Valuable 1-on-1s do three things: they surface information you wouldn't get otherwise, they build capability, and they prevent problems before they're expensive. The structure matters less than the questions.
The best framework is simple: spend one-third on what's working, one-third on what's not, and one-third on what's next. "What's working" isn't just praise—it's pattern recognition. When someone describes a win, dig into why it worked. That's how they learn to replicate success. "What's not" isn't a complaint session—it's obstacle removal. What's blocking them that you can actually fix? "What's next" is development. What skill are they building? What project could stretch them?
Here's the economic logic: A 30-minute 1-on-1 costs roughly $50-150 in salary, depending on the person. If it prevents one resignation (saving you 12-18 months of salary in replacement costs), you'd need to have effective 1-on-1s for years before breaking even on just that one retention event. If it helps someone get 10% better at their role, the productivity gain pays for a year of weekly meetings in a month.
The pattern I see with managers who run great 1-on-1s: They ask questions that can't be answered with "fine" or "good." They take notes and follow up on what was discussed. They make commitments and keep them. They don't dominate the conversation.
The test: If your direct report left the 1-on-1 with something actionable, for example, a decision you made, an obstacle you removed, a skill you helped them build, then it was valuable. If they left with nothing but a half-hour less to do their actual work, you're both wasting time.
Your 1-on-1s should either prevent expensive problems or accelerate growth. Anything else is theater.