What Makes New Teacher Mentoring Actually Work?

Forty-four percent of teachers quit within their first five years. A 2025 survey by the Center for American Progress found that nearly 70 percent of early-career teachers have either left or considered leaving the classroom, with lack of support among the most-cited reasons. One in three reported never participating in a formal induction program at all.

New teacher mentoring is the most consistently supported intervention for early-career retention and effectiveness. Ingersoll and Strong’s review of 15 empirical studies, published in the Review of Educational Research, found that beginning teachers who participated in induction programs had higher retention, better classroom instruction, and higher student achievement. A separate meta-analysis found that mentored teachers had 2.85 times higher odds of remaining in the profession. The evidence is not ambiguous. Mentoring works. The question is why so many mentoring programs don’t.

The Mentoring Gap: What Research Recommends vs. What Schools Deliver

The National Institute for Excellence in Teaching (NIET) published a policy report examining why new teacher mentoring falls short. The core finding: mentoring programs frequently become a “buddy system” rather than a structured learning opportunity. The role of the mentor is not well defined, little training or release time is provided, and the program is disconnected from the school’s instructional improvement goals.

NCTQ research confirms the pattern. Effective mentoring requires same-field mentors, multiple years of support, regular observation and feedback, and collaborative activities with other teachers. As the number of induction components increases, turnover decreases. But most districts provide only the minimum: assign a mentor, hope they connect, check the compliance box.

The structural reason is straightforward. Mentors are classroom teachers. They have their own students, their own lesson planning, their own grading. The time available for mentoring is whatever is left after teaching five periods, attending meetings, and handling the daily demands of a full instructional load. A mentor with 30 minutes of shared planning time per week cannot provide the frequency of coaching that beginning teachers need to develop their practice.

What Effective Mentoring Actually Requires

Ingersoll and Strong’s review found that the most effective induction programs share specific features. Beginning teachers who participated performed better at keeping students on task, developing workable lesson plans, using effective questioning practices, and adjusting classroom activities. These are not abstract skills. They are the practical moves that new teachers struggle with every day, and they require coached practice, not just advice.

NIET identified three elements that distinguish programs that work:

1. Focus on instructional improvement. Mentoring should center on teaching practice, with intentional investment in time, training, and the use of an evidence-based instructional rubric. Conversations about “how are you doing?” are supportive but don’t build skill.
2. Trained and supported mentors. Selecting, training, and compensating mentors signals the priority of the work. Mentors need clear expectations and the skills to provide constructive feedback, not just emotional support.
3. Alignment with school systems. Mentoring that operates as a standalone program, disconnected from school goals, observation frameworks, and professional development, produces weaker results. The mentoring program should reinforce what the school is already working on instructionally.

Kraft, Blazar, and Hogan’s 2018 meta-analysis of 60 causal studies found that coaching programs produce effect sizes of 0.49 standard deviations on instructional quality and 0.18 standard deviations on student achievement. The coaching that drives these effects looks remarkably similar to what effective mentoring looks like: observe, reflect, identify one area to improve, try again, get feedback. The challenge is delivering this cycle frequently enough to build lasting skill in a new teacher’s practice.

The Frequency Problem

The most consistent finding across the mentoring research is that intensity matters. NCTQ found that teachers with mentors from the same field who also participated in collaborative activities were more likely to stay at their school and remain in the profession. Programs that lasted multiple years outperformed single-year programs. The meta-analysis showing 2.85 times higher retention odds measured programs with sustained, structured support.

But sustained support requires someone’s time. A mentor teacher who meets with their mentee once a week for 30 minutes is providing roughly 18 hours of mentoring per school year. A beginning teacher who is struggling with classroom management, lesson pacing, student engagement, and content delivery needs far more than 18 hours of support to develop competence in those areas.

Joyce and Showers’ foundational research on coaching and transfer found that implementation rates for new skills jump from under 20 percent to 80 to 90 percent when sustained coaching accompanies training. The frequency that drives that jump is what most mentoring programs cannot deliver, because the mentor has a full-time teaching job.

This is not a program design problem. It is a capacity problem. And it is the same capacity constraint that limits instructional coaching more broadly.

Extending Mentoring with Technology

The gap between what new teachers need and what mentors can deliver is where AI-powered coaching infrastructure earns its place.

AI Coach by Edthena delivers the structured reflection and coaching cycle that the research identifies as the difference-maker. A new teacher finishes a lesson, reflects on what happened through a guided coaching conversation, identifies one area to work on, and develops a specific action step. The cycle runs as often as the teacher needs it, between mentor meetings, before the next class period, or after a lesson that didn’t go as planned. The coaching cadence the research demands stops being constrained by mentor availability.

For observation and feedback specifically, VC3, Edthena’s video coaching platform, enables mentors and new teachers to work together even when their schedules don’t align. A beginning teacher records a lesson segment. The mentor watches the video and provides time-stamped feedback. The new teacher reviews the feedback and responds. This asynchronous coaching cycle means mentoring is not limited to the 30-minute window when both teachers happen to be free.

And when the principal conducts formal observations of new teachers, Observation Copilot helps turn those observations into timely, framework-aligned feedback. For a first-year teacher, the formal evaluation observation may be one of the few times they receive detailed written feedback on their instruction. Making that feedback specific, timely, and connected to clear next steps is especially high-stakes for someone still building their professional foundation.

New Funding Is Coming for Mentoring Programs

Texas HB-2, signed in June 2025, created the PREP (Preparing and Retaining Educators through Partnerships) allotment, the largest teacher preparation funding investment in the state’s history. The PREP Mentorship track provides $3,000 per beginning teacher for structured induction programs, with mandated mentor stipends of at least $1,000 and required TEA-administered mentorship training. Funding begins in the 2026-27 school year.

Texas is not alone. States across the country are increasing investment in structured mentoring as a retention strategy. The question for districts receiving this funding is not whether to create a mentoring program. It is how to build one that produces the results the research predicts, rather than the buddy-system outcomes that NIET documented.

Districts that build coaching infrastructure into their mentoring programs, using platforms like AI Coach and VC3 to deliver structured weekly cycles and asynchronous observation, are better positioned to deliver the sustained, structured support that the research says works. The funding exists. The evidence base exists. The gap is in implementation infrastructure.

Building a Mentoring Program That Works

If you are designing or redesigning a new teacher mentoring program, start with the research:

1. Select mentors carefully. Same-field, same-building when possible. Mentors should be strong instructors, not just willing volunteers. NCTQ data shows that field-matched mentors with collaborative activities produce the highest retention.
2. Define the mentoring around instruction, not logistics. NIET’s research shows that programs focused on instructional improvement outperform those focused on orientation and emotional support. Use an instructional framework to ground the conversations.
3. Plan for frequency, not just duration. A two-year program with monthly check-ins is not the same as a two-year program with weekly coaching touchpoints. The research ties intensity to outcomes. Reach weekly cadence by adding platforms like AI Coach or video coaching, which deliver the structured virtual coaching cycles a single mentor cannot staff alone.
4. Connect mentoring to the observation system. New teachers are being formally observed and evaluated. If the mentor and the principal are giving different feedback using different frameworks, the beginning teacher gets conflicting signals. Align the mentoring program to the same instructional framework used for evaluation.
5. Measure beyond retention. Retention is the most common metric for mentoring programs, but it is not the only one that matters. Track instructional growth through observation data and student outcomes, not just whether the teacher is still at the school in year two.

Ready to build mentoring infrastructure that extends your coaches’ reach? Learn more about AI Coach, explore VC3 video coaching, or try Observation Copilot for free.

Frequently Asked Questions

Does new teacher mentoring improve retention?

Yes. A meta-analysis of formalized induction programs found that mentored teachers had 2.85 times higher odds of remaining in the profession, with a 27.1 percentage point retention advantage by the fifth month. Ingersoll and Strong’s review of 15 studies confirmed positive effects on retention across multiple program types.

What percentage of new teachers leave the profession?

Research from the University of Pennsylvania found that 44 percent of teachers quit within five years. A 2025 Center for American Progress survey found that nearly 70 percent of early-career teachers have either left or considered leaving, with one in three reporting they never participated in a formal induction program.

What makes a mentoring program effective?

NIET and NCTQ research identifies several factors: same-field mentors, focus on instructional improvement rather than just emotional support, trained and compensated mentors, alignment with school instructional frameworks, multiple years of support, and collaborative activities beyond one-on-one meetings. Programs with more induction components consistently show lower turnover.

How does Texas HB-2 fund new teacher mentoring?

The PREP Mentorship track under Texas HB-2 provides $3,000 per beginning teacher for structured induction programs. Districts must provide mentor stipends of at least $1,000, use TEA-administered mentorship training, and ensure release time for mentoring activities. Funding begins in the 2026-27 school year.