Data sources and methodology
How eoidata.com reads, validates, aggregates and publishes SkillSelect EOI data, including pathway counts, month-end statuses and derived metrics.
Last updated:
Overview
eoidata.com does not replace official sources. It organises public records into a form that is easier to search and compare. This page describes the core month-end SkillSelect EOI dataset; invitation rounds, occupation eligibility, points rules and EOI timelines carry additional source-specific explanations on their respective pages.
Core data source
Month-end EOI snapshots come from the Australian Department of Home Affairs SkillSelect EOI data panel and are used under CC BY 3.0 AU.
We do not hold applicants' names, passport details or private SkillSelect account information. The site publishes counts aggregated by dimensions such as occupation, visa, state, status, month and points score.
Monthly snapshots
The official dataset changes when the panel is reloaded. After new monthly data becomes available, we read the panel and preserve that release as a snapshot with its reference month and reload time. Historical months are not rewritten to match a later status, so trends compare what the public dataset showed at different points in time.
Each data version records:
- the latest snapshot month;
- the official panel reload time;
- our synchronisation and publication times;
- the available historical months;
- publication row counts and a data-version identifier.
Use the month shown on a data page as the time basis for that analysis.
Counting unit
Within one occupation, visa, state, points, month and status combination, core statistics count distinct EOI identifiers. However, one person may create multiple EOIs and one EOI may include multiple visa or state pathways.
This means:
- a number within one filtered combination is an EOI count for that combination;
- a total across states or visas may be a sum of pathways;
- pathway counts are not distinct applicant counts;
- pool size cannot establish an individual's place in a queue.
Status definitions
- Submitted: EOI pathways still marked Submitted at the snapshot.
- Invited: EOI pathways still marked Invited and not yet moved to another status at the snapshot.
- Lodged: EOI pathways marked Lodged at the snapshot.
All three are point-in-time statuses, not events newly occurring in that month. An EOI can be changed, withdrawn, expire or move to another status between snapshots, so a monthly difference cannot be interpreted directly as submissions, invitations or lodgements during the month.
Points and missing values
Points scores come from the official records. Records without a usable numeric score remain unknown or are excluded from distributions that apply only to known scores; pages should identify when a denominator contains known values only. Ranges, minimums, maximums, medians, cumulative counts and proportions are derived by eoidata.com from the aggregated records.
A month-end Invited points distribution is not an invitation-round minimum. Official invitation rounds are a separate dataset shown on the Invitation Rounds page.
Derived results
The following are calculations made by eoidata.com from public records, not conclusions published directly by the Government:
- month-over-month changes and rates;
- cross-dimension totals, proportions and rankings;
- cumulative points counts, percentiles and profile distributions;
- EOI timeline statistics built from historical status records.
We do not generate predictions to fill missing values. If the official source suppresses a small
value using a form such as <5, we preserve that form instead of guessing the number. Derived
features such as timelines display their sample size and additional limitations on the page.
Validation and publication
Before publication, new data is checked for its structure, reference month, statuses, duplicate records, row counts and aggregate results. Published data is separated from source captures, historical snapshots are retained by month, and a data version identifies the current release.
These checks reduce the chance of error but cannot guarantee that the official source, processing code or our explanations will always be correct. Verify consequential figures against the relevant official source.
Corrections
If you find a possible issue, email hello@eoidata.com with the page URL, filters, reference month and figure. We will distinguish source-data issues, methodology questions and processing errors. Material corrections that affect interpretation will be recorded in our Changelog.