How PathScore Finds
Your Hidden Careers
Three scoring layers multiply together. Each layer answers a different question. The result: career matches ranked by fit, desire, and discovery.
What We Collect From You
Everything the algorithm needs to score careers
Your Occupation
Matched to O*NET title
"I work at Amazon warehouse" → Stockers and Order Fillers
Your Skill Profile
Skill and expertise dimensions
Coordination 2.5, Active Listening 2.8, Monitoring 2.5 …
Hidden Skills
Side projects, hobbies, languages
"I sell electronics on eBay" → Sales, Negotiation, E-commerce
AI Enrichment
Context extracted from your words
"lead picker at Amazon" → Team leadership, Performance metrics
Your Priorities
What matters most to you
Higher salary, escape physical labor, work-life balance
Risk Tolerance
How much change you can handle
Play it safe → Moderate → Swing for the fences
Fed into 3 scoring layers
Base Match Score
"Could this person actually do this job?"
Skill Similarity
Compares your skill scores to what the job requires. Like measuring how much two puzzle pieces overlap.
Knowledge Similarity
Do you already know the subject areas this job needs? Transportation, math, customer service, etc.
Ability Match
Physical and cognitive abilities — strength, spatial reasoning, attention to detail.
Gap Size
How much training would you need? Smaller gaps = higher score. A tiny stretch beats a huge leap.
Hidden Skills Bonus
This is where the magic happens. Your eBay business maps to "Sales & Marketing" and "Negotiation" — which lights up careers you'd never have found otherwise.
Example: Warehouse Worker → Logistics Coordinator
Then adjusted by what you want
Preference Multiplier
"Does this career match what the user actually wants?"
"I want higher salary"
Boosts careers paying more. Pay cuts get penalized hard — a big cut makes a career nearly invisible in results.
"I want work-life balance"
Penalizes 50+ hour cultures and shift work. Boosts predictable Mon–Fri roles. A 60-hour job loses almost half its score.
"I want job security"
Three signals combined: Is the field growing? Are there many openings? Is AI unlikely to replace it?
"I want to escape physical labor"
Desk and remote jobs get a massive 1.5× boost. Heavy physical jobs are nearly removed from results.
"I want creative / meaningful work"
Boosts jobs requiring creativity, design, or originality. Also boosts healthcare, education, and social services.
How the multiplier scale works
Example: User wants "higher salary" + "escape physical labor"
Risk tolerance adjusts the spread
Then diversity is enforced
Diversity Bonus
"Are we showing enough hidden gems from different industries?"
New sector? Score goes up.
The first career from a brand-new industry gets a 1.25× boost. This is what surfaces surprising careers you'd never have googled.
Same sector again? Score goes down.
If 3+ results are already in your current industry, the next one gets penalized. Stops the "bus driver → truck driver → train driver" problem.
Example: Building results for a Warehouse Worker
Scores combined, then safety checks
Safety Checks
After scoring, we run rules to catch algorithm failures before results reach the user.
40% Diversity Floor
At least 40% of top 15 results must be cross-sector. If not, promote hidden gems from further down.
Salary Sanity Check
If user said "higher salary," remove any results paying less (unless within 5%).
Effort Reality Check
Validate training hour estimates against real certification databases. Flag anything unrealistic.
Smart Defaults
Auto-set sort and filters based on priorities. "Swing for the fences" → sort by salary. "Play it safe" → sort by easiest.
Final ranked results
Your Ranked Career Matches
Top 15–20 careers, ranked by PathScorer.
Each tagged as cross-sector, adjacent, or same-sector.
Example output: Warehouse Worker, "swing for the fences"
PathScorer Algorithm v1 · Data: O*NET 28.1 + BLS OES 2024 · 1000+ occupations scored per user
You're worth more than
your job title suggests.
Upload your resume. 1000+ occupations scored. Results in minutes.
See my career matches — free