All features
Included on all plans

Fraud detection that
runs itself.

Every click scored against open-source threat signals and rule-based checks. Per-brand enforcement modes (monitor, warn, or block) plus a configurable rules engine so you decide what counts.

app.trcker.io/fraud
Scored
48,210
last 24h
Flagged
612
1.3% of clicks
Blocked
184
before payout

Flagged events

Rule-based, scored on capture.

94
Top fraud score
Held for review before payout
Critical
Datacenter IP + bot signature
clk_8f2a · Frankfurt, DE
94
/ 100
High
Click velocity over threshold
clk_3d71 · Mumbai, IN
78
/ 100
High
Low IP reputation score
clk_a45e · Lagos, NG
71
/ 100
Medium
Duplicate conversion
cnv_b09c · Austin, US
56
/ 100
Bot · datacenter · velocity · IP · duplicate detection

How it works

1

Bot filtering

Known bot signatures, headless-browser detection (Puppeteer, Playwright, Selenium), and user-agent analysis block non-human traffic before it reaches your reports.

2

Rule-based scoring (0-100)

Every click gets a score from deterministic rules: IP reputation against open-source threat feeds, datacenter detection, velocity, geography, and device fingerprint. No ML black box; every score is explainable.

3

BotD interstitial

An optional in-redirect challenge catches sophisticated headless browsers in real time using FingerprintJS BotD. Adds ~80ms to the redirect; only fires on suspicious traffic.

4

Per-brand enforcement

Pick the mode that fits your risk tolerance: monitor (score everything, block nothing), warn (flag for review), or block (auto-reject above threshold). Configure per-offer thresholds anywhere from 0 to 100.

5

Rules engine + dashboard

Build custom rules in a visual editor combining IP range, velocity, geo mismatch, fingerprint recycling, and more. A dedicated fraud dashboard shows partner-level risk, signal breakdowns, and trends.

What gets caught

Datacenter IPs and hosting providers (built-in CIDR list, updated regularly)
Headless browsers: Puppeteer, Playwright, Selenium, and unbranded automation
Velocity spikes: bursts that deviate from a partner's baseline
Recycled device fingerprints across supposedly different visitors
Geographic impossibilities (simultaneous clicks from distant countries)
Sub-ID pattern anomalies that signal click-stuffing campaigns
Conversion-to-click ratio outliers per partner
Duplicate conversions on the same order ID

No add-ons. No per-click fees.

Some platforms charge $200+/mo for a fraud add-on or gate it behind enterprise plans. Trcker includes every layer (scoring, BotD, the rules engine, and the dashboard) on every plan.

View pricing

Frequently asked questions

What happens when a click is flagged as fraud?+

It depends on your brand's enforcement mode. In monitor mode, every click is scored but nothing is blocked, which is useful for tuning thresholds. In warn mode, suspicious clicks are flagged for review. In block mode, clicks above your threshold (default 90) are auto-rejected and excluded from attribution. Rejected clicks still appear in your fraud dashboard so you can audit them.

Can I adjust the fraud score threshold?+

Yes. Per-offer thresholds run from 0 to 100. Lower the threshold to be more aggressive, raise it to let more borderline traffic through. Pair with the rules engine to add brand-specific signals.

Is this machine-learning fraud detection?+

No, by design. The scoring engine is rule-based and uses open-source threat signals (datacenter CIDR lists, IP reputation, headless-browser fingerprints). Every flagged click comes with a human-readable explanation of which rules fired. No black box, no model drift, no surprise blocks. Ezra, Trcker's AI affiliate manager, can read those explanations and recommend partner-level actions.

Does fraud detection cost extra?+

No. Every layer is included on every plan. No add-ons, no per-click fees, no Enterprise gate.

Can partners see their own fraud scores?+

Partners see their rejection rate in the partner portal but not individual click scores. This prevents bad actors from gaming the system.

Related features

Stop paying for fake traffic

Fraud detection runs on every click from day one. Free during early access.

Request Access