Attribution and ROAS measurement for apps and websites
Discovery engine for unidentified data phenomena
Stage
MVP
Not given
Founded
Not given
Not given
Based in
AE
US
Pricing model
Not given
Contact for pricing
Pricing
Free: up to 2,000 conversions per month per product. Pay-as-you-go: $0.04 per conversion after the free allowance. Enterprise: custom pricing for high volume and specific needs.
Not given
Key features
Deterministic Android attribution via Google Play Install Referrer
iOS attribution via SKAdNetwork with versioned conversion value mappings
First-party website sign-up attribution tied to ad clicks
Server SDKs for Node.js, PHP and Python to report conversions from backend
ROAS per campaign, ad set and ad for Meta and TikTok
Meta Conversions API and TikTok Events API forwarding
Revenue measurement from RevenueCat and AdMob
Retention and ARPU cohorts from day zero
No advertising ID or fingerprinting by default
Apple's signed postback stored raw and versioned
Support for seven data shapes including time series, events, networks, and spatial measurements
32 categories spanning domains from financial markets to genomics
Discovery engine for changing relationships and recurring behaviours
Reproducible execution with documented model configuration and runtime context
Confidential compute options including Intel TDX, AWS Nitro Enclaves and AMD SEV
Developer API and GitHub integration for agent-based exploration
Support for investigating evidence and testing explanations
What makes it different
Deterministic attribution by default without advertising IDs or fingerprints
One product on every plan with no feature withdrawal as volume grows
Published pricing with no form or minimum, clear inputs and failure modes
ROAS, attribution and revenue included in every plan including free
SKAdNetwork conversion value mappings versioned instead of overwritten
Separates phenomenon discovery from explanation, allowing rigorous investigation of root causes
Retains all outputs alongside original data for inspection and reproducibility
Works with existing data without requiring pre-identification of phenomena
Keeps execution accountable through documented model identity and configuration
Traqio
Kavara™