Editorial policy & review methodology
This page is the standing answer to the question "how do you score these tools?". We publish our scoring rubric, our conflict-of-interest rules and our update cadence so that readers and vendors can hold us to a stable standard.
1. Scoring rubric
Every click fraud or bot-traffic protection product is scored across six dimensions. Each is scored 0–5; the total out of 30 is converted to a star rating shown at the top of the review.
| Dimension | What we look at | Weight |
|---|---|---|
| Detection coverage | Signal types (IP, device, behaviour, headless, residential proxy), ad platforms supported | 0–5 |
| Blocking action | Time-to-block, IP exclusion sync to ad platforms, false-positive control | 0–5 |
| Reporting | Granularity, raw data export, integration with GA4/looker | 0–5 |
| Pricing transparency | Public list price, billing terms, refund/cancellation friction | 0–5 |
| Setup & UX | Tag install, time-to-first-block, in-product clarity | 0–5 |
| Trust signals | Independent reviews, audit history, support responsiveness | 0–5 |
2. Source hierarchy
Where vendor marketing claims conflict with primary sources, we follow primary sources and footnote the conflict. The hierarchy is:
- Official platform documentation — Google Ads, Microsoft Ads, Meta Business, GA4 help.
- The vendor's own pricing page, terms of service and product changelog.
- Independent review platforms — G2, Capterra, Software Advice — read for both ratings and recent complaint patterns.
- Hands-on testing inside the product, when vendor access is available.
- Vendor marketing and blog content — used as background only, not as fact.
3. Conflict of interest
- We earn affiliate commissions on outbound links to several click-fraud vendors. This is disclosed on every page that contains such a link.
- Commission rates do not affect ranking position. We rank by total score, not by payout.
- We do not accept paid placement, sponsored review content, or paid link insertion.
- We do not allow vendors to preview or edit a review before publication.
4. Update cadence
- Reviews are re-checked every 90 days against the vendor's current pricing page and changelog.
- Pricing changes greater than 10% trigger an out-of-cycle revision.
- If we cannot verify a claim within 30 days, we mark it "unverified" in the review and lower the trust score.
5. Corrections and right of reply
Vendors and readers may submit corrections at our contact page. We do not remove unfavourable content, but we do update factual errors, replace outdated screenshots and add the vendor's response when supplied. Every revision is recorded in the page's changelog.
6. AI assistance
We use language models for drafting, outlines and copy-editing. Every published page is reviewed and fact-checked by a human editor before publication. See our AI content policy for the full process.