Test people programs the way product teams test features.
HR teams make consequential decisions about onboarding, training, hiring, and culture — usually without measurement. Decision Process lets you run controlled comparisons of people programs and measure what actually improves retention, performance, and satisfaction.
Experiment Templates
Ready-to-run experiments
New Hire Onboarding Program
“Structured 90-day onboarding improves 1-year retention vs. 30-day program.”
Conditions
- 30-day onboarding (control)
- 90-day onboarding
Metrics
- Retention Rate (%)
- Performance Rating (1–5)
- Employee Satisfaction (1–10)
Training Format Comparison
“Blended learning (in-person + online) outperforms fully online for skill transfer.”
Conditions
- In-person only (control)
- Online only
- Blended
Metrics
- Training Completion Rate (%)
- Performance Rating (1–5)
- Employee Satisfaction (1–10)
Interview Process Structure A/B
“Structured behavioral interviews improve hiring quality vs. unstructured interviews.”
Conditions
- Unstructured interview (control)
- Structured behavioral
Metrics
- Time to Hire (days)
- Performance Rating (1–5)
- Retention Rate (%)
Worked Example
90-day vs. 30-day onboarding program across two hire cohorts
A company assigns new hires to either a 30-day onboarding program (current standard) or a new 90-day extended program with structured check-ins and mentorship. 48 hires per group. One-year retention and 6-month performance rating used as outcomes.
Results: 1-year retention rate (%)
30-day onboarding (control)
mean: 68.4%
95% CI: 55.2–80.2
90-day extended program
mean: 85.1%
95% CI: 73.6–93.6
P(better) = 93%
The 90-day program improves 1-year retention by +16.7 percentage points (d = +0.61, medium-large effect) at 93% posterior probability. Performance ratings also improved on average. Recommendation: extend standard onboarding to 90 days, with structured 30/60/90-day check-ins.
Explore other domains
Run your first hr & people ops experiment
Private beta — tell us about your use case and we'll get you set up.