Your question is Pitfalls in Industrial Experiments. 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).
What pitfalls would you watch for when running experiments on industrial workflows with shared teams or equipment at Cognite?
Treat this as a practical experimentation-design exercise. Explain how you would define the hypothesis and metrics, choose the randomization unit, estimate sample size and MDE, analyze the results, and decide whether to ship.