What can one do with cross-sectional relationships?

This is somewhat of a broad question, but I think we all would like to find signals that predict something in the future. However, often times, we are just left with cross-sectional relationships (both variables go up in the same time frame, but lagged terms cannot predict each other).

• What are some things we can do, if we stumble upon a new cross-sectional relationship?
• Originally, I am hoping to find signals for things…can a cross-sectional relationship be turned into a signal or strategy of some sort? Like a signal for entry or exit even though it itself offers no predictability.
• Could it be used in a similar way like the Factor models where there is no predictability and it’s all about some sort of compensation for risk? If so, what’s a good resource to kinda get an understanding on how to convert a cross-sectional relationship into risk premium.

I’m looking for broad ideas or some examples on how correlation may be used in finance. And basically next steps to go.

Thanks!