This video is available to ERE Pro subscribers

Subscribe to watch

Already a member? Log in

Source Your Engineering Talent Map

November 9, 2020 · Talent42 Digital ·

About this video

Building an accurate talent map for a technical role starts with treating the job requisition like a puzzle to break apart, then researching the candidate landscape before ever running a search. The approach walks through a real requisition for a principal cloud software engineer role requiring Python, Go, and Kubernetes, showing how to translate each line of a job posting into search terms covering what the candidate does, what platform they work on, which languages and secondary technologies apply, and where they're located.

Key sourcing tactics covered include:

  • Searching for CVs and resumes on unconventional sites like GitHub.io, SlideShare, and Google Sites using strings such as site:github.io "about" OR "resume" OR "CV" combined with job titles and keywords
  • Weighing the risk of scraping LinkedIn Recruiter directly against safer alternatives, and instead running site:linkedin.com searches through Google or Bing
  • Breaking a large keyword string into smaller pieces using minus operators to get around Google's practice of capping visible results at a few hundred, then aggregating each partial search into a spreadsheet for a fuller candidate count
  • Finding personal "About Me" pages across blogs, Medium, and portfolio sites by combining that phrase with job titles, technical keywords, and location, while excluding university and job-board domains
  • Searching Facebook indirectly through Google using a site:facebook.com "profile photo" string paired with general titles and company or industry terms rather than narrow technical keywords

Throughout, the emphasis is on never relying on a single site or network for a talent map, since candidate populations scatter unevenly across platforms and one method alone will always produce a skewed picture. Watching for false positives, such as candidates whose listed location doesn't match where they actually live, and adjusting search terms when a keyword pulls too many or too few results, rounds out a repeatable process for building a talent map from free, public sources rather than paid research services.