Research · Nodes
Research
the paper · arxiv · econ.em · april 2026N=10,765 hires · one F500 carrier
Decision Traces: What Multi-System Data Fusion Reveals About Institutional Knowledge in Enterprise Hiring
- Of 8,181 ATS keywords parsed, the 3,597 testable predicted production zero times after Bonferroni correction; thirty were anti-predictive.
- Every day saved before the first production milestone is worth $54.35 in projected annual production, per producing hire.
- One industry-experience filter screened out hires behind $17.7M in observed annual premium credit.
- Decision Traces in Hiring: Multi-System Data FusionDecision traces connect ATS, HRIS, and behavioral data so you can see whether your screening predicts performance. A 10,765-hire study found it does not.
- Do Resume Keywords Predict Job Performance? 10,765 HiresIn a study of 10,765 hires, zero of 3,597 resume keywords predicted production after correction, and 30 were anti-predictive. Here is what the data showed.
- Why ATS Keywords Fail to Predict Job PerformanceATS keywords measure what a candidate has done, not how they will perform. A 10,765-hire study and the Hidden Workers report explain why keyword screening fails.
- Does a College Degree Predict Job Performance?In a study of 10,765 hires, a college degree was a weak predictor of job performance, with production rates overlapping across education levels. What the data showed.
- The $17.7M Cost of a Single Hiring FilterOne screening rule would have rejected 2,863 producing agents worth $17.7M in annual premium credit. How connecting ATS and HRIS data exposes the cost of a filter.
- AUC 0.647 to 0.735: Data Fusion in Hiring PredictionPersonality assessment reached AUC 0.647 alone and 0.735 fused with ATS and behavioral data in a 10,765-hire study. What the numbers mean, read correctly.
- Which Personality Types Predict Sales Performance?In a 10,765-hire study, personality type was the strongest single predictor of sales production, ranging from 36.8% to 0.0% by type. What the data showed.
- Why Too Much Leadership Drive Hurts Sales PerformanceIn a 10,765-hire study, leadership drive followed an inverted-U. Moderate drive outperformed extreme drive in a structured sales role. What the data showed.
- The Speed-to-Production Constant in Enterprise HiringEach day faster to the production milestone was worth about $54 per agent in a 10,765-hire study. How speed to production became a measurable economic constant.
- The Speed-Volume Trade-Off in High-Volume HiringWhen one carrier doubled hiring, its production rate fell from 41.5% to 21.1%, but producers ramped faster and produced more. The trade-off, and why it stays hidden.
- Prediction vs Moderation: A Better Test for Hiring AIMost hiring AI is judged on whether it predicts who succeeds. A 10,765-hire study suggests the better test is who benefits most from the conditions you control.
- When a Manager Retires, Their Judgment LeavesDecades of hiring judgment disappears when a manager retires, because it lives in no system. How decision traces capture institutional knowledge as a queryable record.