The Coaching Sprint at three years: pick your Coaching Sprint Goal last

Here's the thesis. Pick your Coaching Sprint Goal last, not first. The original post taught the cadence, the anchor on the retro, the 66-day timing. What it under-emphasised: a Coaching Sprint Goal that isn't nested under a Business Objective and a Growth Goal floats. And a floating Sprint Goal produces motion without direction. The fix is a fifteen-minute conversation. The cost of skipping it is months.

The pattern we kept seeing

You finish CSP-SM. You're sold on the methodology. You go back to your team. You open the next retrospective with intent. You ask powerful questions. You hold your tongue. You take notes on what surfaces.

Three weeks later, you don't know what changed. The team feels better, maybe. There's a vague positive direction. But if your sponsor asked you tomorrow "what's the team going to do differently by August?" you'd struggle to answer.

That's the pattern. We've watched it across cohorts in our certification programme, across alumni in the community, across our own engagements. The Coaching Sprint runs. The retros happen. The questions get asked. And six weeks in, no one can name the gap that's getting closed.

The reason is almost always the same: the SM picked a Coaching Sprint Goal in isolation. "Help the team improve their estimates." "Increase psychological safety." "Get better at conflict." Each of these is a reasonable thing for a coach to care about. None of them is a Coaching Sprint Goal.

The goal hierarchy we teach in class but didn't put in the blog

In the CSP-SM class, we walk students through four levels. They sit on top of each other like a pyramid. You write them in the 4DX format: from X to Y by when.

Here's a concrete example. Composite, drawn from the engagements we've supported in 2023 and 2024.

The difference is not the quality of the coaching. It's that the second version has a road. The first version has good intentions and a retro.

Why the original post under-played this

Honestly: because we thought naming the methodology was the hard part. We were wrong. The methodology is easy. The discipline of picking goals top-down, in a domain where everyone is busy and the sponsor isn't always sure what they want, is the hard part. That discipline is what we now spend the most facilitation time on in the CSP-SM class.

And it's the part the original blog skipped over. The post mentioned a coaching agreement. It pointed at the 7 Steps of Systems Entry. It implied the goal-setting work without putting a frame around it. Three years of teaching since then has shown us: if you don't write all four levels before you plan the sprint, you'll keep skipping the top two. The Sprint Goal feels concrete. The Growth Goal feels abstract. The Business Objective feels like someone else's job. So readers picked the concrete one. And then they wondered why the sprint felt aimless.

The fifteen-minute conversation

The fix is small in shape. Before you plan the first Coaching Sprint, sit with the team's PO or the sponsor. Ask three questions. Write the answers in 4DX format.

The three questions, in order:

1. What is the company trying to do in the next 12 months that this team is part of? Get one sentence. If you get a paragraph, push back until it fits in one sentence. That's your Business Objective.
2. What's the one thing about how this team works that, if it improved, would move that business objective faster? Predictability? Quality? Customer-feedback cycle time? People retention? Pick one. That's your Growth Goal.
3. What's the smallest version of that growth goal that this team could plausibly hit in the next three months? The team's version, not the sponsor's. That's your Coaching Sprint Goal.

Write each answer down in from X to Y by when. Read all three aloud as a chain. If the chain doesn't make sense (Sprint Goal doesn't serve Growth Goal, Growth Goal doesn't serve Business Objective) you're not ready to start coaching. Go back and re-do the question that broke the chain.

Fifteen minutes. Usually less.

What changes when you do this

Three things, all of them visible inside the first sprint.

The retro stops feeling exploratory. You walk in with a thesis about what kind of conversation matters this week. Some of those theses are wrong, and you discard them. But you're choosing, not catching. The team feels the difference even before they can name it.

The sponsor stops calling. You sent them three sentences at sprint planning. If they ever want to know what's happening, they re-read those three sentences. They don't ask you for status updates because the status is in the sentences they already have. Maybe one clarifying question a month. That's a lot less.

And you stop running coaching activities. You start running a coaching sprint. The activities you run inside it serve the goal. The ones that don't, you cut. Same shift a delivery team makes the first time it has a Product Goal and stops debating which features to put in the next sprint.

What to do this week

If you have a Coaching Sprint in flight right now, stop. Don't run the next retro yet. Block fifteen minutes with the team's PO before that retro happens, and ask the three questions. Write the three answers in 4DX format. If you get stuck on the Business Objective (you usually will, the first time), don't panic. The question itself is the contribution. Most teams don't have a clear business objective; the SM asking for one is the first useful thing you do.

The original Coaching Sprint post is still the right starting point. The methodology hasn't changed. What's changed is what we now know to put in front of it. The Coaching Sprint is the engine; the four-level goal is the road.

If you're working through this and want a second pair of eyes on the goal chain before you start a sprint, write back when the welcome email lands. Replies reach Erkan and Brock directly. Most of what we see in those conversations is the Business Objective being missing or under-specified, which is also exactly where most agile transformations get stuck. So you'll be in good company.