Design a Rotation Program That Fits Your Company

This is the full playbook from “Good Judgment Needs Direct Contact with Customers.” The book prints a shorter version at the end of that chapter, and this page keeps every step and every example.

Use these steps to build your own rotation program. Don’t import another organization’s program unchanged.

1. Decide who is exempt, and keep the list short. A customer-facing role already in daily contact doesn’t need a separate rotation. Everyone else joins: engineering, product, finance, leadership, including the people who feel furthest from customers. The distance is why the program exists.
2. Set the cadence and calendar it before you need it. Quarterly is a reasonable default. Write it into the standing calendar the way you would a recurring planning meeting, and it survives the first busy sprint that would otherwise be a convenient reason to skip it.
3. Define what completes a rotation. A live call, a shadowed support session, or direct participation in a service interaction all count. The format changes with the role; the requirement does not. The person listens to an unfiltered customer instead of reading a summary.
4. Require one written observation per rotation, logged where people can see it. Not a report. One specific observation: a confusion, a workaround, or a sentence that started with “I wish you could” or “I assumed you couldn’t.”
5. Route each observation into the capability-economics process from “Prioritize Work by What It Earns for the Business”. An observation that never becomes a capability statement with an economic category dies in a log nobody reads again, and direct evidence should outweigh a secondhand report.
6. Go first, in public. Say that you completed your own rotation this quarter, and say what you learned. If leadership doesn’t visibly join, the organization reads the program as one leadership doesn’t believe in.

Run the program for two full quarters before you judge it. The value rarely comes from one dramatic discovery. It comes from a steady stream of small course corrections: catching flawed assumptions, adjusting features before they ship, and changing decisions as new customer evidence emerges. Those corrections compound into a product that reflects how customers actually behave rather than how the team assumed they would months earlier.

Notice what the program actually is, because this is where the book has been heading. Everything since the first chapter has argued that as execution gets cheaper, judgment becomes the engineering organization’s scarce asset, and judgment is only as good as what trained it. A rotation program looks like a customer-service gesture and works as the training pipeline for the one capability you cannot generate, summarize, or buy, kept running because every other force in a growing company cuts it off at the source.

All of the playbooks are listed on the playbooks page.