Back to the library
Article · AI coaching

How long should a session with my AI Coach last?

The ideal session length, what practitioners say about micro-sessions, and the calendar rituals that make short pauses actually happen.

Honestly… it depends. And that's common sense.

AI (and AI coaching) isn't magic. At the end of the day, you have to process the insights and validate the output. Which means you need to dedicate a bit of time and bring your own contribution, especially if you're working on something complex or tied to ambitious goals.

I get it: we all wish there were quick fixes to big challenges. But a few seconds won't cut it. On the other hand, you don't need a full hour either (who even has that in a packed schedule?). Just 10–15 minutes can already take you surprisingly far.

Why 10–15 minutes is the sweet spot

A traditional coaching session lasts an hour because that's what the economics of a human coach's calendar require: travel, scheduling, billing. Strip those constraints away and an interesting question appears: how much of that hour was the actual work?

In our coaching sessions, we see the useful unit of reflection is usually one situation and one question. "How do I open tomorrow's 1:1?" is a 12-minute session. "Who do I want to be as a leader?" is not, but it decomposes into a dozen 12-minute sessions, each anchored to something real that happened that week.

The research backs the intuition that more hours is not the point. Meta-analytic work on workplace coaching finds solid effects overall (Theeboom et al., 2014), and notably does not show that longer engagements automatically produce better outcomes: what matters is that reflection actually happens and connects to real goals. And the randomized trials that put AI coaching to the test found it matched human coaches on goal attainment using regular, short, goal-focused conversations rather than marathon sessions (Terblanche et al., 2022).

What practitioners say about micro-sessions

Talk to executive coaches about what actually changes their clients and few of them credit the session itself. The session is where awareness gets created; the shift happens in the moments between sessions, when the client catches themselves mid-pattern. Graßmann and Schermuly (2021) point to exactly this as the structural advantage of the AI format: because it removes cost and scheduling friction, it makes coaching available at the moment of need, right before the difficult conversation, right after the tense meeting, instead of at the next slot in someone's calendar, days later.

That's the real argument for micro-sessions. Not that short is more convenient than long, but that timely beats thorough. A 10-minute pause the morning of a decision is worth more than a brilliant hour of analysis a week after it.

The trap on the other side: sessions so short they're reflexes, not reflection. If you close the session before you've said anything that surprised you, you checked a box: you didn't pause. That's the difference between motion and progress, a pattern we've dissected in the illusion of progress.

Calendar rituals that make it actually happen

A few tips: the first three haven't changed since the first version of this article, because they still work:

  1. Iterate, don't overload. Don't try to do everything in one go. Block short 10-minute slots in your calendar (personally, if it's not in my calendar, it rarely happens).
  2. Use voice. Just like on WhatsApp, speak into your device and let it turn your words into text. It gives more context, feels more natural, and helps you express yourself better. Yes, you'll need a safe space to do it, but it's worth a try.
  3. Pick up where you left off. You can always come back to your session later and continue from the exact spot you stopped.
  4. Anchor sessions to events, not willpower. The rituals that survive are the ones welded to something already in your week: 10 minutes before every 1:1 with a direct report, 10 minutes after any meeting that raised your pulse, 15 minutes on Friday to close the week. The trigger is the event; the session is the response.
  5. Protect the ritual, not the performance. Some sessions will be flat. That's fine: a flat 10-minute session on a Tuesday keeps the habit alive for the breakthrough one on Thursday. Showing up beats showing off, and we've written the full case for that in consistency over perfection.

If you're wondering what to actually do in those 10-minute slots, we've mapped five concrete scenarios from a manager's week, with the prompts, in how should you use your AI coach?

In short: invest yourself

That's the key. Within your possibilities, of course, but it's the condition for progress. The evidence that coaching in this format pays off is solid and growing (we've reviewed it honestly in does AI coaching work?), but no study can do the ten minutes for you.

Block the first slot this week, and practice it with Lumia, your AI coach.

Frequently asked questions

How long should an AI coaching session last?

For most situations, 10 to 15 minutes: one real situation, one question, one takeaway. Complex topics don't need longer sessions: they need more of them, each anchored to something that actually happened. Reserve longer sessions (30+ minutes) for rarer deep work like a quarterly self-review or preparing a high-stakes negotiation.

Are short coaching sessions actually effective?

Yes, provided they're regular and honest. Research on workplace coaching shows solid effects without longer engagements automatically outperforming shorter ones, and randomized trials found AI coaching matched human coaching on goal attainment through short, regular, goal-focused conversations. The failure mode isn't brevity: it's sessions so rushed you never say anything that surprises you.

How often should I talk to my AI coach?

Two to four short sessions a week is a sustainable rhythm for most managers. More useful than a frequency rule: anchor sessions to recurring events, before 1:1s, after tense meetings, Friday close-of-week, so the trigger is built into your calendar. A timely 10-minute pause beats a thorough hour that happens a week too late.

References

  1. Theeboom, T., Beersma, B., & van Vianen, A. E. M. (2014). Does coaching work? A meta-analysis. The Journal of Positive Psychology, 9(1), 1–18. DOI: 10.1080/17439760.2013.837499
  2. Terblanche, N., Molyn, J., de Haan, E., & Nilsson, V. O. (2022). Comparing artificial intelligence and human coaching goal attainment efficacy. PLoS ONE, 17(6), e0270255. DOI: 10.1371/journal.pone.0270255
  3. Graßmann, C., & Schermuly, C. C. (2021). Coaching with Artificial Intelligence: Concepts and Capabilities. Human Resource Development Review, 20(1), 106–126. DOI: 10.1177/1534484320982891