The CFO wants to know the ROI on coaching. The board wants metrics. And you're standing there thinking: "How do I measure whether a team learned to trust each other? How do I quantify the moment someone felt safe enough to admit they were wrong? How do I show that coaching is valuable when the value lives in conversations, not dashboards?"
Why measuring coaching feels impossible
Coaching is asymmetric work. You invest time in conversations that may not show up in any metric for months. You help a team reframe a problem, and then they solve it faster. You coach an engineer through a confidence crisis, and then they ship better work. You have one conversation with a lead that changes how they listen, and that ripples through a whole organization. But where's the data point? Where's the evidence?
Most coaches respond to this in one of two ways: They give up and accept that coaching is "unmeasurable," or they over-measure and track everything, turning coaching into a compliance exercise. Neither works.
The hidden metrics that matter
"You can't measure human judgment by looking at dashboards. But you can measure whether sustainable agile requires human judgment, and it absolutely does."Mike Cohn · Succeeding with Agile
Here's the trick: you don't measure coaching directly. You measure the downstream effects of good coaching, and you do it without turning coaching into a box-ticking exercise. The metrics are simple — they just live one level of abstraction away from the coaching itself.
When a team is truly coachable, you see: velocity becomes more predictable, action items from retrospectives actually get completed, team members speak up more in planning meetings, context switching decreases, and time-to-value shortens. These aren't coaching metrics. They're team health metrics. But they're a direct result of good coaching.
The framework that works
Start by picking three metrics that matter to your organization. Maybe it's sprint predictability, engineering satisfaction, or time-from-idea-to-production. Maybe it's bug escape rate or technical debt ratio. Pick things the business cares about. Then establish baseline measurements the sprint before you start coaching. Run a coaching intervention — deeper retros, better team conversations, clearer decision-making — for six to eight sprints. Then measure again.
The beauty of this approach is that it's honest about causation. You're not claiming that your coaching caused the improvement — you're showing correlation over time. But correlation is enough for stakeholders. And it gives you something that pure intuition doesn't: credibility.
The human part still matters
The metric tells stakeholders that something shifted. But the real proof lives in the narrative. Can the team articulate what changed? Do they feel different? Would they choose to keep working this way? When a product lead says, "I didn't believe retrospectives could matter, but after three sprints I could see my team shipping faster and taking more ownership," that's evidence. When engineers say, "Someone finally asked what we actually needed instead of just telling us what to build," that's evidence too.
So measure the downstream. Tell the story. And create space for people to notice the change themselves. That combination — metric plus narrative plus self-awareness — is what creates lasting belief in coaching.
The trap to avoid
Don't measure coaching activities. Don't count how many retros you ran or how many coaching conversations you had. Those metrics are useless and actively harmful — they incentivize doing the work instead of doing work that matters. If a team needs a conversation about decision-making authority, that's infinitely more valuable than running five more standups.
And don't optimize for the metric. You'll end up with a team that has great sprint predictability but hates their job. Measure for signal, not for gaming.
How to make the case to your CFO
Frame it this way: "We're running a six-sprint coaching experiment. We've identified three metrics that correlate with delivery quality and team satisfaction. We'll establish baseline, run our coaching work, and remeasure. If the metrics improve by 20%, we continue and expand. If they don't, we change approach." Most CFOs will take that deal. It's honest, bounded, and measurable without being reductive.
What you're actually doing is creating permission to coach the way coaching works — through relationship, attention, and skillful questions — while giving stakeholders the evidence they need. It's not perfect. But it's real.