An organization needs to improve what already works and investigate what might work next. The two activities compete for resources and require different kinds of evidence.
Trouble begins when an early experiment is judged like an established operation, or when an established operation is allowed to remain an experiment indefinitely.
State the stage of the work
For an operating activity, specify the expected result, cost and reliability. For an exploratory activity, specify the question, test and decision the learning will inform.
A small experiment may be worthwhile even when it cannot yet support a full forecast. It should still have limits, an owner and a reason to continue.
Give exploration a route into the business
Agree on what evidence would justify further investment and who can authorize the next commitment. Without that route, successful experiments can remain isolated from the people able to use them.
Protect the operating business from open-ended demands by making the resource tradeoff visible. The work of discovery needs capacity that someone must choose to provide.
Review both activities with appropriate questions. Is the operation delivering? Is the experiment resolving uncertainty? Calling one analytical and the other creative does not remove the need for judgment in either.
Traditional metaphors about complementary qualities may enrich reflection, but organizational design should be assessed through the decisions and work it enables.
