RevContent Automated Widget Optimization Script (Python, API)

A script that runs on a schedule, pulls widget-level performance from the RevContent Stats & Management API, and manages widget bids and blocks based on CPA against a target we set. Spec Auth to the RevContent API (client ID/secret). Pull last [7] days of widget stats per boost: clicks, spend, conversions. Compute CPA per widget. Apply rules from a config file: CPA below target × 0.8 with ≥ [N] conversions → raise bid [15]% CPA above target × 1.2 with ≥ [N] conversions → lower bid [20]% CPA above target × 2.0 with ≥ [N] conversions → block widget ≥ [N] clicks, 0 conversions → block widget Under minimum sample → no change Push bid changes and widget blocks via the API. Guardrails: bid floor and ceiling, max one action per widget per run, max [X] blocks per run, dry-run flag for the first run. Log every action (widget, old value, new value, reason) to a CSV. Runs hourly or daily via cron / GitHub Actions. You Have worked with a native ad platform API (RevContent, Taboola, Outbrain, MGID). Know what a widget is and why widget-level optimization is where native campaigns are won or lost. Python, REST/OAuth2. We provide: API credentials, target CPA, starting thresholds. Engagement: Fixed price, [$X]. Delivered as a repo with a README. More work behind it if this runs clean. How to apply: Widget A has 500 clicks, 3 conversions, $19 CPA against a $15 target. Widget B has 60 clicks, 0 conversions. Widget C has 300 clicks, 2 conversions, $38 CPA. What does your script do to each, and why? Link one thing you’ve built against an ad API.