This is part of GoPerfect Labs — where we publish findings from our data team. We run controlled experiments across our sourcing platform and share what the data actually shows.
We ran an experiment across 500 real recruiting searches — the same role, the same requirements, the same database of 800M+ profiles. One search used traditional Boolean keyword logic. One used GoPerfect's semantic AI, which understands context, trajectory, and meaning rather than exact word matches. Then we scored every result.
.png)
What Boolean search was never designed to do
Boolean search logic — AND, OR, NOT, exact keyword strings — was designed for structured databases. It's precise and controllable. It's also deeply literal in a world where talent is anything but.
A recruiter searching for a "Senior Backend Engineer with Python and AWS experience" using Boolean will find candidates who wrote those exact words. What it won't find: the engineer who listed "cloud infrastructure" instead of AWS, or the one who spent three years at a Python-adjacent stack and could ramp in a week, or the candidate whose skills page says "distributed systems" but whose entire career history maps perfectly to the role.
.png)
The hidden candidate pool you're not reaching
Every keyword-first sourcing approach creates an invisible exclusion layer. Candidates who describe their experience in different language, who came from adjacent industries, who have the skills without the keywords — they don't exist in a Boolean world.
Here's the same search, run both ways, for a real GTM role — VP of Sales at a Series B fintech:
.png)
The Boolean result looked right on paper. The semantic result would have been invisible to a traditional search — but turned out to be exactly what the hiring manager wanted. This isn't a cherry-picked edge case. In our dataset, 1 in 3 of the highest-scoring semantic results were invisible to the equivalent Boolean search.
Where semantic search changes the quality equation
The delta between Boolean and semantic wasn't consistent across role types. We broke it down and found the gap was widest in specific categories — which tells you where keyword sourcing is most actively failing your pipeline.
.png)
The categories with the highest miss rates — career switchers, non-standard titles, strong-trajectory candidates — are also some of the most valuable hires for fast-growing companies. These are the people who bring fresh perspective, grow into the role, and often outperform candidates with the "right" background.
"The best hire for your open role might describe themselves in language your Boolean search was never written to find."
What "semantic" actually means in practice
When recruiters hear "semantic AI," it's easy to picture something vague and magical. In practice, it's specific. GoPerfect's semantic search understands three dimensions that Boolean ignores entirely:
.png)
.png)
