Our quarterly Rock changed halfway through the build. How should our forward-deployed engineer adjust without wasting the work?
Pause long enough to clarify the changed outcome before asking the engineer to add features. A Rock represents a quarterly priority, so a change in direction should be resolved by the accountable leadership team, not quietly absorbed into the build.
Ask the Rock owner to explain what changed, why the original result is no longer the right target, and what must be true by the end of the quarter. Then have the engineer show what is already usable, what is in progress, and which assumptions no longer hold. Separate work that can still serve the revised outcome from work that should stop.
Agree on the smallest useful next deliverable and explicitly identify anything being deferred. If the new direction needs different access, people, or approvals, address those dependencies before promising a date. Record the decision in the team's normal EOS process so the build remains connected to the company's operational priorities.
At the next review, check whether the revised work is actually helping the Rock owner move toward the outcome. A change is not automatically a failure, but repeated changes can signal an issue worth discussing: unclear ownership, an unresolved business decision, or a priority that was not sufficiently defined. Use that discussion to improve the next cycle rather than treating software as a substitute for alignment.
Category: Forward-deployed engineering