Software

Open-core Python packages for advertising measurement, placement auditing, and algorithmic bidding.

Each package is designed for both analysts and AI coding agents (Claude Code, Cursor, Codex, Antigravity): deterministic CLI interfaces with --json output, offline synthetic datasets for testing without credentials, and installable domain skills on skills.sh.

SpotLift

Linear TV & Omnichannel Attribution

Measures the immediate web and sales response of linear TV airings against minute-level Google Analytics 4 (GA4) traffic and Amazon Marketing Cloud (AMC) conversions.

  • Open-Core (spotlift, MIT): Diurnal baseline estimation, single-spot lift integration, weighted collision attribution, and cost-per-incremental-visit (CPIV).
  • Pro (spotlift-pro, Commercial): Shifted-Gamma EM deconvolution for overlapping spots, creative wear-out modeling, 1–36h lag halos, and automated Excel/HTML reporting.
pip install spotlift
npx skills add flovv/spotlift --skill tv-to-web-measurement

DV360 BidKit

Custom Bidding DSL Linter, Simulator & Attention Engine

Validates, simulates, and compiles custom bidding scripts for Google Display & Video 360 (DV360) locally. Implements zero-vendor-fee attention-based dwell bidding alongside Google 2026 DSL compliance.

  • Open-Core (dv360-bidkit, MIT): AST validator for Google DSL 2026, safe local simulator (no eval()), Google S2 readiness checks, and open empirical attention dwell tables (Simonov & Valletti 2024 / Lumen benchmarks).
  • Pro (dv360-bidkit-pro, Commercial): Multi-factor attention calibration (Format × Position × Domain Spillover), automated MFA clutter discounting, Empirical Bayes shrinkage for low-volume publishers, and 65k DSL script minification.
pip install dv360-bidkit
npx skills add flovv/dv360-bidkit --skill dv360-custom-bidding

TubeSentry

YouTube & Video Placement Auditor

Audits placement reports (“Where ads showed”) across YouTube, Performance Max, and DV360. Identifies Made for Kids (MFK) inventory, passive audio/music streams, foreign script delivery, and performance sinkholes, then generates instant exclusion lists for Google Ads Editor.

  • Open-Core (tubesentry, MIT): Deterministic risk pattern detection (MFK, low-intent gaming, passive audio, foreign scripts), spend waste quantification, impression-weighted quality scoring, and Google Ads Editor CSV export.
  • Pro (tubesentry-pro, Commercial): Multimodal creative-to-channel fit engine, channel-level semantic topic clustering, automated exclusion syncing, and white-label client audit decks.
pip install tubesentry
npx skills add flovv/tubesentry --skill youtube-placement-audit

Ad Sentry

Google Performance Max, Demand Gen & Meta Advantage+ Auditor

Audits automated black-box ad campaigns in seconds. Unmasks hidden channel spend splits (Search vs. Display vs. YouTube), deconstructs PMax for Google Meridian MMM, detects non-converting “zombie” shopping SKUs, flags low-quality auto-generated video waste, and prevents internal campaign bidding collisions.

  • Open-Core (ad-sentry, Apache-2.0): True channel spend reconstruction, Google Meridian & PyMC-Marketing MMM export bridge (ad-audit export-meridian), feed zombie SKU detection (clicks with zero conversions), brand cannibalization checks, and offline CSV/CLI diagnostic reporting.
  • Pro (ad-sentry-pro, Commercial): Interactive white-label HTML pitch decks, multi-account MCC batch scanning, creative asset decay modeling, and automated Merchant Center product suppression scripts.
pip install ad-sentry
npx skills add flovv/pmax-sentry --skill pmax-delivery-audit

Commercial Licensing & Support

Open-core packages are free under MIT/Apache licenses on PyPI. Commercial Pro extensions are distributed via Polar.sh as pre-compiled native binary wheels or licensed toolkits with instant license key delivery and a 14-day refund policy. For support or licensing inquiries: teschner@gmail.com.