LLenny's Podcast
← All frameworks
LeadershipClaire Hughes Johnson (former COO of Stripe)

Hypothesis-Based Coaching (Explorer, Not Lecturer)

Coach by exploring a well-informed hunch with someone, not lecturing them from data

Difficulty
Moderate
Time to result
~weeks to results
Steps
4
Confidence
90%

A manager's job is not to be the expert who tells people what to do, but to enable them to be their best. This framework treats a manager's intuition as a scientific hypothesis, a well-informed piece of intuition, and explores it collaboratively with the person rather than waiting for overwhelming data or delivering a lecture. It reframes coaching as mutual discovery.

Origin

Claire Hughes Johnson's management framework developed across roles at Google and as COO of Stripe, reframing 'intuition' as a testable scientific hypothesis to legitimize acting on early signals.

Core principles

  • 01A manager's job is to enable people to be their best, not to be the smartest expert in the room
  • 02Intuition is just a well-informed piece of intuition, the same thing a scientific hypothesis is
  • 03Waiting for a million data points before raising an issue wastes the coaching window
  • 04Good intent plus an owned observation makes exploration safe
  • 05Coaching can be mutual, the person can explore right back at you

How to run it

  1. 1

    Form a hypothesis from partial data

    Take the limited data you have, an observed pattern in someone's work or behavior, and form a hypothesis instead of waiting for complete certainty.

    Pro tip Frame it internally as 'I've seen this pattern of your work, have you seen it?' rather than a proven conclusion.

    Watch out Don't wait until you have a million pieces of data; the moment to coach passes and the pattern entrenches.

  2. 2

    Open with a light, curious question

    Start exploratory: 'How did you feel that presentation went?' Let them self-assess before you offer your read.

    Pro tip Genuine curiosity signals you're trying to help them see something, not catch them out.

  3. 3

    Share your observation as a mirror

    Offer your read as owned perception: 'I felt like you were maybe a little nervous, I'm just exploring, were you?' Add specific physical evidence if they're unsure.

    Pro tip Point to concrete signals (leg shaking, repeating, pushing chair back from the table) so it's observable, not abstract.

    Watch out If they flatly deny it, be willing to back off; the goal is discovery, not winning.

  4. 4

    Explore the fix together

    Move from the observed behavior to a collaborative solution, e.g. how to vocalize discomfort instead of physically exiting the circle.

    Pro tip Invite them to observe you back; mutual exploration builds the trust that makes future coaching land.

In the wild

The chair-pushed-back tell

Claire noticed team members who, when uncomfortable with a meeting topic, would physically push their chair back from the table, ending up several feet from the group. They had no idea they were doing it. Rather than lecture, she explored how they could vocalize the discomfort instead of physically exiting the circle.

Holding up the mirror to an unconscious behavior gave the person a concrete, coachable thing to change.

Common mistakes

Believing the manager's job is to be the expert

Managers who think their role is to know the answer and tell people what to do fail to create the environment where people do their own best work.

Waiting for overwhelming data before coaching

Holding back an observation until it's fully data-proven means the coaching moment is missed; a well-formed hypothesis explored early is more useful.

Is it for you?

Best for

New and mid-level managers coaching direct reports on behavioral or performance patterns

Not ideal for

Moments when someone explicitly asks to be shown exactly how to do a concrete task, where direct instruction is what's needed

From the transcript

your job is to enable people to be their very damn best on your team

37:30

I have a whole other framework which is about hypothesis based coaching

38:00

guess what a scientific hypothesis is it's a well-informed piece of intuition

38:30

take some data form a hypothesis and then explore it with the person

38:30

all you're doing is holding up a mirror you're saying

39:30

From the episode

Lessons from scaling Stripe

Claire Hughes Johnson (former COO of Stripe)