On mentorship — what I got wrong for years

I was a bad mentor for a long time without knowing it.

Not malicious — inattentive. I gave advice generously, answered questions thoroughly, connected people to opportunities when I could. By the visible metrics of mentorship, I was doing the right things. What I was actually doing was treating every conversation as a consulting engagement: identify the problem, provide the solution, move on.

That's not mentorship. That's advice-giving from a position of experience.

The difference took me too long to understand, and I only understood it when I had a mentor who did it properly.

What she did differently: she asked questions I didn't know the answers to. Not difficult technical questions — questions about what I actually wanted, what I was avoiding, what I was afraid of. She was less interested in helping me solve the problem I presented with than in helping me see the problem underneath it.

The result was that the conversations changed me rather than just informing me. That's the distinction.

The version of mentorship I was practising was efficient but shallow. I could generate useful advice at speed because I'd encountered similar situations before. The result was that people left conversations with better answers to the questions they'd asked. They rarely left with different questions.

The other thing I got wrong: I mentored in my own image. When someone came to me with a career question, I implicitly evaluated their options against the path I'd taken and the things I valued. This is a hard bias to see from inside it. Someone who values security differently than I do, or who has family constraints I don't have, or who is optimizing for something I've never prioritized — they don't need my framework. They need their own framework, examined clearly.

The version of mentorship I'm trying to practise now: fewer answers, more questions. Being genuinely curious about what the person wants, not what I would want in their position. Sitting with "I don't know" rather than filling the silence with analogies from my own experience.

I'm still not good enough at it. But I know now what I'm trying to do.

Working on this in production?

We do this work directly alongside engineering teams — architecture review, migration, and hands-on enablement.