In an on-going process, the inclusion of new sources is being conducted through a number of channels:

• Corporate careers websites
• HR blog entries
• Keyword catalogue-based Google research
• Cross-countries check
• Input from the public and our Advisory Board

We assess the informative value and quality of each source through a number of aspects:

• Type of source (e.g. audit, poll, review portal)
• Type of organisation (e.g. non-profit vs. for-profit)
• Type of research (e.g. self-enrolment vs. independent research)
• Validation processes, manipulation barriers, complaint tools
• Academic supervision and/or governmental patronships

The maximum amount of information is collected for each publicly accessible source.

• The above refers to level of detail as well as historical archives
• Our Data Scientists team is led by a PwC-certified Data Analyst
• Quality controls and blind checks ensure continuous accuracy
• We utilise AI-driven tools and scrapers of both major types:
  open-target (web crawlers) and custom-tailored (data extractors)

250 sources come in many different formats. This step converts all data for a common denominator:

• Abbreviations (e.g. “LE” = “Leading Employers”)
• Pre-fixes (e.g. „The Disney Channel” = “Disney Channel”)
• Founder’s first names (e.g. “Smith Group” = “John Smith Group”)
• Languages (e.g. “Swiss Post” = “Schweizer Post”)
• Mergers and re-brandings

‘More’ does not necessarily mean ‘better’! A holistic assessment is what counts.

• Each category of sources has a different base score
  (e.g. multi-stage audit vs. simple self-commitment)
• These are weighted by 50% – 150% for individual findings
  (e.g. rankings, categories, percentages, grades, scores, etc.)
• Negative findings result in points deduction, up to disqualification
• Final score depends on the variance of areas with positive findings