How to Use BotRefund to Organize Evidence for Ad Refund Claims

BotRefund organizes evidence by automatically collecting 110+ independent signals per visit — including click behavior, pointer movement, scroll patterns, browser consistency, and network context — then cross-checking them through an AI model that reaches 99% confidence before packaging everything into a report format that Google and Meta review teams use to evaluate invalid-traffic claims.

To use it, you add the BotRefund script to your landing pages, let it run during your ad campaigns, then download the audit-ready report that ties each flagged session to its click ID (GCLID or fbclid), campaign, placement, timestamp, and the specific behavioral anomalies detected. The report is structured so platform reviewers can verify the evidence without translating raw logs.

What BotRefund evidence organization actually does

BotRefund sits on your website and observes every visitor session that arrives from paid clicks. It does not rely on IP lists or server logs alone. Instead, it runs 106+ client-side checks — such as scrollbar width consistency, clean context iframe behavior, ghost click detection, honeypot trap interactions, robotic mouse movements, superhuman input speed, grid-aligned paths, and absence of humanlike tremor — to build a per-session evidence record.

Each signal is kept as independent evidence, not a verdict. The system cross-checks signals across browser, network, device, and behavior layers, then feeds the complete pattern into a prediction model that classifies the visit as bot or human with 99% accuracy. The output is a session-by-session explanation that includes the click ID, campaign metadata, timestamps, and a signal-by-signal breakdown.

Prerequisites before you start

• Active Google Ads or Meta Ads campaigns sending traffic to pages you control.
• Ability to add a JavaScript snippet to your landing pages or tag manager.
• Access to your ad account click IDs (GCLID for Google, fbclid for Meta) for the periods you want audited.
• A clear goal: either a refund claim, a suppression list for conversion signals, or both.

Step-by-step process to organize evidence with BotRefund

1. Install the tracking script. Add the BotRefund snippet to every landing page that receives paid traffic. The script loads asynchronously and begins capturing behavioral, browser, hardware, network, and attribution signals immediately.
2. Run campaigns normally. Let the script collect data across your active campaigns. It preserves attribution data (campaign, ad set, creative, placement, click identifier) before any changes are made, which is critical for refund claims.
3. Review the audit dashboard. After sufficient traffic volume, open the BotRefund dashboard. You will see flagged sessions grouped by campaign, placement, device, and signal clusters. Each session shows the click ID, timestamp, and the specific anomalies detected (e.g., ghost clicks, honeypot triggers, superhuman speed).
4. Export the refund-ready report. Select the date range and campaigns you want to claim. The export produces a structured document containing: click IDs, campaign details, timestamps, session recordings or replays, and signal-by-signal reasoning for each flagged visit. This format matches what Google and Meta reviewers expect.
5. File the claim with the platform. Submit the report through Google Ads invalid activity credit request or Meta's invalid traffic appeal process. BotRefund's team can assist with claim formatting and negotiation based on experience from 2,500+ audits.
6. Suppress conversion signals (optional). While the claim is pending, you can use BotRefund's conversion protection to stop flagged bot events from feeding back into Google or Meta bidding algorithms, preventing pixel poisoning.

Key evidence types BotRefund captures and organizes

The platform groups evidence into categories that platform reviewers recognize:

• Click behavior: Ghost clicks (activity without human intent sequence), honeypot trap interactions (bots hitting hidden elements), and duplicate click signatures.
• Pointer behavior: Robotic linear movements, absence of humanlike tremor, grid-aligned snapping, and superhuman input speed (<1ms).
• Engagement behavior: Absence of scrolling, no field corrections, uniform click paths, and no meaningful time on offer pages.
• Session behavior: Unnatural durations (too short, too long, or too uniform), missing navigation flow, and rendering anomalies like scrollbar width leaks or clean context iframe mismatches.
• Network and device context: Data center IP ranges, VPN signatures, browser automation framework fingerprints, and hardware consistency checks.
• Attribution linkage: Every session is tied to its GCLID or fbclid, campaign, ad set, creative, placement, and timestamp so the evidence maps directly to billed clicks.

How to verify your evidence package is refund-ready

Before submitting, confirm three things:

1. Click ID coverage: Every flagged session in the report includes a valid GCLID or fbclid that matches your ad account billing data for the claim period.
2. Signal corroboration: No session is flagged on a single anomaly. The report should show multiple independent signals converging (e.g., ghost click + honeypot + superhuman speed + grid-aligned movement).
3. Format compliance: The export uses the structure Google and Meta reviewers use — click ID tables, campaign metadata, session timelines, and signal explanations — not raw JSON or security logs.

If any flagged session lacks a click ID or relies on only one signal, remove it from the claim package. Platform reviewers reject claims built on incomplete attribution or single-rule triggers.

Common mistakes that weaken evidence

Limitations and when this approach does not apply

• Organic or direct traffic: BotRefund only organizes evidence for sessions that carry a paid click ID. It cannot build refund cases for non-paid channels.
• Platforms without invalid-traffic credit programs: The report format is tailored to Google Ads and Meta Ads. Other ad platforms may not accept the same evidence structure.
• Historical claims beyond platform lookback windows: Google and Meta limit how far back you can claim credits. Evidence older than their window (typically 60-90 days) will not be accepted regardless of quality.
• Sites that cannot run client-side scripts: If your landing pages block third-party JavaScript or use strict CSP policies that prevent behavioral data collection, the evidence layer cannot be built.

Key facts

FAQ

How long does it take to collect enough evidence for a claim?

Depends on traffic volume. Most advertisers see actionable clusters within 7-14 days of installation. High-spend campaigns may produce a claim-ready report in 3-5 days.

Can I use BotRefund evidence for a claim I already filed and lost?

Only if the platform allows re-opening with new evidence. BotRefund's report format is designed for first-time submissions; retrospective claims depend on each platform's appeal policy.

Does BotRefund automatically file the refund request?

No. It prepares the evidence package and can guide the submission. You or your agency files the claim through Google Ads or Meta's support channels.

What if my site uses a strict Content Security Policy?

You must allow the BotRefund script domain in your CSP directives. Without client-side execution, behavioral signals cannot be captured and evidence cannot be organized.

How does this differ from Google's automatic invalid activity credits?

Google's automatic system catches server-level patterns (rapid clicking, known bad IPs). BotRefund adds client-side behavioral proof — mouse movement, scroll behavior, browser consistency — that server logs miss, enabling claims for activity Google's automation did not flag.

Can I use the evidence to build exclusion audiences?

Yes. The placement, audience, and device breakdowns in the report let you create suppression lists in Google Ads and Meta to stop bidding on traffic sources with high bot concentrations.

What happens to the evidence after the claim is resolved?

You retain the full report. It can be reused for future claims, shared with auditors, or referenced when adjusting targeting. BotRefund does not delete your session data unless you request it.