A job title is a compression of a much longer story: years of experience, scope of ownership, the actual mix of tasks someone did day to day, squeezed into two or three words that mean something different at every company that uses them. "Senior engineer" at a ten-person startup and a Fortune 500 platform team can describe two genuinely different jobs: different scope, different decision authority, sometimes an entirely different day-to-day skill set.
Screening primarily by title, or by the adjacent proxy of years-of-experience, imports that compression error directly into a hiring decision. It filters out people whose title undersold their actual scope, and lets through people whose title oversold it, and it does both silently, without anyone noticing until well after an offer's been made.
What skills-based screening actually requires
The phrase gets used loosely enough that it's worth being specific about what it takes to do properly. It is not a looser version of resume screening, and it is not just adding a coding test. It means defining, before you look at a single candidate, the specific capabilities the role actually requires (not the capabilities a template job description usually lists) and then assessing directly against those, through structured work samples, scenario-based interviews, or a real technical exercise rather than a title match.
Done properly, this is more work up front, not less. Someone has to actually define what "can operate independently in a legacy codebase with incomplete test coverage" looks like as an assessable skill, rather than assuming a senior title implies it. Skipped, the process just reverts to title-matching with extra steps.
Where this matters most
The return is highest exactly where title compression is worst: career-changers and people moving between company sizes or industries, where the title genuinely doesn't map cleanly to the target role's requirements. It's lowest where the role is narrow and well-defined enough that title and scope reliably line up, in which case the extra rigor is overhead, not signal.
The honest version of this practice isn't a slogan about looking past resumes. It's a specific, more expensive assessment process that pays off precisely in the cases where the shortcut it replaces was most likely to be wrong.