unnatural session durations – Catches visit lengths that are too short, too long, or too uniform to be human.

Each signal is a piece of evidence. A single anomaly is not a bot verdict, but when several signals agree, the case becomes strong.

How Independent Checks Work

Independent checks are the building blocks of reliable detection. BotRefund uses 106 independent checks to build a reliable picture of whether a visit is human or automated. Each check captures a different fact about the visit, from network data to pointer behavior.

No single check is definitive. A VPN can mask location. A privacy browser can block scripts. A touch device may not produce mouse movements. That is why every signal is treated as evidence, not a verdict. The system then cross-checks each signal against independent browser, network, device, and behavior data. It looks for corroboration, not a single tell.

For example, a real visitor on a mobile network might show a slightly unusual port because of carrier settings. But that same visitor would still have coherent timing, click patterns, and scrolling. A bot, on the other hand, often shows multiple mismatches at once: a strange port, a robotic mouse path, superhuman speed, and no natural tremor. The AI prediction model weighs the complete pattern across all signals. Accuracy comes from corroboration, not from one browser tell.

Bot versus human behavior: a real person hesitates, pauses, and moves with slight jitter. They read, then scroll, then click. Bots often act in straight lines and snap to grid coordinates. Their clicks occur in milliseconds, and their session durations look mechanical. When a check catches an anomaly, it is not enough to convict. But when many checks align, the evidence becomes hard to dismiss.

How the Evidence Is Generated Step by Step

1. Add the detection script. You place a small JavaScript snippet on your website. This typically takes about one minute and requires no credit card.
2. Collect behavioral data. The script records mouse movements, clicks, scrolls, session duration, and other interactions in real time.
3. Cross-check signals. The system compares each signal against independent browser, network, device, and behavior data. It looks for corroboration, not a single tell.
4. Run AI prediction. A machine learning model weighs the complete pattern across all signals to classify the visit as bot or human.
5. Generate a report. The system produces a detailed report with timestamps, signal descriptions, and a verdict. This report is your evidence.
6. Export and submit. You export the report and send it to your Google or Meta representative to claim a refund.

Key Facts About BotRefund's Evidence Process

Limitations and When Evidence May Not Be Conclusive

No single signal is a definitive bot verdict. Privacy tools, travel, corporate networks, and unusual devices can produce unexpected behavior for genuine people. That is why the evidence process cross-checks multiple signals and uses AI to weigh the complete pattern.

If a visitor uses a VPN or a privacy browser, some signals may look suspicious even though the person is real. The system accounts for this by keeping each signal as evidence—not a verdict—and looking for corroboration.

Also, the evidence is only as good as the data collected. If your site does not have the script installed, no evidence is generated. And if you wait too long, you may miss the refund window for older clicks.

How to Use the Evidence to Claim Refunds

Once you have a report, the next step is to submit it to the ad platform. BotRefund's process is designed to make this easy: you turn on the free AI audit, export your report, send it to your Google or Meta rep, and claim your refund.

Here is a concrete timeline of the refund submission w:

1. Day 1: Install the script. Start collecting data.
2. Day 2–7: The system runs live. You check the free bot audit to see flagged visits.
3. Day 8: Export your report for the previous week.
4. Day 9: Send the report to your Google or Meta rep with a clear refund request.
5. Day 15–30: Ad platforms review the evidence. You may need to answer follow-up questions.
6. Day 30–60: Refund approval and recovery, depending on the platform.

The report should clearly show which signals were triggered and why the visit was classified as a bot. This gives the platform a concrete reason to approve your claim.

Frequently Asked Questions

How long does it take to generate bot evidence?

Setup takes about one minute. After that, evidence is generated continuously as traffic comes in. You can run a free bot audit to see results immediately.

What if a real user triggers a bot signal?

That is why the process uses cross-checking and AI. A single anomaly is not enough to classify a visit as a bot. The system looks for corroboration of signals before making a verdict.

Can I use this evidence for both Google and Meta refunds?

Yes. BotRefund is designed to prove bot clicks for both Google Ads and Meta ads, and it negotiates with both platforms on your behalf.

Do I need technical skills to install the script?

No. The script is added in about one minute, and no credit card is required for the free audit. The process is designed for non-technical users.

How far back can I claim refunds?

BotRefund can recover bot-click refunds from Google Ads spend dating back to 2017, so you may be able to reclaim older losses.

What does the evidence report look like?

The report includes timestamps, the specific signals triggered, and a clear verdict. It is formatted to be submitted directly to ad platforms.