Your question is Design Test for Novelty and Spillovers. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You work on a streaming product and are testing a new content discovery experience. Early results look promising, but you’re worried the lift may fade once users get used to the change, or that behavior could spill across users and distort the test.
How would you design the experiment if you were worried about novelty effects or network interference? What would you measure, how would you randomize, and how would you decide whether the result is trustworthy enough to ship?