How to Mentor Without Losing Your Billable Hours

Why this matters

The reason good mentors quit mentoring is not that they stop caring. It is that six weeks in, their own numbers are down, somebody notices, and nobody has connected the two facts. The tech quietly goes back to running solo, the apprentice gets handed to whoever is next, and the shop concludes that mentoring "did not take."

This is a measurement problem wearing a motivation costume. A mentor's output genuinely drops for the first stretch, then genuinely exceeds where it started. If the shop only ever looks at the dip, mentoring will always look like a bad trade. If the mentor cannot describe the curve out loud, they cannot defend themselves in the moment somebody asks why their week was light.

This is about protecting the mentor's numbers specifically. The shop-wide question of how many hours to budget for training, and the tactics for teaching inside a single call, are separate and covered elsewhere.

Step 1: Change what you are measured on before you take a learner

Do this first, in writing, before the pairing starts. Not after the first bad week.

The default is that a paired senior is still scored as an individual, against the same target they hit solo. That is not a hard target, it is an impossible one, because the person is now producing work and producing a technician at the same time and only one of those is being counted.

Ask for one of three changes, in order of preference:

• Score the pair, not the person. Your target becomes the pair's combined output against a stated baseline. This is the cleanest and it is the one that makes the crossover visible.
• Apply a stated adjustment for the ramp weeks. A named percentage off your individual target for a named number of weeks, agreed up front. Less good than pair scoring because it expires on a calendar rather than on evidence.
• Carve out the teaching hours. Your target is calculated on your non-teaching hours only, and the teaching hours are logged as a separate code.

If none of the three is available, understand what that tells you: the shop wants the outcome without funding it, and you should not take the learner. Taking them anyway is how a mentor ends up with a worse review and a resentment that lasts years.

Step 2: Know the shape of the curve so you can name it

The curve has four phases. Numbers vary by trade, learner, and job mix, but the shape holds and being able to state it turns a defensive conversation into a factual one.

Phase 1, pure drag. The learner produces nothing and consumes your attention. Every task takes longer because you are narrating, watching, and correcting. Your output is meaningfully below your solo baseline.

Phase 2, partial offset. The learner reliably handles the non-technical load - hauling, staging, cleanup, photos, paperwork, the parts runner role. Your output recovers most of the way back because those minutes were coming out of your day anyway.

Phase 3, crossover. The learner performs signed-off portions of the work unaided while you do the parts that need you. The pair now produces more than you did alone. This is the point everything is aimed at.

Phase 4, return. The learner runs the simple calls solo some days, and the pair's combined output clears the baseline by a wide margin.

The number you want in your head is not how long each phase takes. It is how big the hole gets at its deepest and how many weeks to fill it back in. Track that once and you never have to argue about mentoring again.

Step 3: Hand over the minutes you were already losing, first

The single largest error mentors make is teaching the interesting part first. The interesting part is the diagnosis, which is the part where your speed advantage is largest and the learner's slowness costs most.

Hand over the low-skill, high-minute work first, on day one:

• Staging tools and materials at the work location before you walk in
• Protective coverings down and picked up
• Photos, meter readings written down, the job record filled in
• Cleanup and haul-out
• The parts run

None of that requires a sign-off. All of it comes out of your clock every single day. In most residential work this is a real and recoverable slice of the call, and moving it to the learner buys back most of Phase 1 in the first two weeks.

The bonus is pedagogical rather than economic: a learner who stages your tools learns what tools the task needs, and a learner who writes down your readings sees every number you take before they understand what it means. That is free curriculum.

Step 4: Put the teaching in the slack, never in the critical path

Every call has parts where a minute costs you a minute and parts where it does not. Teach in the second kind.

Slack: the drive, the walk-out to the truck, the wait while a system stabilizes, the wait while something drains or fills, the write-up. Critical path: the customer standing there, the diagnosis with the clock running, anything with a hard finish time.

Concretely, on a call where you have to wait several minutes for a reading to settle, that wait is the highest-value teaching window in your day and most techs spend it on their phone. Use it for the question you were going to ask anyway: "what do you think this is going to read, and why."

The rule to hold: if the customer can see you and the clock is running, you demonstrate and narrate. If nobody is watching and you have slack, you hand over the tool.

Step 5: Use windshield time as the real classroom

Drive time is already a sunk cost and it is the only genuinely uninterrupted block in a field day. It is also where diagnostic reasoning transfers best, because you can talk through a case with no equipment in front of you and no pressure to be right.

Two formats that work:

• Pre-brief on the way in. "Here is what dispatch says. What do you expect to find, and what is the first thing you would measure." Two minutes. It forces a prediction, and a prediction that turns out wrong is the most teachable event in the trade.
• Debrief on the way out. "What did we actually find, and where was your prediction off." One minute. Do it before the next call overwrites it.

Neither costs a billable minute. Both are the highest-yield teaching in the week.

Step 6: Bank the returns, deliberately and early

Phase 3 does not arrive on its own. It arrives when somebody makes a dispatch decision to send the learner without you on a task they are signed off on. If nobody makes that decision, the pair stays paired indefinitely and the mentor never gets their hours back - which is the most common way this whole arrangement quietly fails.

Push for it. As soon as a learner has cleared your sign-off bar on a task, that task's simple calls get dispatched to them alone, with you reachable by phone. Your first solo release should happen well before you feel entirely comfortable, because the discomfort is the point and a phone call is a cheap safety net.

Start with calls where the failure mode is a callback rather than a hazard, and where the customer situation is forgiving. Do not start with an emergency, a commercial account, or anything with a hard deadline.

Step 7: Declare the weeks you cannot mentor

Some weeks you cannot teach. A run of emergencies, a big install, a short-handed crew. Say so, out loud, to the learner and to dispatch: "this week I am not teaching, you are helping, we pick it back up Monday."

That sentence does two things. It stops the learner from reading a bad week as a personal signal, which is the most common cause of a learner disengaging in month two. And it keeps you from doing the worst version of mentoring, which is half-teaching under time pressure - the learner gets the tool, gets interrupted, gets the tool taken back, and learns that they are an obstacle.

Do not let declared weeks exceed roughly one in four. Past that you do not have a mentoring arrangement, you have a helper, and you should relabel it honestly.

A worked example: twelve weeks of a pair, in hours

Illustrative numbers, and the shape is the point - run your own from your own board.

A senior tech working a 40-hour week produces 32.5 billable hours solo. That is the baseline every week below is measured against.

Weeks 1 and 2 are pure drag. Weeks 3 and 4 recover most of it because Step 3 moved the staging, cleanup, and documentation to the learner. Weeks 5 and 6 are the crossover, driven by the learner performing one signed-off portion of the work unaided. From week 9 the learner runs simple calls solo two days a week, which is what produces the jump.

Now total it. The cumulative shortfall through week 6 is 5.5 plus 5.5 plus 2.5 plus 2.5, less 0.5 and 0.5, which is 15.0 hours behind baseline at the deepest point. Against a 32.5-hour baseline week, the entire investment is worth about 46 percent of one week's output, spread across six weeks so that no single week looks alarming.

Filling the hole: weeks 7 and 8 return 3.5 each for 7.0, leaving 8.0 to recover. Week 9 returns 9.5 and clears it. So the pairing is fully paid back partway through week 9, and by the end of week 12 the shop is 30.0 hours ahead of where the senior would have been alone.

The point of doing this arithmetic is what it does to the conversation in week 4. In week 4 the mentor is 16.0 hours behind cumulatively and looks like a problem. With the curve written down in advance, week 4 is on plan. Without it, week 4 is where mentoring gets cancelled - about five weeks before it would have started paying.

What changes the answer

A learner who is not new to the trade. Someone cross-training from an adjacent trade skips most of Phase 1, because they already own the tool handling, the customer manner, and the safety instincts. Expect crossover much earlier and a shallower hole.

Flat-rate versus time-and-materials pricing. Under time and materials with two techs billing, a second body can add billable hours immediately and the hole is shallower but the incentive to release the learner solo is weaker. Under flat rate, the pair adds no revenue until the learner splits off, so Step 6 matters more.

Emergency-heavy job mix. A board that runs mostly urgent calls has very little slack, so Step 4 has nowhere to put the teaching. Either accept a much longer ramp or deliberately route a few non-urgent calls to the pair each week.

A mentor who is the only person who can do a critical task. Their hours are the shop's constraint, so the drag is more expensive than the table suggests. Pair them anyway, because that concentration is itself the bigger risk, but shorten the ramp by planting the practice repetitions on a bench rather than on live calls.

How to verify you got this right

Four checks, at week 6.

Pull the pair's output against the baseline and see whether the curve is flattening. If week 6 is still as deep as week 2, the learner is not absorbing the non-technical minutes and Step 3 was never really done.

Check the sign-off record. If there is not a single signed-off task after six weeks, Phase 3 has nothing to trigger on and the pairing will not cross over on its own.

Ask the learner what they ran unaided last week. If the answer is "I handed him things," you have an assistant.

Ask the mentor whether anyone has questioned their numbers. If they have been questioned and had no answer ready, go back to Step 1, because the arrangement is running on goodwill and goodwill runs out in about eight weeks.

References

• See related: How to Train When You Are Too Busy to Train, for setting the shop-wide hours budget this sits inside
• See related: How to Teach a Skill in the Field Without Losing the Day, for tactics inside a single call
• See related: Why Most Small-Shop Training Fails in the Second Month
• See related: The Competency Sign-Off SOP, for the sign-off that triggers the solo release in Step 6
• Trade-standard practice for field-service utilization measurement in small contracting shops