The Challenge

TeachStart, a competitive teacher fellowship program, faced a critical operational challenge: their recruitment team was spending thousands of hours conducting phone screens, creating a significant bottleneck in their hiring pipeline while stretching their resources thin.

Key Pain Points:

• Recruiters conducted over 2,500 phone screens annually, with only 47% of invited candidates completing them

• High candidate attrition between stages wasted recruiter time and delayed hiring

• No data-driven way to identify which early application or resume screen factors were predictive of candidates that would receive or accept offers

• Uncertainty about which parts of the recruitment process added genuine value to the candidate experience

The Manta Solution

TeachStart deployed Manta to analyze their recruitment data and identify optimization opportunities. Within minutes, the system analyzed over 11,000 candidate records, performed simulations, and provided actionable insights.

Manta identified a key opportunity: candidates with strong resume screening scores on two factors out of the 30 that were assessed (representing about 16% of applicants) could potentially skip phone screens without compromising hiring quality.

The Analysis

Manta conducted multiple sophisticated analyses in minutes that would have taken a team of human data scientists days or weeks:

1. Competency Correlation Analysis: Analyzed 30 rubricized data points that are collected in the interview process against hiring outcomes and program performance

2. Predictive Modeling: Identified which early-stage competencies best predicted successful program completion

3. Pipeline Simulation: Created detailed models comparing current vs. modified recruitment processes

4. Risk Assessment: Tested multiple scenarios including pessimistic assumptions to validate findings

Organization Impact

Projected Efficiency Gains:

• 400 hours of recruiter time saved annually

• 31% reduction in total screening workload

• Zero additional technology investment required

Projected Quality Impact:

• Potential to increase offers to qualified candidates by over 500 

• Faster progression of high-quality candidates to interviews

• More recruiter time available for sourcing and engaging candidates

• Even when modeling pessimistic assumptions, the model still showed significant improvements in finding more qualified candidates to hire

Key Takeaways

1. Hidden Efficiency: TeachStart discovered certain metrics (like "Writing Skill" and "Why TeachStart" scores) were highly predictive of success, while others (like prior teaching experience) had no predictive value
   
   

2. Data-Driven Decisions: AI can uncover insights that challenge conventional recruiting wisdom, about where to maintain a “high bar” to hire the best candidates
   
   

3. Rapid Time-to-Value: Finding the most predictive assessments in minutes enables process optimization to improve organizational efficiency