Mentorship is genuinely valuable and the formal version is overrated. Most useful mentorship happens as an accumulation of specific conversations with several people, not a single designated mentor relationship — and that version is far easier to build.
**Why the formal ask usually fails**: 'Will you be my mentor?' asks for an open-ended, indefinite commitment from someone who doesn't yet know whether the relationship will be worth their time. It's a large request wearing the clothes of a small one, and most people decline or agree and then quietly disengage.
**What works instead:**
1. **Ask a specific question about a specific thing.** 'I'm deciding between specialising in X or Y — you made that choice a few years ago, would you have twenty minutes to tell me how you thought about it?' Small, bounded, flattering, and easy to say yes to.
2. **Follow up with what you did.** This is the step that converts a one-off into a relationship. 'You suggested I try X — I did, here's what happened.' People become invested in others who act on their advice, and this is what causes them to offer more.
3. **Repeat, occasionally.** After three or four of these, you have a mentor. Nobody had to use the word.
**Where to find these people:**
- **Inside your company first.** Senior colleagues, people one or two levels ahead, people on adjacent teams. Proximity and shared context make this the highest-yield source, and it's the one people overlook while searching for a famous stranger.
- **Your manager**, for some things — though the evaluative relationship limits what you can safely discuss.
- Communities, open-source projects, local meetups.
- Former colleagues and university contacts.
- Formal programmes, if your company runs one — these do work, and the structure removes the awkwardness.
**What to look for**: someone two to five years ahead of you is often more useful than someone twenty years ahead. They remember the specifics of your current problem, their advice is current, and they're more available.
**How to be worth mentoring**, since this determines whether it continues: come with specific questions rather than 'any advice?', do the thing they suggested, be respectful of their time, and offer something back where you can — a perspective from your area, help with something, or simply being someone whose progress is satisfying to watch.
**And the reframe on 'how important'**: important, but not a prerequisite. Plenty of people progress well without one, and one excellent conversation a quarter with someone slightly ahead of you delivers most of the value that people imagine comes from a formal arrangement.