Coverage & methodology

Know the coverage.
Understand the evidence.

Exende observes reviewed public career sources. The catalogue gives you a consistent view of those sources and the changes recorded while they are observed.

Current catalogue

Real counts. Clear definitions.

The latest values are fetched from the public aggregate API. They describe observed catalogue coverage, not the whole hiring market.

Active jobs
Companies observed
Reviewed sources
Data freshness

From source to record

A deliberate observation process.

01

Reviewed public sources

Collection starts from deliberately enabled public career sources. A source review and a technical observation are distinct from a grant of unrestricted redistribution rights.

02

Consistent normalization

Listings retain a source URL alongside normalized employment and workplace types. Skills use a deterministic taxonomy. Missing values remain unknown.

03

Recorded changes

Meaningful changes create immutable versions. Unchanged listings receive at most one visibility confirmation per UTC day; a complete source snapshot determines closures.

Reading the data

What the fields can tell you.

Locations & countries

Location text comes from the source and may describe more than one place. It is not a normalized city identifier. Country codes are assigned only from explicit supported country evidence; city-only and ambiguous locations can remain null.

First & last observed

These dates mark Exende observations, not necessarily a listing’s original publication date. The public freshness timestamp is the latest canonical job observation; it does not mean that every source was checked then.

History & movement

Created, updated, and closed events describe recorded versions. Initial collection can create many records at once. Historical coverage does not establish a company’s complete hiring history.

Hiring metrics

Role mix and velocity describe observed activity. Acceleration is withheld until the history spans the requested window. These metrics are descriptive evidence, not predictions or guarantees.

Jobs remain source-scoped. Similar roles across different sources are not automatically merged. Use the metrics documentation to interpret history readiness and normalization limits.

Use data with context

Keep the source and the boundary.