Trcker's pricing engine uses Bayesian modeling and machine learning to calculate the optimal cost-per-click for every partner and traffic segment — automatically.
The engine takes daily snapshots of earnings-per-click across partners, offers, and traffic segments. Bayesian priors smooth out noisy data from low-volume partners, so you get stable estimates even with limited history.
Before changing any bid, the engine simulates the impact against historical data. Built-in guardrails prevent bids from moving more than 20% in a single adjustment and enforce floor/ceiling constraints.
A LightGBM model scores every incoming click for conversion probability in real time. High-intent traffic gets higher bids. Bot-like patterns get suppressed. The model retrains automatically as data accumulates.
SHAP feature importance and decision tree analysis automatically discover which traffic segments convert best — by device, geo, time of day, referrer, and more. New segments surface without manual analysis.
Other affiliate platforms give you a flat CPC field and call it a day. You set $0.50, hope it works, and manually adjust when it doesn't. Trcker's pricing engine eliminates guesswork — it calculates the optimal bid for every click automatically.
No. The engine runs automatically once you have enough conversion data (typically 50+ conversions). You can set floor and ceiling prices per offer, but the optimization is fully autonomous.
The model uses click-level features: device type, operating system, geographic region, time of day, referrer domain, partner history, and fraud score. It does not use personal data or cookies.
Yes. You can set manual CPC overrides per partner or per offer at any time. The engine will respect your overrides and optimize around them.
The pricing engine is available on Growth and Pro plans. Starter plans use static CPC pricing.
The pricing engine activates automatically on Growth and Pro plans. Free during early access.
Request Access →