Diversity data methodology

Data Sources

All team-level data was taken from our Human Resources Information System (HRIS). All company-level data was taken from our HRIS where possible. This provided the most complete data set on which to base our analyses. LGBTI* identification data was collected via an anonymous, optional survey conducted in January and February 2016.




All data is representative of full-time Atlassians, except where indicated below. The data excludes contractors and interns. Team-level statistics exclude "teams" of one.

All data is accurate as of Feb. 1 of the year indicated. 12-month hiring refers to the hiring period from February 1, 2015 to February 1, 2016.




Team: Teams were defined using the "Team" classification within HRIS (e.g., "Recruitment" or "JIRA Product Marketing"). We chose to rely on this definition because we assume, on average, that Atlassians work more closely with similarly-classified Atlassians. We also chose this definition because it provides us with the most standardized unit of analysis and ensures the replicability of our analyses over time.

Female/Women: This category describes all full-time Atlassians who have identified as female within our HRIS. We have coded gender as binary due to data limitations and for simplicity, although we recognize that gender exists on a spectrum and these categorizations may not accurately represent all Atlassians.

Total workforce: Percentage of full-time Atlassians who have self-identified as female.

Technical: All development, product management, and design roles on our Software teams as well as non-Software roles where coding is a primary job requirement.

Leadership: All Atlassians with direct reports and senior-level individual contributors.

Race: All race data is for full-time U.S. Atlassians only, as race and ethnicity data is not available in our other locations. Our chosen categories are modeled off of the U.S. government's Equal Employment Opportunity Commission (EEOC) ontology, although we acknowledge these are imperfect categorizations.

LGBTI* Identification: All data was provided as a self-identification, without specific categories imposed. The (*) in our chosen acronym implicitly recognizes all of the categories not explicitly named in the acronym. Sub-categories are not mutually exclusive.

  • Gender identity: All Atlassians who have self-identified that their gender identity falls within the LGBTI* community, broadly defined.
  • Sexual orientation: All Atlassians who self-identified that their sexual orientation falls within the LGBTI* community, broadly defined.

International: This category currently accounts for Atlassians under a work visa. This does not include Atlassians who have become permanent residents or citizens of countries in which they work either before, or during, their time with Atlassian. This showcases the truly global nature of our workforce. 


Team Analysis


Our first priority when undertaking this analysis was to decide what constitutes a "team". The traditional definition is, "a number of persons associated in some joint action." We chose to use this definition, and proxy it within our data using the method described in "Definitions" above.

The next step was to decide the schema by which we would group our teams. We chose an ontology that aligns with the kinds of teams we see our customers are on. Here are the kinds of functions within each of our departments listed in the report (in alphabetical order):

  • Customer Support: Customers for Life, Technical Support, Account Management, Advocates, etc.
  • Finance: Analytics, Planning, Business Development, Tax, etc.
  • HR: HR, Recruitment, Experience
  • IT: Administration, Business Systems, Security, Workplace Technology, etc.
  • Legal: Legal, Risk & Compliance, Corporate Development
  • Marketing: Corporate Branding, Public Relations, Marketing, Experts, etc.
  • Software Development: Engineering, Design, Product Management, Infrastructure, etc.


Team Averages by Department


In order to look at the average team within each department, we first found the total representation of each demographic group within each team. We then took the averages of the representation as a percentage “to find the average representation per team per department.

Analyses for race data excludes teams without a U.S.-based team member. The percentages shown represent only data for US-based team members.

Team Diversity Highlights

We looked at each individual team within a department. The summary percentage in this table represents the percentage of teams within that department that share the characteristic described in the header (e.g., 15% of our Software Development teams have a black team member, while 36% of our Customer Support teams do.)