Mentoring Tips That Actually Work (Not the LinkedIn Kind)

Most mentoring advice reads like a fortune cookie. "Listen more than you talk." "Lead by example." True, and useless without specifics. Here's what actually moves the needle, drawn from running cohorts of career-changers at Masterschool.

Ask before you answer. A mentee brings a broken function. The instinct is to fix it. Don't. Ask what they've already tried and what they expected to happen. Half the time they find the bug explaining it to you. The other half, you've learned exactly where the gap in their model is — which is worth more than the fix.

Give them the wrong answer sometimes. Not maliciously — as a check. If a mentee accepts a subtly wrong suggestion without pushing back, that tells you they're not yet evaluating advice, just absorbing it. That's the actual skill to build.

Match the correction to the stage, not the mistake. A beginner writing a for-loop instead of a list comprehension doesn't need a lecture on Pythonic style — they need to see their loop work. A mentee six weeks in making the same mistake needs the "why," because they're ready for it and repetition means the first explanation didn't stick.

Name the pattern, not just the instance. "This function is too long" is a one-time fix. "You're solving three problems in one function — notice that pattern, it'll come up again" is a lesson. The difference between mentoring and code review is whether the lesson generalises past the file you're looking at.

Let them be wrong in low-stakes moments. If you catch every mistake before it ships, you've optimised for your own comfort, not their growth. Save the intervention for mistakes that actually cost something to learn from the hard way, and let the cheap ones through.

Show your own uncertainty. "I'd guess X, let's check" models the actual practice of engineering — which is mostly forming and testing hypotheses — far better than arriving with answers already assembled. Mentees who only ever see confident answers learn to fake confidence instead of building competence.

Debrief the win, not just the failure. When something goes right, ask why. Most mentees can tell you what went wrong in detail but go blank when asked what they did well. That blank spot is exactly where the next round of imposter syndrome comes from.

None of this is complicated. It's just slower than giving the answer — and that's the point.