How we report graduate outcomes.

Data sources

Outcome figures on this site come from two sources, always named on the figure itself:

• Public alumni outcomes (interim). Until partner universities supply verified placement reports, we research publicly visible alumni outcomes on LinkedIn — self-reported employers and roles for a named graduating cohort. This is the source of every figure currently marked “Illustrative demo data”.
• Partner-reported data. Once a data-sharing agreement is in place, universities supply cohort placement reports directly. These rows replace the interim dataset and are labelled with the partner report and verification date.

Collection method

For the interim dataset, researchers enumerate public LinkedIn profiles of a university's India-campus graduating cohort and record employment status, employer, and — where publicly stated — compensation. Only profiles that can be attributed to the named cohort year are counted. The number of respondents (n) and the cohort year are published next to every figure; a percentage or package number is never shown without them.

Minimum sample size (min-n suppression)

We do not publish outcome statistics for cohorts with fewer than 10 respondents. Below that threshold a median can be swung by a single person, and publishing it would be misleading precision. Instead you will see “Cohort too small to report” with the n itself disclosed — you always know how much we know.

Update cadence

The interim dataset is refreshed at most once per quarter and each row carries its collection date. Partner-reported rows are refreshed per cohort cycle. Any figure whose source data is older than 12 months fails our build-time validation and is removed from the site until re-verified.

Illustrative demo data

The current dataset is an illustrative demo dataset: agent-researched from public alumni outcomes and not yet human-verified against university records. Every surface using it carries an “Illustrative demo data” badge. The badge is removed figure by figure as rows are human-verified or replaced by partner-reported data.