Collection & Freshness

This page documents how often each metric is updated and how far back history goes. Use it to understand data freshness when building integrations or models.

Collection Approach

We pull metrics via a combination of methods:

• Official APIs -- Used where available and compliant
• Public data collection -- When APIs are limited, we use compliant collection focused on publicly visible counters and public discussion data
• Standardization -- Raw data is normalized to consistent formats across platforms

The goal is reliable, comparable metrics that can be analyzed consistently across companies and over time.

Time Conventions

• Daily aggregates use America/Los_Angeles day boundaries
• Daily values reflect completed prior-day rollups under that day boundary
• Intraday ingestion and end-of-day aggregates are separate views: intraday is event-driven, daily is normalized for one-value-per-day workflows

Point-in-Time

AltIndex history is event-time stamped, not point-in-time versioned, and there is no vintage tracking today. Each value is stored against the date it describes, and a value can be revised in place when a source corrects it. If your workflow requires knowing what was visible on a given past date, treat this dataset as event-time history and ask us about point-in-time delivery before building on it.

Every figure below is measured directly against the production database, and the build that publishes this page fails if the measurements are stale or disagree with this catalogue. When a dataset's collection ends, it is either removed from this catalogue or explicitly marked historical; it is never left presented as live.

Data Quality

Our collection approach prioritizes:

• Consistency -- Same methodology applied across all companies
• Accuracy -- Direct mapping to official accounts via our knowledge graph
• Completeness -- Coverage across all major social platforms
• Timeliness -- Daily updates for core metrics, near real-time for discussion data