If you have a powerful compounding product, just make the product more what it wants to be.

Don't overthink it, just do what the people who care about it and have thought about it a lot think is an obvious improvement.

If a company has a product that has compounding value, then if engineers self report "I'm doing the best work of my career" then the company is thriving.

The more you overthink it, the more you'll interfere with the emergent magic of it.

People implicitly assume things that don't have a clean monocausal narrative don't exist, which causes them to miss the power of simply letting engaged employees work on the details they think matter.