How to Build a Mentoring Program That Actually Works

Most corporate mentoring programs do not fail loudly. They just fade. Three months in, half the pairs have stopped meeting. Six months in, the whole thing has quietly disappeared, and nobody wants to be the one to say so.

This is the full method, written by Darren McFarlane, founder of Next Level Mentoring. I have designed and delivered mentoring programs for government departments and ASX20 companies, trained more than 300 people in mentoring, and run one government program for four consecutive years. No theory borrowed from a textbook. This is what has actually worked.

What this guide covers

1. Why most programs fail
2. Start with why
3. Get an executive champion
4. Make it opt-in
5. Let mentees choose
6. Train before day one
7. Set real goals
8. Launch properly
9. Survive the middle
10. Close it well
11. Measure and improve
12. How long it should run
13. How many people
14. Do you need software
15. What it costs
16. What good looks like
17. Three ways to build yours
18. Questions people ask

A mentoring program is one of the cheapest, most effective ways to develop your people, keep your best ones, and move hard-won knowledge from your experienced leaders to the next generation. Done well, it pays for itself many times over. Done the usual way, it becomes another good intention that runs out of steam by winter.

The difference between the two is not budget, and it is not luck. It is structure. This guide walks through the whole thing, step by step, in the order you would actually build it. If you follow it, you will have a program that lasts beyond its first year and becomes part of how your organisation works.

Why most mentoring programs fail

Before the how, it is worth being honest about the why, because nearly every failed program fails for the same handful of reasons.

It usually gets handed to someone in HR, on top of their actual job, with no real structure and no backing from the top. People get matched, then left to figure it out on their own. There are no goals, so the pairs meet for a pleasant chat with no direction, and after a few months there is nothing left to talk about. Momentum drops around month three, and because there is no one whose job it is to notice, the program simply stops.

None of that is a problem with mentoring. It is a problem with the setup. Mentoring runs on trust, structure and senior support, and when you treat it as a tick-box exercise it behaves like one. The good news is that every one of those failure points is fixable, and the rest of this guide is how you fix them.

The mentoring is almost never the problem. The setup is.

Start with why, not who

The first question is not who should be in the program. It is what you are trying to solve. A mentoring program aimed at retaining your mid-level leaders looks different from one aimed at onboarding graduates, or one aimed at breaking down silos between divisions.

Pick the problem first. Are you losing good people? Do your future leaders lack confidence? Is knowledge walking out the door as experienced staff retire? Write the goal down in plain language, and agree how you will know whether you achieved it. Everything downstream, from who you invite to how you measure success, follows from that one decision.

This is also the moment to be realistic about scope. You do not need to roll mentoring out to five hundred people to prove it works. You need to prove it with a smaller group first, then grow it. More on that later.

Step 2

Get an executive champion, not just an HR owner

A long-running program needs a champion at the executive level. Not a sponsor in name only, but someone senior who will talk about it, encourage people to join, and ideally take part themselves.

This matters more than almost anything else. When people see a director or a deputy volunteering as a mentor, the program stops being an HR initiative and becomes something the organisation genuinely values. In the program I have run for the Department of Water and Environmental Regulation, the Director General mentors on the program himself and speaks about it publicly. That single fact does more for participation than any internal email ever could.

Find your champion before you launch. If you cannot find one, that is useful information. It usually means the organisation is not ready to back this properly yet, and it is better to know that now.

Step 3

Make it opt-in, never "voluntold"

If you asked me the single thing that predicts whether a program works, it is not the platform, the materials, or even how good the mentors are. It is whether people chose to be there.

Every strong program I have run is built on people who put their hand up. Mentees who want to grow. Mentors who want to give something back, rather than tick a leadership box. Mandate it and you get bodies in a room. Make it something people apply for and you get energy, commitment and results.

So advertise the program internally, explain what it is and what it is not, and let people register their interest. You cannot force a relationship into existence, and the ones you force tend to be the ones that fizzle first.

Step 4

Let the mentee choose their mentor

How you match people is one of the most important design decisions you will make, and it is the one most programs get lazy about. The research and my own experience both point the same way: the more involved the mentee is in choosing their mentor, the better the relationship works.

The process I use is mentee-led. Each mentee completes a short form about what they want to work on. Based on that, they receive a shortlist of suitable mentors with a little about each, and they rank them in order of preference. The pairing is then finalised from those preferences.

It sounds like a small thing. It is not. A mentee who chose their mentor feels ownership of the relationship from day one. A mentee who was assigned one is already half looking for the exit. Every year, participants ask us to match more carefully, by goals rather than surface similarity, and every year we tighten it further. Matching is the foundation. Get it right and the relationship has a real chance. Get it lazy and even two good people can struggle to click.

Step 5

Train mentors and mentees before day one

Most people have never been formally mentored, and most mentors have never been trained to mentor. If you pair them up and hope, you get the classic failure: the mentor spends every session solving the mentee's problems for them, the mentee stops learning, and both quietly lose interest.

So before the program starts, give everyone the same base of knowledge. What mentoring is, and what it is not. Why a mentor guides rather than gives answers. How to get value from the sessions. It does not need to be a two-day course. An hour of good training, done up front, changes the quality of every conversation that follows and means everyone starts from the same place.

Step 6

Set real goals with every pair

This is the step that separates a mentoring program from a series of nice coffees. Early on, each pair should sit down, ideally with some facilitation, and work out what the mentee actually wants from the year.

Do not expect those goals to arrive fully formed. Most people have only a loose sense of what they want, and the goals almost always shift once they start talking them through with their mentor. That is the point. The conversation refines a vague wish into something the mentee genuinely cares about, and a mentee working towards something they care about is a mentee who keeps showing up.

A word of warning. Do not force these into SMART goals out of habit. Mentoring is a journey, not a performance review. Some of the best goals are closer to themes, and squeezing them into a specific, measurable, time-bound box drains the life out of them. In my programs, this goal-setting session is consistently rated the highlight of the year.

Step 7

Launch with a kick-off that creates energy

The kick-off is where the program becomes real. Get everyone together, in person if you can. Reinforce the training, answer questions, and set out the key dates for the year ahead.

One recommendation: reveal the pairings at the kick-off. It adds a genuine lift of excitement to the room and marks a clear start to the journey.

This is also a rare and valuable moment as you have your current leaders and your future leaders in one place, engaged and paying attention. Use it wisely. A mentoring kick-off is a good time to communicate something that matters to the organisation as a whole.

Step 8

Keep the program alive through the middle

Most pairs hit the ground running and need very little from you after the goal-setting session. The middle of the program is where the risk sits, because this is where momentum quietly drops and where unsupported programs die.

Two things keep it alive. First, light-touch check-ins, so that if a pairing has stalled, someone notices and helps it back on track before it lapses for good. Second, a mid-program group session that brings everyone back together. It gathers deeper feedback, lets mentees see how their peers are driving their own journeys, which lifts everyone's effort, and builds connections across the group. You do not need to hover. You need to be there at the moments that matter.

Step 9

Close it properly and accredit your mentors

A program that just stops leaves value on the table. Close it with a proper ceremony that celebrates the journey and gives each pair one last session to wrap up well. Let people reflect on what they gained. That reflection reinforces the learning and, usefully, produces much better feedback.

Recognise your mentors. They gave their time, and a certificate they can share on their profile is a small thing that means a lot, and it makes them far more likely to volunteer again next year. Building a culture of mentoring depends on mentors who want to come back.

Step 10

Measure what matters, then improve

Gather feedback from every mentor and every mentee, evaluate honestly, and use what you learn to make the next program better. A mentoring program is not something you launch once. It is something you sharpen every year until it becomes part of your culture.

Measure the things that actually tell you whether it worked: how many relationships continued after the program formally ended, whether mentees progressed, how satisfied both sides were, and whether more people wanted in next time. Those numbers build the case for growing the program, and they are the difference between a program that survives a budget cut and one that does not. This closed loop, where each year's feedback improves the next, is the heart of the five-stage framework I use for every program.

How long should a mentoring program run?

For a first cohort, around ten months to a year works well, with pairs meeting roughly once a month. That gives enough time for the relationship to deepen and for real progress against goals, without dragging. Programs shorter than a few months rarely produce meaningful outcomes, because mentoring compounds. The value grows with each session as the pair get to know each other. A one-off session barely scratches the surface.

How many people should be in a mentoring program?

You can start small. A first program of around ten to thirty pairs is very manageable and proves the concept properly. As it succeeds, it grows on its own. The program I run for one government department started with sixteen pairs and grew to thirty over three years, driven almost entirely by participants telling their colleagues. For a managed, high-touch program of the kind that develops your key people, somewhere between thirty and 150 participants is the sweet spot. Beyond that scale, the nature of the program changes.

Do you need mentoring software?

This is worth being clear about, because a lot of organisations reach for software first and wonder later why the mentoring did not happen.

Software solves a logistics problem. It matches names, sends reminders, and generates reports. For very large populations, thousands of people, that is genuinely useful. But logistics is not where programs fail. Programs fail because mentors are not trained, mentees have no goals, pairings are left to figure it out alone, and no one steps in when momentum drops.

Software cannot train a mentor. It cannot coach someone through their goals. And it cannot notice, in month four, that six pairs have quietly stopped meeting. Buying a platform and assuming you now have a program is a bit like buying a gym membership and assuming you are now fit. The tool supports the work. It does not replace it. If your goal is genuine development for your key people rather than administering mentoring at massive scale, the human parts are the whole point.

How much does a mentoring program cost?

It depends on how you build it. Running it entirely in-house costs mostly time, your time, and the risk is that it loses momentum without dedicated structure and accountability. Bringing in outside help costs money but buys you a proven framework, facilitation, and the outside credibility that internal programs often lack.

The more useful way to think about cost is against the cost of doing nothing. Replacing one experienced person can cost anywhere from 150 to 400%t of their salary once you add recruitment, lost productivity and the effect on the team. If a program keeps even one good person who was thinking of leaving, it has paid for itself several times over.

The real question is not what a program costs. It is what losing your best people costs.